On the Brown Moses Blog I was frequently working on authenticating videos, and one very useful technique is using information in the video to confirm the location the video was filmed. In this article, I’ll explain some of the techniques I use to identify those locations, using tools available to anyone with a internet connection.
\n
This first example shows how from a static position it’s possible to establish the precise location a video was filmed. On August 15th 2011, the Libya opposition claimed to have captured the small town of Tiji, posting the following video online as proof
\n\n
The first 4 minutes of the video shows fighting in fairly featureless surroundings, but at 4:00 there’s a shot showing a street with buildings either side.
\n
The first thing to look for are landmarks that might help narrow down the search, and on the same side of the road as the cameraman is a mosque
\n
\n
Tiji is a pretty small town, so there won’t be many mosques, but we can make our search even easier by looking for the kind of road the mosque is on
\n
\n
We can get a good idea what the road looks like from this shot. It has a small median, the lane closest to the camera is about two tanks wide, and has single story buildings across from the mosque.
\n
Now to find Tiji on a map so we can take a look at the roads. A number of sites are available with satellite map imagery, including Google Maps and Bing Maps, but I find Wikimapia very useful as you can select different source satellite map imagery from Google Maps, Bing Maps, and Yahoo! Maps, which can be useful when one set of satellite map imagery isn’t very clear.
\n
Finding Tiji is a simple enough search on Wikimapia, but in some cases with Arab countries you’ll only be able to find Arabic place names, something to keep in mind when searching for more obscure locations. Below we have a satellite view of the town
\n
\n
If we zoom in closer we can see that there’s a number of roads in the centre of the town, but only the main road running through the town appears to have a median
\n
\n
In fact, we can zoom in close enough to see individual cars on the road, which is clearly wide enough to accommodate at least two large vehicles, maybe even three, matching what we can see in the video
\n
\n
So we have a road that would see like a pretty good match to what’s in the video, and we know there’s a mosque on it, so all we have to do now is follow the road until we find a mosque.
\n
\n
Now we have a mosque on the road, we can do more to confirm it’s the correct location. The most obvious feature is the minaret on the south side of the mosque, visible in the satellite map, and frequently a useful point of reference in videos from the Middle East. We can clearly see in the video that the minaret is on the side closest to the road (south), with the mosque’s dome to the right of the minaret (the north). It’s also possible to make out the wall outside of the mosque from the satellite map, and where the road curves northwards
\n
\n
\n
From this we can say the location is almost certainly here
\n
\n
You’ll note to the south of the mosque, across the road, there’s also a number of trees, and these are visible in the background as the tank drives by
\n
\n
It’s also worth checking for any tall buildings in the video, as the satellite map appears to show no buildings that are taller than the mosque, and if there were that would suggest this is the wrong location.
\n
That’s a pretty straight forward example with a distinct landmark, and in a small town. This next video, from August 11th 2011, claimed to show the Libyan opposition taking control of the key town of Brega, with fighters taking a tour of a urban area
\n\n
For a long time Brega was the front-line between the opposition forces based in Benghazi, and loyalist in the Gaddafi stronghold of Sirte. For months, there were constant reports of the opposition capturing the town, but very little evidence to support the claim. Brega itself is a mixture of small residential areas and industrial complexes, spread out over around 10km of coast line. From the video we can see the opposition are in a residential area, made up of rows of single and two storey buildings. This doesn’t match one of the three residential areas in Brega, which appears to be made up of multi-storey buildings that are spaced apart
\n
\n
But it could match the western and eastern residential areas
\n
\n
While the buildings in the video are fairly featureless, there’s something else that we can work from that’s likely to be very unique. During the video, the cameraman walks around Brega, along the roads, and this means that from watching the video we can actually map out the roads. We can see he starts on a sharply curving road, at 42s he turns left, and shows the road he was on met a T-junction, and at 1:20 he comes to another junction. He turns left, walks past a junction on the right hand side of the road, and the road he’s on curves to the left.
\n
\n
But what are we looking for? Simply, the layout of the roads. We can see the roads, the way the curve in some areas, and where the junctions are. You can simply sketch it out on a piece of paper, which I’d recommend as it’s easier to rotate when you’re looking for a match, and then search the streets until you find a match. Here’s a very simple MS Paint sketch of the road layout
\n
\n
We can compare this to the two areas we’re looking at to see if there’s any matching pattern. The western residential area appears to lack the curving roads seen in the video, but the eastern residential area is a different matter
\n
\n
Now you take sketch of the road layout and see if there’s any matches in that area. In the eastern residential areas there’s one location that’s a match
\n
\n
\n
To double-check the location is correct, you watch the video again, and check for anything that matches or doesn’t match. The area is fairly nondescript, but you can see some areas that match what’s in the video. For example, at 2:47, a long open area is visible behind the cameraman
\n
\n
As we know the route he’s walked, we can check the map if there’s an open area in the same position, and as we see below, there is
\n
\n
Currently, these are relatively unknown techniques, but as more and more unverified videos are produced from conflict zones, these kinds of techniques will be key part of the verification process.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nOn the Brown Moses Blog I was frequently working on authenticating videos, and one very useful technique is using information in the video to confirm the location the video was filmed. In this article, I’ll explain some of the techniques I use to identify those locations, using tools available to anyone with a internet connection.\nThis first example shows how from a static position it’s possible to establish the precise location a video was filmed. On August 15th 2011, the Libya opposition claimed to have captured the small town of Tiji, posting the following video online as proof\nThe first 4 minutes of the video shows fighting in fairly featureless surroundings, but at 4:00 there’s a shot showing a street with buildings either side.\nThe first thing to look for are landmarks that might help narrow down the search, and on the same side of the road as the cameraman is a mosque\nTiji is a pretty small town, so there won’t be many mosques, but we can make our search even easier by looking for the kind of road the mosque is on\nWe can get a good idea what the road looks like from this shot. It has a small median, the lane closest to the camera is about two tanks wide, and has single story buildings across from the mosque.\nNow to find Tiji on a map so we can take a look at the roads. A number of sites are available with satellite map imagery, including Google Maps and Bing Maps, but I find Wikimapia very useful as you can select different source satellite map imagery from Google Maps, Bing Maps, and Yahoo! Maps, which can be useful when one set of satellite map imagery isn’t very clear.\nFinding Tiji is a simple enough search on Wikimapia, but in some cases with Arab countries you’ll only be able to find Arabic place names, something to keep in mind when searching for more obscure locations. Below we have a satellite view of the town\nIf we zoom in closer we can see that there’s a number of roads in the centre of the town, but only the main road running through the town appears to have a median\nIn fact, we can zoom in close enough to see individual cars on the road, which is clearly wide enough to accommodate at least two large vehicles, maybe even three, matching what we can see in the video\nSo we have a road that would see like a pretty good match to what’s in the video, and we know there’s a mosque on it, so all we have to do now is follow the road until we find a mosque.\nNow we have a mosque on the road, we can do more to confirm it’s the correct location. The most obvious feature is the minaret on the south side of the mosque, visible in the satellite map, and frequently a useful point of reference in videos from the Middle East. We can clearly see in the video that the minaret is on the side closest to the road (south), with the mosque’s dome to the right of the minaret (the north). It’s also possible to make out the wall outside of the mosque from the satellite map, and where the road curves northwards\nFrom this we can say the location is almost certainly here\nYou’ll note to the south of the mosque, across the road, there’s also a number of trees, and these are visible in the background as the tank drives by\nIt’s also worth checking for any tall buildings in the video, as the satellite map appears to show no buildings that are taller than the mosque, and if there were that would suggest this is the wrong location.\nThat’s a pretty straight forward example with a distinct landmark, and in a small town. This next video, from August 11th 2011, claimed to show the Libyan opposition taking control of the key town of Brega, with fighters taking a tour of a urban area\nFor a long time Brega was the front-line between the opposition forces based in Benghazi, and loyalist in the Gaddafi stronghold of Sirte. For months, there were constant reports of the opposition capturing the town, but very little evidence to support the claim. Brega itself is a mixture of small residential areas and industrial complexes, spread out over around 10km of coast line. From the video we can see the opposition are in a residential area, made up of rows of single and two storey buildings. This doesn’t match one of the three residential areas in Brega, which appears to be made up of multi-storey buildings that are spaced apart\nBut it could match the western and eastern residential areas\nWhile the buildings in the video are fairly featureless, there’s something else that we can work from that’s likely to be very unique. During the video, the cameraman walks around Brega, along the roads, and this means that from watching the video we can actually map out the roads. We can see he starts on a sharply curving road, at 42s he turns left, and shows the road he was on met a T-junction, and at 1:20 he comes to another junction. He turns left, walks past a junction on the right hand side of the road, and the road he’s on curves to the left.\nBut what are we looking for? Simply, the layout of the roads. We can see the roads, the way the curve in some areas, and where the junctions are. You can simply sketch it out on a piece of paper, which I’d recommend as it’s easier to rotate when you’re looking for a match, and then search the streets until you find a match. Here’s a very simple MS Paint sketch of the road layout\nWe can compare this to the two areas we’re looking at to see if there’s any matching pattern. The western residential area appears to lack the curving roads seen in the video, but the eastern residential area is a different matter\nNow you take sketch of the road layout and see if there’s any matches in that area. In the eastern residential areas there’s one location that’s a match\nTo double-check the location is correct, you watch the video again, and check for anything that matches or doesn’t match. The area is fairly nondescript, but you can see some areas that match what’s in the video. For example, at 2:47, a long open area is visible behind the cameraman\nAs we know the route he’s walked, we can check the map if there’s an open area in the same position, and as we see below, there is\nCurrently, these are relatively unknown techniques, but as more and more unverified videos are produced from conflict zones, these kinds of techniques will be key part of the verification process.",
+ "images": [
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/brown1.png",
+ "alt": "brown1",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/brown2.png",
+ "alt": "brown2",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/brown3.png",
+ "alt": "brown3",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/brown5.png",
+ "alt": "brown5",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/brown6.png",
+ "alt": "brown6",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/group7.png",
+ "alt": "group7",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/group8.png",
+ "alt": "group8",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/group9.png",
+ "alt": "group9",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/group11.png",
+ "alt": "group11",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/group12.png",
+ "alt": "group12",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/hello1.png",
+ "alt": "hello1",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/hello2.png",
+ "alt": "hello2",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/hello3.png",
+ "alt": "hello3",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/hello4.png",
+ "alt": "hello4",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/hello5.png",
+ "alt": "hello5",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/hello6.png",
+ "alt": "hello6",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/hello7.png",
+ "alt": "hello7",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/hello8.png",
+ "alt": "hello8",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/hello9.png",
+ "alt": "hello9",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-07-09_verification-and-geolocation-tricks-and-tips-with-google-earth.json b/bellingcat_geolocation_json/2014-07-09_verification-and-geolocation-tricks-and-tips-with-google-earth.json
new file mode 100644
index 0000000000000000000000000000000000000000..968cb5175feaf84c3e89b1d2bd9b707277da47d8
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-07-09_verification-and-geolocation-tricks-and-tips-with-google-earth.json
@@ -0,0 +1,52 @@
+{
+ "slug": "verification-and-geolocation-tricks-and-tips-with-google-earth",
+ "title": "Verification and Geolocation Tricks and Tips with Google Earth",
+ "url": "https://www.bellingcat.com/resources/how-tos/2014/07/09/verification-and-geolocation-tricks-and-tips-with-google-earth/",
+ "date": "2014-07-09T14:04:35+00:00",
+ "modified": "2014-07-09T14:04:45+00:00",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "One piece of software that’s freely available to anyone and has become a very useful tool in the open source investigators toolbox is Google Earth.",
+ "content_html": "
One piece of software that’s freely available to anyone and has become a very useful tool in the open source investigators toolbox is Google Earth. Google Earth shows the entire world pieced together with satellite map imagery, with a number of features that are useful for anyone working to verify and geolocate images from locations across the world. There’s a number of techniques that I’ve used with Google Earth that has made it an essential tool in my own work, and the work of others studying content from conflict zones.
\n
Historical imagery
\n
Under the “view” menu you can find the historical imagery option. This allows you to look at earlier satellite map images stored by Google, which can be a useful tool in a number of ways. A simple example is tracking the destruction caused by conflict, in the below images we see a busy neighbourhood in Damascus wiped out between May and August 2012
\n
\n
\n
On a number of occasions I’ve found myself watching a video where I can see one side of a building, but the most up to date map imagery only shows the other side of the building. You might encounter this situation when you want to look at the way in which windows are arranged on one side of a building to see if it matches what you can see in a video. With historical imagery you can go through older satellite images and sometimes find images taken from a different angle, and even in some cases images of each side of a building, such as the example shown below
\n
\n
In this example I was trying to establish whether or not a partly constructed building matched a building I could see on a satellite map. Using the historical satellite map imagery I could see that the structure was only recently built, and therefore there was a strong possibility it was a recent construction project that may have been abandoned at the conflict took hold
\n
\n
3D Terrain
\n
Google Earth has mapped a great deal of terrain in 3D, and this can be extremely useful when geolocating and verifying videos. If you suspect a video or image was taken facing a certain direction, you can place the Street View cursors on the ground, and get an idea of the hills and mountains you would expect to see in that same location. Below are a couple of examples from videos I’ve examined in the past
\n
\n
\n
Landmarks
\n
In an area where large amounts of video footage is being produced from, and there’s unique structures visible in the videos, it’s possible to speed up the geolocation process by marking each point of interest on a map, and finding an image of it to add to the Google Earth map. For example, below I’ve identified the mosques in part of Damascus, and added an image of the mosque’s minaret to each marker I’ve placed
\n
\n
Any time I see a mosque in a video from that area I can quickly review the images of each mosque, and find a match. If two are visible then it’s possible to deduce which direction the cameraman is filming from, and in many cases find the exact location. It’s possible to do this with any landmark, for example water towers, large communication towers, apartment buildings, etc.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nOne piece of software that’s freely available to anyone and has become a very useful tool in the open source investigators toolbox is Google Earth. Google Earth shows the entire world pieced together with satellite map imagery, with a number of features that are useful for anyone working to verify and geolocate images from locations across the world. There’s a number of techniques that I’ve used with Google Earth that has made it an essential tool in my own work, and the work of others studying content from conflict zones.\nHistorical imagery\nUnder the “view” menu you can find the historical imagery option. This allows you to look at earlier satellite map images stored by Google, which can be a useful tool in a number of ways. A simple example is tracking the destruction caused by conflict, in the below images we see a busy neighbourhood in Damascus wiped out between May and August 2012\nOn a number of occasions I’ve found myself watching a video where I can see one side of a building, but the most up to date map imagery only shows the other side of the building. You might encounter this situation when you want to look at the way in which windows are arranged on one side of a building to see if it matches what you can see in a video. With historical imagery you can go through older satellite images and sometimes find images taken from a different angle, and even in some cases images of each side of a building, such as the example shown below\nIn this example I was trying to establish whether or not a partly constructed building matched a building I could see on a satellite map. Using the historical satellite map imagery I could see that the structure was only recently built, and therefore there was a strong possibility it was a recent construction project that may have been abandoned at the conflict took hold\n3D Terrain\nGoogle Earth has mapped a great deal of terrain in 3D, and this can be extremely useful when geolocating and verifying videos. If you suspect a video or image was taken facing a certain direction, you can place the Street View cursors on the ground, and get an idea of the hills and mountains you would expect to see in that same location. Below are a couple of examples from videos I’ve examined in the past\nLandmarks\nIn an area where large amounts of video footage is being produced from, and there’s unique structures visible in the videos, it’s possible to speed up the geolocation process by marking each point of interest on a map, and finding an image of it to add to the Google Earth map. For example, below I’ve identified the mosques in part of Damascus, and added an image of the mosque’s minaret to each marker I’ve placed\nAny time I see a mosque in a video from that area I can quickly review the images of each mosque, and find a match. If two are visible then it’s possible to deduce which direction the cameraman is filming from, and in many cases find the exact location. It’s possible to do this with any landmark, for example water towers, large communication towers, apartment buildings, etc.",
+ "images": [
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/IMAGE-1.jpg",
+ "alt": "IMAGE 1",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/IMAGE-2.jpg",
+ "alt": "IMAGE 2",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/IMAGE-3.jpg",
+ "alt": "IMAGE 3",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/IMAGE-4.jpg",
+ "alt": "IMAGE 4",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/IMAGE-5-1024x151.jpg",
+ "alt": "IMAGE 5",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/IMAGE-6.jpg",
+ "alt": "IMAGE 6",
+ "caption": ""
+ },
+ {
+ "url": "http://bellingcat.com/app/uploads/2014/07/IMAGE-7.jpg",
+ "alt": "IMAGE 7",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-07-15_geolocation-techniques-mapping-landmarks.json b/bellingcat_geolocation_json/2014-07-15_geolocation-techniques-mapping-landmarks.json
new file mode 100644
index 0000000000000000000000000000000000000000..e5a5f683dd952a308d10a23a6c99be51a4224429
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-07-15_geolocation-techniques-mapping-landmarks.json
@@ -0,0 +1,42 @@
+{
+ "slug": "geolocation-techniques-mapping-landmarks",
+ "title": "Geolocation Techniques – Mapping Landmarks",
+ "url": "https://www.bellingcat.com/resources/how-tos/2014/07/15/geolocation-techniques-mapping-landmarks/",
+ "date": "2014-07-15T20:12:45+00:00",
+ "modified": "",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)While frequently you may find yourself trying to find the location on one video, sometimes you may be faced with the opposite, finding the location of multiple videos. In this article I’ll explain one technique that can be used to make the task of finding the locations of multiple videos from one geographical area […]",
+ "content_html": "
While frequently you may find yourself trying to find the location on one video, sometimes you may be faced with the opposite, finding the location of multiple videos. In this article I’ll explain one technique that can be used to make the task of finding the locations of multiple videos from one geographical area a lot easier.
\n
Damascus – Summer 2013
\n
On August 21st 2013, the district of Zamalka in Damascus was hit by a Sarin attack, killing hundreds. To understand the attack I wanted to build up a picture of events before and after the attack, and as part of that I wanted to review the videos uploaded to YouTube channels used by groups in Zamalka and the surrounding areas.
\n
Dozens, if not hundreds, of videos had been posted in the period I was interested in, and geolocating each one would have been a very time-consuming process. I realised there was one thing I could do to make things a lot more easier for me, and that was cataloguing landmarks, in particular the many mosques in the area.
\n
The following can be done on both Google Earth and Google Maps, but for this example I’ll use Google Earth.
\n
I began by looking at the district of Jobar, to the west of Zamalka
\n
\n
My first task was to find mosques in the area, so for that I turned to Wikimapia. Wikimapia allows users to annotate maps, include adding the names and locations of various structures, including mosques. All I need to do is find the area on the map I’m interested in, search for “mosque”, and it lists the mosques that other users have marked on the map. Then, I just have to hover my cursor over each one to see the location as a red marker.
\n
\n
Now on Google Earth I began to add place marks for each mosque I’ve identified on Wikimapia, a sample of which can be shown below
\n
\n
It’s one thing to have the mosques marked on a map, but quite another to know what they look like. You can only see so much information on satellite maps, but there is another source of images of landmarks online that might not be immediately obvious.
\n
Panoramio is a website that combine maps with photographs tagged to specific locations. This means you can find the location you are interested in, then check if anyone happens to have taken photographs of the landmarks you’re interested in. In this example, I’m on the lookout for mosques and their minarets, and I quickly come across a number of photographs
\n
\n
Google Earth and Google Maps both allow the user to add images to place marks, which appear when the place mark is clicked on, as shown below
\n
\n
This means that there’s now reference images for several very noticeable landmarks, and now when I come to review videos and I spot a minaret, I’m now able to quickly review all the images I have, and I know exactly where they are.
\n
The same process can be used for any sort of landmark, and as I become more familiar with less obvious landmarks in the video I’m watching, such as specific apartment buildings, I can also add place marks with images for those locations. This technique can be used to build a set of reference images for a region that can be referred back to during future geolocation attempts, making it very useful for locating videos in those areas you believe you may revisit in the future, and with long conflicts like Syria, it can be something that’s worth doing early on.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nWhile frequently you may find yourself trying to find the location on one video, sometimes you may be faced with the opposite, finding the location of multiple videos. In this article I’ll explain one technique that can be used to make the task of finding the locations of multiple videos from one geographical area a lot easier.\nDamascus – Summer 2013\nOn August 21st 2013, the district of Zamalka in Damascus was hit by a Sarin attack, killing hundreds. To understand the attack I wanted to build up a picture of events before and after the attack, and as part of that I wanted to review the videos uploaded to YouTube channels used by groups in Zamalka and the surrounding areas.\nDozens, if not hundreds, of videos had been posted in the period I was interested in, and geolocating each one would have been a very time-consuming process. I realised there was one thing I could do to make things a lot more easier for me, and that was cataloguing landmarks, in particular the many mosques in the area.\nThe following can be done on both\nGoogle Earth\nand\nGoogle Maps\n, but for this example I’ll use Google Earth.\nI began by looking at the district of Jobar, to the west of Zamalka\nMy first task was to find mosques in the area, so for that I turned to\nWikimapia\n. Wikimapia allows users to annotate maps, include adding the names and locations of various structures, including mosques. All I need to do is find the area on the map I’m interested in, search for “mosque”, and it lists the mosques that other users have marked on the map. Then, I just have to hover my cursor over each one to see the location as a red marker.\nNow on Google Earth I began to add place marks for each mosque I’ve identified on Wikimapia, a sample of which can be shown below\nIt’s one thing to have the mosques marked on a map, but quite another to know what they look like. You can only see so much information on satellite maps, but there is another source of images of landmarks online that might not be immediately obvious.\nPanoramio\nis a website that combine maps with photographs tagged to specific locations. This means you can find the location you are interested in, then check if anyone happens to have taken photographs of the landmarks you’re interested in. In this example, I’m on the lookout for mosques and their minarets, and I quickly come across a number of photographs\nGoogle Earth and Google Maps both allow the user to add images to place marks, which appear when the place mark is clicked on, as shown below\nThis means that there’s now reference images for several very noticeable landmarks, and now when I come to review videos and I spot a minaret, I’m now able to quickly review all the images I have, and I know exactly where they are.\nThe same process can be used for any sort of landmark, and as I become more familiar with less obvious landmarks in the video I’m watching, such as specific apartment buildings, I can also add place marks with images for those locations. This technique can be used to build a set of reference images for a region that can be referred back to during future geolocation attempts, making it very useful for locating videos in those areas you believe you may revisit in the future, and with long conflicts like Syria, it can be something that’s worth doing early on.",
+ "images": [
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/IMAGE-14.jpg",
+ "alt": "IMAGE 1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/IMAGE-22-1200x580.jpg",
+ "alt": "IMAGE 2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/IMAGE-32.jpg",
+ "alt": "IMAGE 3",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/IMAGE-42-1200x633.jpg",
+ "alt": "IMAGE 4",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/IAMGE-5-1200x623.jpg",
+ "alt": "IAMGE 5",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-07-17_geolocating-the-missile-launcher-linked-to-the-downing-of-mh17.json b/bellingcat_geolocation_json/2014-07-17_geolocating-the-missile-launcher-linked-to-the-downing-of-mh17.json
new file mode 100644
index 0000000000000000000000000000000000000000..a1b03090d2a0a6a7e453c8753bcd6014cfe219c6
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-07-17_geolocating-the-missile-launcher-linked-to-the-downing-of-mh17.json
@@ -0,0 +1,52 @@
+{
+ "slug": "geolocating-the-missile-launcher-linked-to-the-downing-of-mh17",
+ "title": "Geolocating the Missile Launcher Linked to the Downing of MH17",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/07/17/geolocating-the-missile-launcher-linked-to-the-downing-of-mh17/",
+ "date": "2014-07-17T21:37:25+00:00",
+ "modified": "2014-08-12T10:15:59+00:00",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Shortly after reports of flight MH17 being shot down in Ukraine began emerging, the following video was posted online, claiming to show a Buk missile launcher travelling through Snizhne, a rebel held town near the Russian border. The original video was quickly deleted from YouTube for unknown reasons, but one lesson to learn earlier on […]",
+ "content_html": "
Shortly after reports of flight MH17 being shot down in Ukraine began emerging, the following video was posted online, claiming to show a Buk missile launcher travelling through Snizhne, a rebel held town near the Russian border.
\n\n
The original video was quickly deleted from YouTube for unknown reasons, but one lesson to learn earlier on with any video of interest is to immediately download it. A variety tools are available to do this, including Keepvid, which I used in this instance.
\n
The obvious question is whether or not this is actually Snizhne, so the first thing to do is see if there’s any obvious landmarks in the video. In this case, the road is a dual-carriageway with trees in the middle, and looking at satellite images from Snizhne it seems quite unusual. I shared the video with my Twitter followers, asking if anyone could find the area (crowd-sourcing these things can be useful), and quickly several people pointed to an area south of the center of Snizhne
\n
\n
The road appeared to share the same layout, with the trees in the middle, and had the slight turn visible in the video. Based off this, it seemed the camera would have to be north of the road, facing south. Aside from the road layout, something else stood out in the video, the height of the camera. It was clear the camera was positioned much higher than the buildings to the south, suggesting it was either on a hill or on top of a tall building. To test this, I used the ground-level view option in Google Earth, and positioned the cursor on the road, facing north, towards the camera
\n
\n
It seemed the only hills in the area were far in the distance, and by examining the map I was able to see apartment buildings to the north of the road
\n
\n
Based on that I now had the approximate position of the camera, and I was able to identify other features I would expect to see from that vantage point. First, three trees positioned to the north of the bend in the road
\n
\n
South of the bend it’s possible to identify two junctions on the road, points 1 and 2, with one tree visible between the junction on both the video and map. At point 3 the red roof of a house is also visible in both the video and satellite map
\n
\n
Based on this information it seems likely that this is the correct location of what’s claimed to by the Buk missile launcher, heading south away from Snizhne. Based on initial information about the location of the crash site, it appears this location was around 10-15km of the crash site.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nShortly after reports of flight MH17 being shot down in Ukraine began emerging, the following video was posted online, claiming to show a Buk missile launcher travelling through\nSnizhne\n, a\nrebel held town\nnear the Russian border.\nThe original video was quickly deleted from YouTube for unknown reasons, but one lesson to learn earlier on with any video of interest is to immediately download it. A variety tools are available to do this, including\nKeepvid\n, which I used in this instance.\nThe obvious question is whether or not this is actually Snizhne, so the first thing to do is see if there’s any obvious landmarks in the video. In this case, the road is a dual-carriageway with trees in the middle, and looking at satellite images from Snizhne it seems quite unusual. I shared the video with my Twitter followers, asking if anyone could find the area (crowd-sourcing these things can be useful), and quickly several people pointed to an area\nsouth of the center of Snizhne\nThe road appeared to share the same layout, with the trees in the middle, and had the slight turn visible in the video. Based off this, it seemed the camera would have to be north of the road, facing south. Aside from the road layout, something else stood out in the video, the height of the camera. It was clear the camera was positioned much higher than the buildings to the south, suggesting it was either on a hill or on top of a tall building. To test this, I used the ground-level view option in Google Earth, and positioned the cursor on the road, facing north, towards the camera\nIt seemed the only hills in the area were far in the distance, and by examining the map I was able to see apartment buildings to the north of the road\nBased on that I now had the approximate position of the camera, and I was able to identify other features I would expect to see from that vantage point. First, three trees positioned to the north of the bend in the road\nSouth of the bend it’s possible to identify two junctions on the road, points 1 and 2, with one tree visible between the junction on both the video and map. At point 3 the red roof of a house is also visible in both the video and satellite map\nBased on this information it seems likely that this is the correct location of what’s claimed to by the Buk missile launcher, heading south away from Snizhne. Based on initial information about the location of the crash site, it appears this location was around 10-15km of the crash site.",
+ "images": [
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/bukmap1.jpg",
+ "alt": "bukmap1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/bukmap2.jpg",
+ "alt": "bukmap2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/bukmap3.jpg",
+ "alt": "bukmap3",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/bukmap4.jpg",
+ "alt": "bukmap4",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/bukmap5.jpg",
+ "alt": "bukmap5",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/bukmap6.jpg",
+ "alt": "bukmap6",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/bukmap7.jpg",
+ "alt": "bukmap7",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-07-18_identifying-the-location-of-the-mh17-linked-missile-launcher-from-one-photograph.json b/bellingcat_geolocation_json/2014-07-18_identifying-the-location-of-the-mh17-linked-missile-launcher-from-one-photograph.json
new file mode 100644
index 0000000000000000000000000000000000000000..c75bb24a83ff05bb20b27f9ccaa1b8e79e64a2d3
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-07-18_identifying-the-location-of-the-mh17-linked-missile-launcher-from-one-photograph.json
@@ -0,0 +1,47 @@
+{
+ "slug": "identifying-the-location-of-the-mh17-linked-missile-launcher-from-one-photograph",
+ "title": "Identifying the Location of the MH17 Linked Missile Launcher From One Photograph",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/07/18/identifying-the-location-of-the-mh17-linked-missile-launcher-from-one-photograph/",
+ "date": "2014-07-18T21:44:33+00:00",
+ "modified": "2014-08-12T10:15:11+00:00",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)In the aftermath of the MH17 downing in Ukraine various images and videos were shared on social media sites claiming to show the Buk missile launcher that was alleged by some to have been used to shoot down flight MH17. One photograph was shared on social media sites, reportedly showing the missile launcher in Snizhne, […]",
+ "content_html": "
In the aftermath of the MH17 downing in Ukraine various images and videos were shared on social media sites claiming to show the Buk missile launcher that was alleged by some to have been used to shoot down flight MH17. One photograph was shared on social media sites, reportedly showing the missile launcher in Snizhne, close to where a video show a Buk missile launcher was filmed
\n
\n
\n
There doesn’t seem much to go on in this picture, but one of my long term followers, Aric Toler, managed to piece together the location using a variety of open source information, in a way that impressed myself and others who have spent their time geolocating images from conflict zones. I asked him to talk me through the process, and here’s what he told me.
\n
His first step was to identify the store with the yellow facade in the picture. Aric has knowledge of Russian, and he was able to piece together the visible letters on the stores sign, and from that he was able to search for chain stores which contained those characters, which he discovered was Строй Дом. He searched for Строй Дом in Snizhne, but was unable to find any. He searched through Google results for the name, and eventually found a reference to one here, located in the town of Torez. He then found the specific address by search the terms “стройдом” “торез” “50 лет ссср” on Google, the name of the shop, the town, and the street. This found a court document with the name of the shop and it’s address as part of a testimony from a witness. From knowing the address he was then able to identify the location on Yandex Maps
\n
\n
Something else that came up in the searches were videos filmed by someone who had been posting videos of himself driving around Torez
\n
\n\n\n
These videos stated the roads they were filmed on, and include shots of the store from the east heading west, and the north heading south
\n
\n
Visible in both videos were the store and the distinct apartment building with black and white stripes down each corner of the building. It was then possible to follow the path of the car in the videos on satellite maps, confirming the position of further landmarks along the route.
\n
Based on that information it was possible to say the camera was positioned (marked in red) in the forecourt of a garage south of the store (marked in yellow), and the approximate position of the Buk missile launcher (marked in green)
\n
\n
Aric didn’t use any specialised knowledge, beyond his language skills, to do this that would have been beyond anyone else doing the same investigation. All the information was available online, and from a single photograph he was able to establish the exact location of the missile launcher. Speaking as someone who has done this for a while, it’s very difficult to establish a location from one photograph alone, so the fact this was his first attempt to geolocate something like this is really impressive, and sets a great example to anyone who might be thinking about trying some geolocation.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nIn the aftermath of the MH17 downing in Ukraine various images and videos were shared on social media sites claiming to show the Buk missile launcher that was alleged by some to have been used to shoot down flight MH17. One photograph was shared on social media sites, reportedly showing the missile launcher in Snizhne, close to where a video show a Buk missile launcher\nwas filmed\nThere doesn’t seem much to go on in this picture, but one of my long term followers,\nAric Toler\n, managed to piece together the location using a variety of open source information, in a way that impressed myself and others who have spent their time geolocating images from conflict zones. I asked him to talk me through the process, and here’s what he told me.\nHis first step was to identify the store with the yellow facade in the picture. Aric has knowledge of Russian, and he was able to piece together the visible letters on the stores sign, and from that he was able to search for chain stores which contained those characters, which he discovered was Строй Дом. He searched for Строй Дом in Snizhne, but was unable to find any. He searched through Google results for the name, and eventually found a reference to one\nhere\n, located in the town of Torez. He then found the specific address by search the terms “стройдом” “торез” “50 лет ссср” on Google, the name of the shop, the town, and the street. This found\na court document\nwith the name of the shop and it’s address as part of a testimony from a witness. From knowing the address he was then able to identify the location on\nYandex Maps\nSomething else that came up in the searches were videos filmed by someone who had been posting videos of himself driving around Torez\nThese videos stated the roads they were filmed on, and include shots of the store from the east heading west, and the north heading south\nVisible in both videos were the store and the distinct apartment building with black and white stripes down each corner of the building. It was then possible to follow the path of the car in the videos on satellite maps, confirming the position of further landmarks along the route.\nBased on that information it was possible to say the camera was positioned (marked in red) in the forecourt of a garage south of the store (marked in yellow), and the approximate position of the Buk missile launcher (marked in green)\nAric didn’t use any specialised knowledge, beyond his language skills, to do this that would have been beyond anyone else doing the same investigation. All the information was available online, and from a single photograph he was able to establish the exact location of the missile launcher. Speaking as someone who has done this for a while, it’s very difficult to establish a location from one photograph alone, so the fact this was his first attempt to geolocate something like this is really impressive, and sets a great example to anyone who might be thinking about trying some geolocation.",
+ "images": [
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/buk-possible-site.jpg-1200x900.jpg",
+ "alt": "buk possible site.jpg",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/map1.jpg",
+ "alt": "map1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/7Wseuzow.jpg",
+ "alt": "7Wseuzow",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/map3-1200x630.jpg",
+ "alt": "map3",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/map2-1200x621.jpg",
+ "alt": "map2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/07/Map4.jpg",
+ "alt": "Map4",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-08-10_locating-the-rockets-used-during-the-august-21st-sarin-attacks-in-damascus.json b/bellingcat_geolocation_json/2014-08-10_locating-the-rockets-used-during-the-august-21st-sarin-attacks-in-damascus.json
new file mode 100644
index 0000000000000000000000000000000000000000..39fadbb61615e46c83b538fa6b996ad56ba6a736
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-08-10_locating-the-rockets-used-during-the-august-21st-sarin-attacks-in-damascus.json
@@ -0,0 +1,127 @@
+{
+ "slug": "locating-the-rockets-used-during-the-august-21st-sarin-attacks-in-damascus",
+ "title": "Locating the Rockets Used During the August 21st Sarin Attacks in Damascus",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/08/10/locating-the-rockets-used-during-the-august-21st-sarin-attacks-in-damascus/",
+ "date": "2014-08-10T21:23:37+00:00",
+ "modified": "2014-08-12T10:10:56+00:00",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Following the August 21st 2013 Sarin attacks in Damascus a number of images were posted online showing the remains of the rockets used in the attack, along with claims of where they had landed. For example, in this video we have a press conference with Zamalka local council where they show what they claim […]",
+ "content_html": "
Following the August 21st 2013 Sarin attacks in Damascus a number of images were posted online showing the remains of the rockets used in the attack, along with claims of where they had landed. For example, in this video we have a press conference with Zamalka local council where they show what they claim to be impact sites of the rockets
\n
\n
For those of us working to confirm as much information about the August 21st attacks it was possible to use this information to find the exact impact locations of several rockets. The process for identifying the impact locations for each rocket provides great examples of the variety of tools and techniques that an investigator can use when attempt to find specific locations with open source information.
\n
Rocket 1
\n
A number of videos and photographs were posted of one rocket in particular, marked with “197”, an example of which can be seen below
\n\n
It was clear the rocket was in a field next to some apartment buildings, with no roads visible, and we had a rough idea where the rockets had landed (at least according to the local co-ordination committee). A number of people started searching satellite map imagery for potential locations, and it appeared there were very few locations in the area that could have been a possible match. It can be extremely useful to have multiple people searching for the same location to quickly identify possible areas to examine. Often, once I’ve established a general area for a search, I’ll simply asking my followers on Twitter if they can find any potential spots, and generally I’ll receive several suggestions which can each be quickly investigated.
\n
The best match that was found in this instance was a field next to some apartment buildings near the reported impact sites
\n
\n
It was then a case of examining the images we had to see if it was a good fit. For example, the below picture helped establish the position of the rocket in relation to the apartment building, which we could then use to establish the direction other images were taken
\n
\n
In the above satellite image I drew what I believed the line of sight was behind the building, and although only a small area was visible it did confirm there was nothing to the side of the building. It’s important to remember you’re not only looking for what should be there, but also what shouldn’t be there.
\n
This next image was an awkward angle, but it did give a clear view of buildings that were to the north, behind the building closest to the rocket.
\n
\n
Here it was possible to count the windows on the building at point 2 on both the satellite map and in the image, and the void to the right of the building was also a match in both the satellite map and image. It was even possible to see the tops of some trees in the photograph that were visible in the satellite map.
\n
This image shows the view to the north, but slightly to the right of the above image.
\n
\n
Here it’s possible to count the windows at point 2, and even make out the balcony on the corner of the building on the satellite map imagery. Further images to the east and west also show features that match, which allows us to confirm the position this rocket was recorded at. Based on the position of the rocket it appears the rocket would have likely come from the north.
\n
Rocket 2
\n
In this example we had one crucial piece of information, GPS data, that was sent to us by the person who filmed the following footage, originally broadcast by ITN
\n\n
The above video is made up from several clips filmed in different locations, and in this example we’re looking at one impact location, shown in the below clip
\n\n
In the above clip we have views of the location in opposite directions. The GPS information provided was not entirely accurate, and it’s always worth keeping in mind GPS information provided in these circumstances might not always be 100% reliable for a number of reasons. The GPS co-ordinated pointed to the area below
\n
\n
It wasn’t immediately obvious if a wall was present in this image, but using Google Earth historical imagery it was possible to find an earlier image where shadows were being cast by the structures in the area (more details on using Google Earth historical imagery here)
\n
\n
Here it’s possible to make out the shadow cast by the wall, as well as the shadows of the surrounding buildings. The longer the shadows, the taller the structure, so we can see here the buildings surrounding the wall are much taller, something that’s visible in the videos.
\n
It also appears that in the above images the building to the left of the wall is jutting out into the road. In fact, in the video we can just make out that the upper floor juts out over the lower floor
\n
\n
We can also see the buildings to the south of that building, across the road, are not parallel with the road, something visible on satellite map imagery
\n
\n
The view east is also visible in the video
\n
\n
The below image shows the approximate field of view visible in the video, with a large building visible to the east
\n
\n
Using historical imagery it’s possible to find an image that shows the west side of the building, allowing up to count the number of windows and their position on the building in both the video and the satellite map imagery
\n
\n
It’s also interesting to note the rocket appears to have passed through a wall just before impact, suggesting it would have come from the northwest, although there’s a possibility it hit the building to the west of the wall before its final impact, so it would be difficult to be sure exactly where it came from. This rocket is located around 600m north-east of Rocket 1
\n
\n
Rocket 3
\n
This rocket was more of a challenge as its final resting place wasn’t outside where we could expect to see lots of useful landmarks, but inside an apartment building
\n\n
The key with this rocket was establishing the location of the building the video was filmed in. It was first filmed when UN/OPCW inspectors visited the attack sites in Damascus, and by establishing which videos featured the same inspectors it was possible to find the location of the building. There was a lot of footage of the inspectors at work, so we examined the footage filmed showing the inspectors examining the rocket, and compared it to footage filmed outside buildings on the same day. Only one group of videos filmed outside buildings matched, and we were able to make out the following three individuals present both at the inspection site and outside one building
\n
\n
Next was finding the location of the building. There wasn’t too much to go on in the video, but thanks to the VDC and local co-ordination committee we had a number of locations they claimed were impact sites that could be used to direct our investigation. After examining several location, it became apparent this was the site visible in the video
\n
\n
Marked in the above image are three structures that helped verify the video. In red we have a building site that’s visible when the inspectors leave the building
\n
\n
In green, a building with a distinct layout of windows visible in the satellite map imagery
\n
\n
To the south, marked in blue, two apartment buildings
\n
\n
For these apartment buildings we had to go back to historical imagery, which gave us several views of the building
\n
\n
In the top left image it’s possible to count the windows, four rows on one, two rows on the other, just as it is in the video. Based on having established the location it appears the inspectors are coming out of the building to the north of the building marked by the VDC and LCC, so it seems likely this was the building hit by the attack.
\n
This demonstrates how a variety of information can be used to establish important facts that might otherwise be impossible to establish. As I demonstrate in “Identifying Government Positions During The August 21st Sarin Attacks” this information then becomes part of a broader understanding of the attacks, giving us a clearer understanding of where the attacks could have originated from.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nFollowing the August 21st 2013 Sarin attacks in Damascus a number of images were posted online showing the remains of the rockets used in the attack, along with claims of where they had landed. For example, in\nthis video\nwe have a press conference with Zamalka local council where they show what they claim to be impact sites of the rockets\nFor those of us working to confirm as much information about the August 21st attacks it was possible to use this information to find the exact impact locations of several rockets. The process for identifying the impact locations for each rocket provides great examples of the variety of tools and techniques that an investigator can use when attempt to find specific locations with open source information.\nRocket 1\nA number of\nvideos\nand\nphotographs\nwere posted of one rocket in particular, marked with “197”, an example of which can be seen below\nIt was clear the rocket was in a field next to some apartment buildings, with no roads visible, and we had a rough idea where the rockets had landed (at least according to the local co-ordination committee). A number of people started searching satellite map imagery for potential locations, and it appeared there were very few locations in the area that could have been a possible match. It can be extremely useful to have multiple people searching for the same location to quickly identify possible areas to examine. Often, once I’ve established a general area for a search, I’ll simply asking my followers on Twitter if they can find any potential spots, and generally I’ll receive several suggestions which can each be quickly investigated.\nThe best match that was found in this instance was a field next to some apartment buildings near the reported impact sites\nIt was then a case of examining the images we had to see if it was a good fit. For example, the below picture helped establish the position of the rocket in relation to the apartment building, which we could then use to establish the direction other images were taken\nIn the above satellite image I drew what I believed the line of sight was behind the building, and although only a small area was visible it did confirm there was nothing to the side of the building. It’s important to remember you’re not only looking for what should be there, but also what shouldn’t be there.\nThis next image was an awkward angle, but it did give a clear view of buildings that were to the north, behind the building closest to the rocket.\nHere it was possible to count the windows on the building at point 2 on both the satellite map and in the image, and the void to the right of the building was also a match in both the satellite map and image. It was even possible to see the tops of some trees in the photograph that were visible in the satellite map.\nThis image shows the view to the north, but slightly to the right of the above image.\nHere it’s possible to count the windows at point 2, and even make out the balcony on the corner of the building on the satellite map imagery.\nFurther images\nto the east and west also show features that match, which allows us to confirm the position this rocket was recorded at. Based on the position of the rocket it appears the rocket would have likely come from the north.\nRocket 2\nIn this example we had one crucial piece of information, GPS data, that was sent to us by the person who filmed the following footage, originally broadcast by ITN\nThe above video is made up from several clips filmed in different locations, and in this example we’re looking at one impact location, shown in the below clip\nIn the above clip we have views of the location in opposite directions. The GPS information provided was not entirely accurate, and it’s always worth keeping in mind GPS information provided in these circumstances might not always be 100% reliable for a number of reasons. The GPS co-ordinated pointed to the area below\nIt wasn’t immediately obvious if a wall was present in this image, but using Google Earth historical imagery it was possible to find an earlier image where shadows were being cast by the structures in the area (more details on using Google Earth historical imagery\nhere\n)\nHere it’s possible to make out the shadow cast by the wall, as well as the shadows of the surrounding buildings. The longer the shadows, the taller the structure, so we can see here the buildings surrounding the wall are much taller, something that’s visible in the videos.\nIt also appears that in the above images the building to the left of the wall is jutting out into the road. In fact, in the video we can just make out that the upper floor juts out over the lower floor\nWe can also see the buildings to the south of that building, across the road, are not parallel with the road, something visible on satellite map imagery\nThe view east is also visible in the video\nThe below image shows the approximate field of view visible in the video, with a large building visible to the east\nUsing historical imagery it’s possible to find an image that shows the west side of the building, allowing up to count the number of windows and their position on the building in both the video and the satellite map imagery\nIt’s also interesting to note the rocket appears to have passed through a wall just before impact, suggesting it would have come from the northwest, although there’s a possibility it hit the building to the west of the wall before its final impact, so it would be difficult to be sure exactly where it came from. This rocket is located around 600m north-east of Rocket 1\nRocket 3\nThis rocket\nwas more of a challenge as its final resting place wasn’t outside where we could expect to see lots of useful landmarks, but inside an apartment building\nThe key with this rocket was establishing the location of the building the video was filmed in. It was first filmed when UN/OPCW inspectors visited the attack sites in Damascus, and by establishing which videos featured the same inspectors it was possible to find the location of the building. There was a lot of footage of the inspectors at work, so we examined the footage filmed showing the inspectors examining the rocket, and compared it to footage filmed outside buildings on the same day. Only one group of videos filmed outside buildings matched, and we were able to make out the following three individuals present both at the inspection site and outside one building\nNext was finding the location of the building. There wasn’t too much to go on in the video, but thanks to the\nVDC\nand local co-ordination committee we had a number of locations they claimed were impact sites that could be used to direct our investigation. After examining several location, it became apparent this was the site visible in the video\nMarked in the above image are three structures that helped verify the video. In red we have a building site that’s visible when the inspectors leave the building\nIn green, a building with a distinct layout of windows visible in the satellite map imagery\nTo the south, marked in blue, two apartment buildings\nFor these apartment buildings we had to go back to historical imagery, which gave us several views of the building\nIn the top left image it’s possible to count the windows, four rows on one, two rows on the other, just as it is in the video. Based on having established the location it appears the inspectors are coming out of the building to the north of the building marked by the VDC and LCC, so it seems likely this was the building hit by the attack.\nThis demonstrates how a variety of information can be used to establish important facts that might otherwise be impossible to establish. As I demonstrate in “\nIdentifying Government Positions During The August 21st Sarin Attacks\n” this information then becomes part of a broader understanding of the attacks, giving us a clearer understanding of where the attacks could have originated from.",
+ "images": [
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/المؤتمر_الصحفي_للمجلس_المحلي_في_زملكا_بريف_دمشق_-_YouTube-11.jpg",
+ "alt": "المؤتمر_الصحفي_للمجلس_المحلي_في_زملكا_بريف_دمشق_-_YouTube-11",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Rocket-map.jpg",
+ "alt": "Rocket map",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Comp-2.jpg",
+ "alt": "Comp 2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Comp-1-1200x426.jpg",
+ "alt": "Comp 1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Comp-3-1200x387.jpg",
+ "alt": "Comp 3",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/2.jpg",
+ "alt": "2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/1.jpg",
+ "alt": "1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/3-1200x710.jpg",
+ "alt": "3",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/5-1200x667.jpg",
+ "alt": "5",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/4.jpg",
+ "alt": "4",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/6-1200x670.jpg",
+ "alt": "6",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/7.jpg",
+ "alt": "7",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/12.jpg",
+ "alt": "12",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/8-1200x674.jpg",
+ "alt": "8",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/OPCW3-1200x716.jpg",
+ "alt": "OPCW3",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/OPCW2.jpg",
+ "alt": "OPCW2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/OPCW1.jpg",
+ "alt": "OPCW1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/positions1.jpg",
+ "alt": "positions",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/OPCW4-1200x657.jpg",
+ "alt": "OPCW4",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/OCPW5.jpg",
+ "alt": "OCPW5",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/OPCW6-1200x621.jpg",
+ "alt": "OPCW6",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/IMAGE-3.jpg",
+ "alt": "IMAGE-3",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-08-15_how-to-locate-a-secret-pro-russian-training-camp.json b/bellingcat_geolocation_json/2014-08-15_how-to-locate-a-secret-pro-russian-training-camp.json
new file mode 100644
index 0000000000000000000000000000000000000000..9c67fad5d2ece82576994fde8acfed3a38ce487b
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-08-15_how-to-locate-a-secret-pro-russian-training-camp.json
@@ -0,0 +1,162 @@
+{
+ "slug": "how-to-locate-a-secret-pro-russian-training-camp",
+ "title": "How to Locate a “Secret” Pro-Russian Training Camp",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/08/15/how-to-locate-a-secret-pro-russian-training-camp/",
+ "date": "2014-08-15T14:38:02+00:00",
+ "modified": "2014-08-22T12:52:04+00:00",
+ "author": "Bellingcat Investigation Team",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Sometimes when geolocating events we can’t rely solely on imagery from sites like Google Earth and Yandex Maps, but instead, we have to utilize several tools and combine multiple sources to come up with an accurate location and complete assessment. In late May, BBC published a YouTube Video titled “Secret training base for Ukraine’s […]",
+ "content_html": "
Sometimes when geolocating events we can’t rely solely on imagery from sites like Google Earth and Yandex Maps, but instead, we have to utilize several tools and combine multiple sources to come up with an accurate location and complete assessment.
\n
In late May, BBC published a YouTube Video titled “Secret training base for Ukraine’s militias”. The video highlighted a secret militia camp used to train over 100 new recruits seeking to join the fight in eastern Ukraine. The video was obviously created in such a way as to not give away its location and lucky for us, giving open source investigators a decent challenge
\n
\n
The first step to geolocating a location from a video is always to gather as much information about the location to help narrow our search. As we go through this particular video, we can conclude three major points about the location that will get us started in right direction.
\n
1) The camp is most likely near the city of Luhans’k. In this screenshot, we can see recent graduates from the camp standing in a formation in front of the Luhans’k SBU building.
\n
\n
2) Its close enough to the water to allow the rebels to go for a refreshing dip between intense training sessions.
\n
\n
3) Based off the colouring and condition of the shacks in the video, the camp may be located at an old camping ground or children’s camp.
\n
\n
In order to begin our search for a Luhans’k waterfront camping ground we turn to a Wikimapia.org, an open-content collaborative mapping tool which allows users to mark locations on a map similar to the way they would add information into a Wikipedia page. Wikimapia gives us the option to search for certain categories such as military bases, schools, hospitals, and yes… even campsites.
\n
Here is a screen shot of the map with the ‘campsite’ category selected. You can see that we there is a handful of marked locations that are near water.
\n
\n
Now we can go through each location to see if we can either find a match. Here is a campsite right along the water, but it doesn’t have a dock which should be visible in satellite imagery.
\n
\n
The same is true with this one…
\n
\n
And then there’s this one, “Campsite Yaseny”, located approximately 800 meters from the Russian border. Not only does the location make sense, but you can even see a dock in the water the same shape and size as the one in the video. But due to the lack of recent satellite imagery and tree cover, we cannot verify the location using only imagery.
\n
\n
\n
\n
Wikimapia provides users with the ability to add details about the location they created. At this particular location, a user added a note implying the campsite may be the venue for a music festival.
\n
\n
Here is the text Google translated into English.
\n
\n
The user also added a website into the details: venichkina.narod.ru. Let’s investigate further. This is a screen shot from the home page of the website provided… The title of the page Google translates into”Venichkina Rainbow 2010 held!” From this information we can gather than a music festival may have been held at this campsite in 2010. Lucky for us, we can assume there are pictures somewhere on the World Wide Web that may allow us to confirm our location.
\n
\n
We can also assume that someone may post the pictures on the Russian social media site, VKontakte. So we turn to our trusty friend Google and search for “Веничкина радуга vk.com”. Most of the time you will get better results by searching in the target language than you would if you search in English. Anyway, here are the results.
\n
\n
It doesn’t take long to see that the first result is exactly what we were looking for. This VK page seems to be the music festival’s community page where attendees discuss future/pass festivals and of course, post photos.
\n
\n
Now we can view the page’s photo albums and see there are several mentioning the festival in 2010. Let’s take a look and see if any help us verify our location.
\n
\n
And there it is. Same paint scheme on the outside.
\n
\n
Same paint scheme on the inside.
\n
\n
Same three buildings from a different angle.
\n
\n
And then there’s the dock that lead us to this location. I’d say we’ve confirmed the location!
\n
\n
Even though we don’t always have clear satellite imagery, Streetview, Panoramio photos, or may be looking for a campsite deep in the forest along the Russian border, it is in fact possible to find by combining multiple tools and techniques learned here at Bellingcat.
\n
Extra:
\n
Here is another video posted in June showing heavy weapons training at the same training camp… Just 800 meters from the Russian border.
\n
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nSometimes when geolocating events we can’t rely solely on imagery from sites like Google Earth and Yandex Maps, but instead, we have to utilize several tools and combine multiple sources to come up with an accurate location and complete assessment.\nIn late May, BBC published a\nYouTube Video\ntitled “Secret training base for Ukraine’s militias”. The video highlighted a secret militia camp used to train over 100 new recruits seeking to join the fight in eastern Ukraine. The video was obviously created in such a way as to not give away its location and lucky for us, giving open source investigators a decent challenge\nThe first step to geolocating a location from a video is always to gather as much information about the location to help narrow our search. As we go through this particular video, we can conclude three major points about the location that will get us started in right direction.\n1) The camp is most likely near the city of Luhans’k. In this screenshot, we can see recent graduates from the camp standing in a formation in front of the Luhans’k SBU building.\n2) Its close enough to the water to allow the rebels to go for a refreshing dip between intense training sessions.\n3) Based off the colouring and condition of the shacks in the video, the camp may be located at an old camping ground or children’s camp.\nIn order to begin our search for a Luhans’k waterfront camping ground we turn to a\nWikimapia.org\n, an open-content collaborative mapping tool which allows users to mark locations on a map similar to the way they would add information into a Wikipedia page. Wikimapia gives us the option to search for certain categories such as military bases, schools, hospitals, and yes… even campsites.\nHere is a screen shot of the map with the ‘campsite’ category selected. You can see that we there is a handful of marked locations that are near water.\nNow we can go through each location to see if we can either find a match. Here is a campsite right along the water, but it doesn’t have a dock which should be visible in satellite imagery.\nThe same is true with this one…\nAnd then there’s this one, “Campsite Yaseny”, located approximately 800 meters from the Russian border. Not only does the location make sense, but you can even see a dock in the water the same shape and size as the one in the video. But due to the lack of recent satellite imagery and tree cover, we cannot verify the location using only imagery.\nWikimapia provides users with the ability to add details about the location they created. At this particular location, a user added a note implying the campsite may be the venue for a music festival.\nHere is the text Google translated into English.\nThe user also added a website into the details: venichkina.narod.ru. Let’s investigate further. This is a screen shot from the home page of the website provided… The title of the page Google translates into”Venichkina Rainbow 2010 held!” From this information we can gather than a music festival may have been held at this campsite in 2010. Lucky for us, we can assume there are pictures somewhere on the World Wide Web that may allow us to confirm our location.\nWe can also assume that someone may post the pictures on the Russian social media site, VKontakte. So we turn to our trusty friend Google and search for “Веничкина радуга vk.com”. Most of the time you will get better results by searching in the target language than you would if you search in English. Anyway, here are the results.\nIt doesn’t take long to see that the first result is exactly what we were looking for. This VK page seems to be the music festival’s community page where attendees discuss future/pass festivals and of course, post photos.\nNow we can view the page’s photo albums and see there are several mentioning the festival in 2010. Let’s take a look and see if any help us verify our location.\nAnd there it is. Same paint scheme on the outside.\nSame paint scheme on the inside.\nSame three buildings from a different angle.\nAnd then there’s the dock that lead us to this location. I’d say we’ve confirmed the location!\nEven though we don’t always have clear satellite imagery, Streetview, Panoramio photos, or may be looking for a campsite deep in the forest along the Russian border, it is in fact possible to find by combining multiple tools and techniques learned here at Bellingcat.\nExtra:\nHere is another\nvideo\nposted in June showing heavy weapons training at the same training camp… Just 800 meters from the Russian border.",
+ "images": [
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/21.jpg",
+ "alt": "2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/11.jpg",
+ "alt": "1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/31.jpg",
+ "alt": "3",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/41.jpg",
+ "alt": "4",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/51.jpg",
+ "alt": "5",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/61.jpg",
+ "alt": "6",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/71.jpg",
+ "alt": "7",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/81.jpg",
+ "alt": "8",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/9.jpg",
+ "alt": "9",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/10-1200x1000.jpg",
+ "alt": "10",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/111.jpg",
+ "alt": "11",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/121.jpg",
+ "alt": "12",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/13.jpg",
+ "alt": "13",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/14.jpg",
+ "alt": "14",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/15.jpg",
+ "alt": "15",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/16.jpg",
+ "alt": "16",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/17.jpg",
+ "alt": "17",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/19.jpg",
+ "alt": "19",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/18.jpg",
+ "alt": "18",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/211.jpg",
+ "alt": "21",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/20.jpg",
+ "alt": "20",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/23.jpg",
+ "alt": "23",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/22.jpg",
+ "alt": "22",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/25.jpg",
+ "alt": "25",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/24.jpg",
+ "alt": "24",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/29.jpg",
+ "alt": "29",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/28.jpg",
+ "alt": "28",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/27.jpg",
+ "alt": "27",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/26.jpg",
+ "alt": "26",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-08-23_the-hills-of-raqqa-geolocating-the-james-foley-video.json b/bellingcat_geolocation_json/2014-08-23_the-hills-of-raqqa-geolocating-the-james-foley-video.json
new file mode 100644
index 0000000000000000000000000000000000000000..7549346b41cccbde637254730bcebb083937ea95
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-08-23_the-hills-of-raqqa-geolocating-the-james-foley-video.json
@@ -0,0 +1,102 @@
+{
+ "slug": "the-hills-of-raqqa-geolocating-the-james-foley-video",
+ "title": "The Hills of Raqqa - Geolocating the James Foley Video",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/08/23/the-hills-of-raqqa-geolocating-the-james-foley-video/",
+ "date": "2014-08-23T16:46:26+00:00",
+ "modified": "2014-09-03T16:58:30+00:00",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Earlier this week the Guardian reported on claims made by hostages held by the Islamic State, identifying the speaker in the James Foley video as being an Islamic State member based in Raqqa, an Islamic State stronghold in north-central Syria. Based on that claim, and information in the James Foley video it’s possible to narrow down the […]",
+ "content_html": "
Earlier this week the Guardian reported on claims made by hostages held by the Islamic State, identifying the speaker in the James Foley video as being an Islamic State member based in Raqqa, an Islamic State stronghold in north-central Syria. Based on that claim, and information in the James Foley video it’s possible to narrow down the location the video was filmed at.
\n
In the video there’s not a great deal of information, however, it is possible to determine some information from the images.
\n
\n
It appears to be a hilly area, with barely any plants or grass, and in the distance we can see a green plain below the hills. In the below map we can see Raqqa, surrounded by green plains, with hills to the south
\n
\n
The position of the shadows in the video suggest this was filmed in the morning, with the camera pointing northwards. Based on that it seems reasonable to search the hills to the south of Raqqa for a possible location. But what details are available in the video to match to the featureless terrain in the area?
\n
It appears the video was filmed on an area of raised ground, with the edge of the raised area circled in red in the below image.
\n
\n
Also visible is the edge of a track, circled in green, leading to a break in the rocks, circled in purple, which is also visible in this picture.
\n
\n
In this image, filmed face on to James Foley, the break in the rocks is visible on the left, with another small break in the rocks visible on the right.
\n
\n
It also appears that in the distance there’s no obvious structures beyond what may be trees or small individual buildings.
The below images show the track entrance to the area, and you can see in the satellite map images the rock outcrop would be partly covering the entryway, as it does in the video.
\n
\n
Also visible on the satellite map imagery is the raised area the video is filmed on.
\n
\n
Based on that information it appears these are the approximate positions of what we can see in the video.
\n
\n
When the camera is positioned to the right it appears trees are visible in the distance that are a possible match to the trees on the road to the north-west. It’s also worth noting buildings or other structures are not visible on the satellite map or on the video.
\n
\n
\n
In the below images we see the view north, through the break in the hillside.
\n
\n
Again, we see very little beyond fields, and the occasional tree or structure. For example, this tree is roughly 1km directly north of the camera.
\n
\n
Having reviewed locations along the hills south of Raqqa this appears to be the only location which fits what little information is available in the video. Along the hills this appears to be the most isolated spot with any sort of road access, with most other areas with road access appearing to have structures that would be visible in the video. Based on all available information we can say that at the very minimum the James Foley video was filmed in the hills south of Raqqa, and likely at the location shown above, supporting the claims made in the Guardian that Raqqa is the possible location of the remaining hostages.
",
+ "content_text": "Translations:\nEnglish (UK)\nEarlier this week\nthe Guardian reported\non claims made by hostages held by the Islamic State, identifying the speaker in\nthe James Foley video\nas being an Islamic State member based in\nRaqqa\n, an Islamic State stronghold in north-central Syria. Based on that claim, and information in the James Foley video it’s possible to narrow down the location the video was filmed at.\nIn the video there’s not a great deal of information, however, it is possible to determine some information from the images.\nIt appears to be a hilly area, with barely any plants or grass, and in the distance we can see a green plain below the hills. In the below map we can see\nRaqqa\n, surrounded by green plains, with hills to the south\nThe position of the shadows in the video suggest this was filmed in the morning, with the camera pointing northwards. Based on that it seems reasonable to search the hills to the south of Raqqa for a possible location. But what details are available in the video to match to the featureless terrain in the area?\nIt appears the video was filmed on an area of raised ground, with the edge of the raised area circled in red in the below image.\nAlso visible is the edge of a track, circled in green, leading to a break in the rocks, circled in purple, which is also visible in this picture.\nIn this image, filmed face on to James Foley, the break in the rocks is visible on the left, with another small break in the rocks visible on the right.\nIt also appears that in the distance there’s no obvious structures beyond what may be trees or small individual buildings.\nAfter examining the hill in the area,\nI believe this is a likely match for what’s visible in the video\n.\nThe below images show the track entrance to the area, and you can see in the satellite map images the rock outcrop would be partly covering the entryway, as it does in the video.\nAlso visible on the satellite map imagery is the raised area the video is filmed on.\nBased on that information it appears these are the approximate positions of what we can see in the video.\nWhen the camera is positioned to the right it appears trees are visible in the distance that are a possible match to the trees on the road to the north-west. It’s also worth noting buildings or other structures are not visible on the satellite map or on the video.\nIn the below images we see the view north, through the break in the hillside.\nAgain, we see very little beyond fields, and the occasional tree or structure. For example,\nthis tree\nis roughly 1km directly north of the camera.\nHaving reviewed locations along the hills south of Raqqa this appears to be the only location which fits what little information is available in the video. Along the hills this appears to be the most isolated spot with any sort of road access, with most other areas with road access appearing to have structures that would be visible in the video. Based on all available information we can say that at the very minimum the James Foley video was filmed in the hills south of Raqqa, and likely at the location shown above, supporting the claims made in the Guardian that Raqqa is the possible location of the remaining hostages.\nUpdate\nISSE Net has produced\nthis interesting counter-argument\nto the above location being the site of the James Foley execution.",
+ "images": [
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Angle-5.jpg",
+ "alt": "Angle 5",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Sat-map-Raqqa.jpg",
+ "alt": "Sat map Raqqa",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Raised-1.jpg",
+ "alt": "Raised 1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Raised-2.jpg",
+ "alt": "Raised 2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Raised-3.jpg",
+ "alt": "Raised 3",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Break-1.jpg",
+ "alt": "Break 1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Break-2.jpg",
+ "alt": "Break 2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Location-1.jpg",
+ "alt": "Location 1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Jut-1.jpg",
+ "alt": "Jut 1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Jut-2.jpg",
+ "alt": "Jut 2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Outcrop.jpg",
+ "alt": "Outcrop",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Location-2.jpg",
+ "alt": "Location 2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Background-1.jpg",
+ "alt": "Background 1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Background-2.jpg",
+ "alt": "Background 2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Background-3.jpg",
+ "alt": "Background 3",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Background-4.jpg",
+ "alt": "Background 4",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/08/Background-5.jpg",
+ "alt": "Background 5",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-09-29_geolocating-the-mh17-buk-convoy-in-russia.json b/bellingcat_geolocation_json/2014-09-29_geolocating-the-mh17-buk-convoy-in-russia.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f7fc809bf4ea256f2afcb7fff52aec0b8352a59
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-09-29_geolocating-the-mh17-buk-convoy-in-russia.json
@@ -0,0 +1,102 @@
+{
+ "slug": "geolocating-the-mh17-buk-convoy-in-russia",
+ "title": "Geolocating the MH17 Buk Convoy in Russia",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/09/29/geolocating-the-mh17-buk-convoy-in-russia/",
+ "date": "2014-09-29T10:34:52+00:00",
+ "modified": "2014-09-29T11:08:23+00:00",
+ "author": "Veli-Pekka Kivimäki",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Over the last few weeks Bellingcat has been examining footage of a Buk missile launcher filmed inside Russia that matches the Buk missile launcher linked to the down of MH17. The movements of the Buk launcher unit linked to the downing of MH17 were previously examined on Bellingcat, and the reporting relied on several […]",
+ "content_html": "
Over the last few weeks Bellingcat has been examining footage of a Buk missile launcher filmed inside Russia that matches the Buk missile launcher linked to the down of MH17. The movements of the Buk launcher unit linked to the downing of MH17 were previously examined on Bellingcat, and the reporting relied on several videos of a convoy moving through Belgorod oblast in Russia. Further analysis of the videos yields information which can be used to accurately locate where the videos were filmed.
\n
The first video examined was one posted on Instagram, with the location tagged as Neznamovo (Незнамово), south of Stary Oskol and uploaded on June 23rd. First observation upon watching the video is that at the beginning of it the vehicles of the convoy are on a road slightly elevated compared to the vehicles where the video was filmed from, sloping downwards as the clip progresses and finally merging with the main road.
\n
\n
Towards the end of the video, an important clue is revealed: the sign with the letters ‘ОГК’ and ‘АГЗС’, the latter being a Russian shorthand for a gas station. Searching for these terms quickly yields results to the proximity Stary Oskol and Neznamovo, and a Wikimapia link gives us a location where the area features (roads, road markings, buildings) are a strong match to the video. The elevated stretch of the road is an on-ramp to road P-188 heading south. Street View being available for this location, the site can be confirmed with confidence. [Coordinates: 51.233946, 37.940584]
\n
\n
The second video referenced was posted to VK on the morning of June 24th, said to show the convoy in Alexeyevka (Алексеевка), also referencing ‘Magnit’, a major Russian supermarket chain. This narrows down the search area, and section of the road where the trucks are turning give a good reference point to look for in satellite images. A Yandex map search for “Алексеевка Магнит” for gives list of seven locations, one of which is next to a section of road which resembles the area where the trucks can be seen turning.
\n
\n
A closer look confirms this location as a strong candidate, as buildings opposite to the reported Magnit location seem like they could match what was shown on the video. Checking Google Maps reveals Street View is available at this location, making verification easy: the location is identical to what can be seen on the video. [Coordinates: 50.624196, 38.649911]
\n
\n
A third video, reportedly from Stary Oskol, seems to have been filmed at dusk or dawn. Viewing the metadata for the video reveals the upload time to be 04:16 ‘Zulu time’, another term used for Coordinated Universal Time, a.k.a. UTC. During the summer months Ukraine is on daylight savings time, meaning the time difference to UTC is +3 hours, so the local upload time of the video has been 07:16 in the morning of June 24th. The video alone provides a few clues about the shooting location, but it’s useful to first look at the other videos to narrow down the search area.
\n
Video number four shows the convoy of vehicles during the daytime, reportedly at Krasnensky district of Belgorod oblast. The scenery in the video appears to show countryside with scarce buildings, but it does offer some details for analysis. First, there are fairly long stretches of straight road, with a curve at around 1 minute into the video. Near the curve, a pair of bus stops is visible, as well as an intersection and a tall structure to the right. Two minutes in, two large structures are visible to the right of the road.
\n
\n
As the Krasnensky district presents a fairly large search area, the road shape gives a good starting point for narrowing the possibilities down. Looking at the main roadways of the area, a stretch of road near Raskhovets (Расховец) appears to be a possible match. Overview satellite imagery also shows two large structures along the road, and zooming in on the satellite imagery of this area reveals two promising details: a pair of bus stops near an intersection, as well as a tall structure casting a long shadow. As there’s Street View images available of the location, these findings can be verified, and the location definitely matches the video. [Coordinates: 50.902533, 38.458406]
\n
\n
As the third video’s upload time was morning on June 24th, it likely was shot sometime between the Neznamovo and Raskhovets videos. With the knowledge of the shooting location of the previous videos, it’s possible to make estimations of the routes the convoy may have taken between Neznamovo and Raskhovets.
\n
\n
One such suggested route goes through the village of Gorodishche (Городище) in the Starooskolsky district. In the video, some small buildings are visible on both sides of the road, mixed together with trees. Gorodishche matched this description, warranting a further look.
\n
\n
Fortunately, this route is covered by Street View, so a virtual drive-through of the village is possible. One unique feature clearly visible at the beginning of the video is an arch on the building next to where the video was shot, and scrolling through the Street View images, on the east side of the village a building can be found with arches similar to the video. A closer look at the site confirms this to very likely be the shooting location of the video: the fence and buildings on the opposite side of the road match the video, the intersection briefly visible can be seen, and from the frames where the camera is pointing towards the convoy, a nearby church’s dome is visible, with the buildings and treeline also matching the video. [Coordinates: 51.137286, 38.064599]
\n
\n
While not referenced in the published article, Magnitsky had located another video of the same convoy, again from Stary Oskol. The video is fairly short and shaky with lots of panning, but the units are clearly identifiable. With the footage, it was possible to create a rudimentary panorama to get a better sense of the overall area where the video was filmed.
\n
\n
From the footage, a few key details stood out: on the very left of the scene, rails and cables are visible. This seemed to indicate the footage might have been filmed near a light rail or tram station. English Wikipedia confirmed that there was one operating in the city, and Russian Wikipedia provided more details on this. It became apparent the Stary Oskol light rail operated on a single rail line, so checking the assumption would be fairly simple: find the end of the line, and follow it down to see if there are any matching locations.
\n
Following the rail line south, a location can be found on Street View where the brick wall and the building on the right appear to match the video. Further orienting the camera confirms the features match the site exactly, with the video seemingly filmed from the light rail station. [Coordinates: 51.311605, 37.897013]
\n
\n
\n
Exploring the city streets on Street View also gave an indication that the Instagram image of the convoy might be from along the same route. In fact, the picture can be located to approx. 1km of the previous video, based on the red apartment building and pavement features. The picture appears to have been taken from the Eldorado store parking lot. [Coordinates: 51.323770, 37.882489]
\n
\n
Further searches for additional material from the same dates has uncovered another video from Kursk, reportedly from June 23rd, the same date the convoy had been filmed passing through Stary Oskol. The dashcam video is of poor quality and many details are not visible, but composition of the convoy gives a strong indication this may indeed be the same one. In fact, compared to the convoy seen near Raskhovets, the first 16 vehicles seem to be in the same order.
\n
\n
To locate this video, there are a few clues: reported location (Kursk), treelines on both sides of the road, a large blue sign on the side of the road, a crosswalk and a small reddish object on road’s wide shoulder on the right. From overhead satellite imagery, it can be seen that the road A144 on the east side of Kursk is lined by trees, and zooming in on this area further reveals crosswalks on this road. One section of this road has a wide shoulder near an intersection, and having a look with Street View again reveals matching details: a wide shoulder as in the video with a dark orange colored bus stop and a crosswalk next to it, and looking in the opposite direction a big blue sign on the side of the road. [Coordinates: 51.722592, 36.336530]
\n
\n
Putting all these data points together we can plot these points on a map, which would fit on a possible route from Kursk to Alexeevka. Compared to the shortest routing suggested by Google, only one manual adjustment to the route has been made in Stary Oskol, which adds 1km to the route, but avoids smaller roads and turns. The green markers show the locations at which the BUK unit had been filmed as part of the convoy, and the yellow markers show where other vehicles of the same convoy were filmed.
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nOver the last few weeks Bellingcat\nhas been examining footage\nof a Buk missile launcher filmed inside Russia that matches the Buk missile launcher linked to the down of MH17. The movements of the Buk launcher unit linked to the downing of MH17 were previously examined on Bellingcat, and the reporting relied on several videos of a convoy moving through Belgorod oblast in Russia. Further analysis of the videos yields information which can be used to accurately locate where the videos were filmed.\nThe first video examined was one posted on Instagram\n, with the location tagged as Neznamovo (Незнамово), south of Stary Oskol and uploaded on June 23rd. First observation upon watching the video is that at the beginning of it the vehicles of the convoy are on a road slightly elevated compared to the vehicles where the video was filmed from, sloping downwards as the clip progresses and finally merging with the main road.\nTowards the end of the video, an important clue is revealed: the sign with the letters ‘ОГК’ and ‘АГЗС’, the latter being a Russian shorthand for a gas station. Searching for these terms quickly yields results to the proximity Stary Oskol and Neznamovo, and a\nWikimapia link\ngives us a location where the area features (roads, road markings, buildings) are a strong match to the video. The elevated stretch of the road is an on-ramp to road P-188 heading south. Street View being available for this location, the site can be confirmed with confidence. [Coordinates: 51.233946, 37.940584]\nThe second video referenced was posted to VK\non the morning of June 24\nth\n, said to show the convoy in Alexeyevka (Алексеевка), also referencing ‘Magnit’, a major Russian supermarket chain. This narrows down the search area, and section of the road where the trucks are turning give a good reference point to look for in satellite images. A Yandex map search for “Алексеевка Магнит” for gives list of seven locations, one of which is next to a section of road which resembles the area where the trucks can be seen turning.\nA closer look confirms this location as a strong candidate, as buildings opposite to the reported Magnit location seem like they could match what was shown on the video. Checking Google Maps reveals Street View is available at this location, making verification easy: the location is identical to what can be seen on the video. [Coordinates: 50.624196, 38.649911]\nA third video\n, reportedly from Stary Oskol, seems to have been filmed at dusk or dawn.\nViewing the metadata for the video\nreveals the upload time to be 04:16 ‘Zulu time’, another term used for Coordinated Universal Time, a.k.a. UTC. During the summer months Ukraine is on daylight savings time, meaning the time difference to UTC is +3 hours, so the local upload time of the video has been 07:16 in the morning of June 24\nth\n. The video alone provides a few clues about the shooting location, but it’s useful to first look at the other videos to narrow down the search area.\nVideo number four\nshows the convoy of vehicles during the daytime, reportedly at Krasnensky district of Belgorod oblast. The scenery in the video appears to show countryside with scarce buildings, but it does offer some details for analysis. First, there are fairly long stretches of straight road, with a curve at around 1 minute into the video. Near the curve, a pair of bus stops is visible, as well as an intersection and a tall structure to the right. Two minutes in, two large structures are visible to the right of the road.\nAs the Krasnensky district presents a fairly large search area, the road shape gives a good starting point for narrowing the possibilities down. Looking at the main roadways of the area, a stretch of road near Raskhovets (Расховец) appears to be a possible match. Overview satellite imagery also shows two large structures along the road, and zooming in on the satellite imagery of this area reveals two promising details: a pair of bus stops near an intersection, as well as a tall structure casting a long shadow. As there’s Street View images available of the location, these findings can be verified, and the location definitely matches the video. [Coordinates: 50.902533, 38.458406]\nAs the third video\n’s upload time was morning on June 24\nth\n, it likely was shot sometime between the Neznamovo and Raskhovets videos. With the knowledge of the shooting location of the previous videos, it’s possible to make estimations of the routes the convoy may have taken between Neznamovo and Raskhovets.\nOne such suggested route goes through the village of Gorodishche (Городище) in the Starooskolsky district. In the video, some small buildings are visible on both sides of the road, mixed together with trees. Gorodishche matched this description, warranting a further look.\nFortunately, this route is covered by Street View, so a virtual drive-through of the village is possible. One unique feature clearly visible at the beginning of the video is an arch on the building next to where the video was shot, and scrolling through the Street View images, on the east side of the village a building can be found with arches similar to the video. A closer look at the site confirms this to very likely be the shooting location of the video: the fence and buildings on the opposite side of the road match the video, the intersection briefly visible can be seen, and from the frames where the camera is pointing towards the convoy, a nearby church’s dome is visible, with the buildings and treeline also matching the video. [Coordinates: 51.137286, 38.064599]\nWhile not referenced in the published article, Magnitsky had located\nanother video of the same convoy\n, again from Stary Oskol. The video is fairly short and shaky with lots of panning, but the units are clearly identifiable. With the footage, it was possible to create a rudimentary panorama to get a better sense of the overall area where the video was filmed.\nFrom the footage, a few key details stood out: on the very left of the scene, rails and cables are visible. This seemed to indicate the footage might have been filmed near a light rail or tram station.\nEnglish Wikipedia confirmed\nthat there was one operating in the city, and\nRussian Wikipedia provided more details\non this. It became apparent the Stary Oskol light rail operated on a single rail line, so checking the assumption would be fairly simple: find the end of the line, and follow it down to see if there are any matching locations.\nFollowing the rail line south, a location can be found on Street View where the brick wall and the building on the right appear to match the video. Further orienting the camera confirms the features match the site exactly, with the video seemingly filmed from the light rail station. [Coordinates: 51.311605, 37.897013]\nExploring the city streets on Street View also gave an indication that the\nInstagram image\nof the convoy might be from along the same route. In fact, the picture can be located to approx. 1km of the previous video, based on the red apartment building and pavement features. The picture appears to have been taken from the Eldorado store parking lot. [Coordinates: 51.323770, 37.882489]\nFurther searches for additional material from the same dates has uncovered\nanother video from Kursk\n, reportedly from June 23\nrd\n, the same date the convoy had been filmed passing through Stary Oskol. The dashcam video is of poor quality and many details are not visible, but composition of the convoy gives a strong indication this may indeed be the same one. In fact, compared to the convoy seen near Raskhovets, the first 16 vehicles seem to be in the same order.\nTo locate this video, there are a few clues: reported location (Kursk), treelines on both sides of the road, a large blue sign on the side of the road, a crosswalk and a small reddish object on road’s wide shoulder on the right. From overhead satellite imagery, it can be seen that the road A144 on the east side of Kursk is lined by trees, and zooming in on this area further reveals crosswalks on this road. One section of this road has a wide shoulder near an intersection, and having a look with Street View again reveals matching details: a wide shoulder as in the video with a dark orange colored bus stop and a crosswalk next to it, and looking in the opposite direction a big blue sign on the side of the road. [Coordinates: 51.722592, 36.336530]\nPutting all these data points together we can plot these points on a map, which would fit on a possible route from Kursk to Alexeevka. Compared to the shortest routing suggested by Google, only one manual adjustment to the route has been made in Stary Oskol, which adds 1km to the route, but avoids smaller roads and turns. The green markers show the locations at which the BUK unit had been filmed as part of the convoy, and the yellow markers show where other vehicles of the same convoy were filmed.",
+ "images": [
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/16.jpg",
+ "alt": "1",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/2.jpg",
+ "alt": "2",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/31.jpg",
+ "alt": "3",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/41.jpg",
+ "alt": "4",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/52.jpg",
+ "alt": "5",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/6.jpg",
+ "alt": "6",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/71.jpg",
+ "alt": "7",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/81.jpg",
+ "alt": "8",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/91.jpg",
+ "alt": "9",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/101.jpg",
+ "alt": "10",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/111.jpg",
+ "alt": "11",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/121.jpg",
+ "alt": "12",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/131.jpg",
+ "alt": "13",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/141.jpg",
+ "alt": "14",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/151.jpg",
+ "alt": "15",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/161.jpg",
+ "alt": "16",
+ "caption": ""
+ },
+ {
+ "url": "https://bellingcat.com/app/uploads/2014/09/17.jpg",
+ "alt": "17",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-11-07_geolocated-july-buk-convoy-videos-in-russia.json b/bellingcat_geolocation_json/2014-11-07_geolocated-july-buk-convoy-videos-in-russia.json
new file mode 100644
index 0000000000000000000000000000000000000000..ce479ffad10bc5f89e77424e433f5855b8924f41
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-11-07_geolocated-july-buk-convoy-videos-in-russia.json
@@ -0,0 +1,92 @@
+{
+ "slug": "geolocated-july-buk-convoy-videos-in-russia",
+ "title": "Geolocated July BUK convoy videos in Russia",
+ "url": "https://www.bellingcat.com/news/europe/2014/11/07/geolocated-july-buk-convoy-videos-in-russia/",
+ "date": "2014-11-07T22:38:16+00:00",
+ "modified": "2014-11-08T23:53:51+00:00",
+ "author": "Veli-Pekka Kivimäki",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)In July, movements of Buk system elements were again on the move in Russia. The Bellingcat team has discovered and analyzed 12 videos reportedly from between July 19-20. Some of the sightings of individual units have been previously discussed in relation to MH17, but they can now be placed in the greater context of […]",
+ "content_html": "
In July, movements of Buk system elements were again on the move in Russia. The Bellingcat team has discovered and analyzed 12 videos reportedly from between July 19-20. Some of the sightings of individual units have been previously discussed in relation to MH17, but they can now be placed in the greater context of a military convoy’s movements. In addition, there is a newly discovered video of a unit from June’s convoy moving alone through Kamensk-Shaktinsky down a road which leads to the Russian border town of Donetsk.
\n
The first video examined has a self-reported location of “Федосе”, which can be found to refer to Федосеевка, or Fedoseevka, north of Stary Oskol. The location in the video can be found on eastbound P189, which passes through Fedoseevka.
\n
\nAbove, screen shot from the video. Below, location on Street View.
\n
Another video with the reported location of Stary Oskol can be linked to the same exact location, seemingly during the time the convoy was stopped by the side of the road.
\n
\nLeft, two screen shots from the video. Right, same landmarks on Street View.
\n
Video number three is not dashcam footage, but shot while stationary. The location provided is Stary Oskol, and based on the visible poles of the light rail power line can be narrowed down to P188 through Stary Oskol. Landmarks reveal the exact location.
\n
\nAbove, screen shot from the video. Below, location on Street View.
\n
The fourth video of the convoy is again said to be located in Stary Oskol. In this case, the video is long enough to give several landmarks, which can be used to locate the video.
\n
\nAbove, screen shot from the video. Below, location on Street View.
\n
With video number five, we see the reported location as Stary Oskol, but the date marked as July 30th. This is interesting, because we see an exact match in the vehicles of the convoy to all the videos marked as having been filmed on July 19th. Comparing the tarp-covered trailers, we can see an exact match to a July 19th video. It seems highly unlikely two identifical convoys with exactly the same kind of vehicles, including the tarped loads, would have moved through the city 11 days apart, especially when no other footage from July 30th has been found.
\n
\nLeft, screen shot from the video. Right, comparisons to vehicles from July 19th videos.
\n
As for the location of the video, it can be placed at the center of Stary Oskol, with match confirmed on Street View.
\n
\nAbove, screen shot from the video. Below, location on Street View.
\n
The sixth video shows familiar surroundings from previous geolocation work, and the filming location can be found in Stary Oskol.
\n
\nAbove, screen shot from the video. Below, location on Street View.
\n
Moving out of the city, the next video said to be from Stary Oskol can be located to south of the city. The immediate surroundings provide few clues, but by also looking at the view from the rear view mirror, the general area can be narrowed down, and exact the location found.
\n
\nAbove, screen shot from the video. Below, location on Street View.
\n
The eigth video has has quite a suggestive title, talking about a “Buk driven to dismantling”. This would seem unlikely, as the vehicles are in Neznamovo, far from Luhansk, and heading south. The ninth video is related, and filmed in the same area.
\n
\nAbove, screen shot from the video. Below, location on Street View.
\n
\nAbove, screen shot from the video. Below, location on Street View.
\n
Next, as number 10, we have a video reportedly from Alexeevka. The quality of the video is not great, but it does provide an overview of the surrounding terrain, which was useful in matches from satellite imagery. Ultimately, the search leads to the following location, giving a very likely match as the location, south of Alexeyevka, with the vehicles heading east.
\n
\nAbove, location on Street View. Below, screen shot from the video.
\n
The final video of this convoy, number 11, provided no reference location, and was one of the more difficult ones to locate. One tool used was SunCalc, as likely time of the video filmed was known, so the shadows of the vehicles provided a clue about road orientation. Then, the matching road shapes were searched from the likely convoy route based on the June convoys movements, giving a match to the Olkhovatka. Finally, Street View was used to confirm the match.
\n
\nAbove, screen shot from the video. Below, location on Street View.
\n
Summary of the convoy’s movements on a map looks as follows:
\n
\n
The last video discussed here is a very interesting one, as it shows a lone Buk TEL unit with obscured markings traveling alone. The video is short and provides very few clues about the likely location. The reported location, “пос. Южный” or “pos. Yuzhnyy”, is alone ambiguous, as it is a common name for villages in Russia. The necessary context was provided by the uploaders reported location on Russian social media sites, placing him in Kamensk-Shaktinsky. With this additional information, it was possible to find a church listing its location as “пос. Южный г. Каменск-Шахтинский”, on the south side of Kamensk-Shaktinsky, and searching the surroundings there lead to the the location pictured below. The unit was traveling south on E40/M21, which leads to the Russian border town of Donetsk.
\n
\nAbove, screen shot from the video. Below, location on Street View.
\n
Location on a map:
\n
\n
To wrap up, here is a summary table of all the locations.
",
+ "content_text": "Translations:\nEnglish (UK)\nIn July, movements of Buk system elements were again on the move in Russia. The Bellingcat team has discovered and analyzed 12 videos reportedly from between July 19-20. Some of the sightings of individual units have been previously discussed in relation to MH17, but they can now be placed in the greater context of a military convoy’s movements. In addition, there is a newly discovered video of a unit from June’s convoy moving alone through Kamensk-Shaktinsky down a road which leads to the Russian border town of Donetsk.\nThe\nfirst video\nexamined has a self-reported location of “Федосе”, which can be found to refer to Федосеевка, or Fedoseevka, north of Stary Oskol. The location in the video can be found on eastbound P189, which passes through Fedoseevka.\nAbove, screen shot from the video. Below, location on Street View.\nAnother video\nwith the reported location of Stary Oskol can be linked to the same exact location, seemingly during the time the convoy was stopped by the side of the road.\nLeft, two screen shots from the video. Right, same landmarks on Street View.\nVideo number three\nis not dashcam footage, but shot while stationary. The location provided is Stary Oskol, and based on the visible poles of the light rail power line can be narrowed down to P188 through Stary Oskol. Landmarks reveal the exact location.\nAbove, screen shot from the video. Below, location on Street View.\nThe fourth video\nof the convoy is again said to be located in Stary Oskol. In this case, the video is long enough to give several landmarks, which can be used to locate the video.\nAbove, screen shot from the video. Below, location on Street View.\nWith\nvideo number five\n, we see the reported location as Stary Oskol, but the date marked as July 30th. This is interesting, because we see an exact match in the vehicles of the convoy to all the videos marked as having been filmed on July 19th. Comparing the tarp-covered trailers, we can see an exact match to a July 19th video. It seems highly unlikely two identifical convoys with exactly the same kind of vehicles, including the tarped loads, would have moved through the city 11 days apart, especially when no other footage from July 30th has been found.\nLeft, screen shot from the video. Right, comparisons to vehicles from July 19th videos.\nAs for the location of the video, it can be placed at the center of Stary Oskol, with match confirmed on Street View.\nAbove, screen shot from the video. Below, location on Street View.\nThe sixth video\nshows familiar surroundings from previous geolocation work, and the filming location can be found in Stary Oskol.\nAbove, screen shot from the video. Below, location on Street View.\nMoving out of the city,\nthe next video\nsaid to be from Stary Oskol can be located to south of the city. The immediate surroundings provide few clues, but by also looking at the view from the rear view mirror, the general area can be narrowed down, and exact the location found.\nAbove, screen shot from the video. Below, location on Street View.\nThe eigth video\nhas has quite a suggestive title, talking about a “Buk driven to dismantling”. This would seem unlikely, as the vehicles are in Neznamovo, far from Luhansk, and heading south.\nThe ninth video\nis related, and filmed in the same area.\nAbove, screen shot from the video. Below, location on Street View.\nAbove, screen shot from the video. Below, location on Street View.\nNext, as\nnumber 10\n, we have a video reportedly from Alexeevka. The quality of the video is not great, but it does provide an overview of the surrounding terrain, which was useful in matches from satellite imagery. Ultimately, the search leads to the following location, giving a very likely match as the location, south of Alexeyevka, with the vehicles heading east.\nAbove, location on Street View. Below, screen shot from the video.\nThe final video of this convoy\n, number 11, provided no reference location, and was one of the more difficult ones to locate. One tool used was\nSunCalc\n, as likely time of the video filmed was known, so the shadows of the vehicles provided a clue about road orientation. Then, the matching road shapes were searched from the likely convoy route based on the June convoys movements, giving a match to the Olkhovatka. Finally, Street View was used to confirm the match.\nAbove, screen shot from the video. Below, location on Street View.\nSummary of the convoy’s movements on a map looks as follows:\nThe last video\ndiscussed here is a very interesting one, as it shows a lone Buk TEL unit with obscured markings traveling alone. The video is short and provides very few clues about the likely location. The reported location, “пос. Южный” or “pos. Yuzhnyy”, is alone ambiguous, as it is a common name for villages in Russia. The necessary context was provided by the uploaders reported location on Russian social media sites, placing him in Kamensk-Shaktinsky. With this additional information, it was possible to find a church listing its location as “пос. Южный г. Каменск-Шахтинский”, on the south side of Kamensk-Shaktinsky, and searching the surroundings there lead to the the location pictured below. The unit was traveling south on E40/M21, which leads to the Russian border town of Donetsk.\nAbove, screen shot from the video. Below, location on Street View.\nLocation on a map:\nTo wrap up, here is a summary table of all the locations.\n#\nLocation\nCoordinates\n1\nFedoseevka\n51.340488,37.7983139\n2\nFedoseevka\n51.3404795,37.7977605\n3\nStary Oskol\n51.3324537,37.8736081\n4\nStary Oskol\n51.3247281,37.8818285\n5\nStary Oskol\n51.3164918,37.8912864\n6\nStary Oskol\n51.3114993,37.8968766\n7\nStary Oskol\n51.2874736,37.9203471\n8\nNeznamovo\n51.2491165,37.9340878\n9\nNeznamovo\n51.242226,37.9368639\n10\nAlexeyevka\n50.6090816,38.645594\n11\nOlkhovatka\n50.6090816,38.645594\n12\nKamensk-Shakhtinsky\n48.2985814,40.2435022",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-01-fedoseevka.jpg",
+ "alt": "july-01-fedoseevka",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-02-fedoseevka.jpg",
+ "alt": "july-02-fedoseevka",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-03-staryoskol.jpg",
+ "alt": "july-03-staryoskol",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-04-staryoskol.jpg",
+ "alt": "july-04-staryoskol",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/truck-comparison-left-right.png",
+ "alt": "truck-comparison-left-right",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-05-staryoskol.jpg",
+ "alt": "july-05-staryoskol",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-06-staryoskol.jpg",
+ "alt": "july-06-staryoskol",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-07-staryoskol.jpg",
+ "alt": "july-07-staryoskol",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-08-neznamovo.jpg",
+ "alt": "july-08-neznamovo",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-09-neznamovo.jpg",
+ "alt": "july-09-neznamovo",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-11-alexeyevka.jpg",
+ "alt": "july-11-alexeyevka",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-12-olkhovatka.jpg",
+ "alt": "july-12-olkhovatka",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/map-july1.jpg",
+ "alt": "map-july1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/july-10-kamensk-shakhtinsky.jpg",
+ "alt": "july-10-kamensk-shakhtinsky",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/map-july2.jpg",
+ "alt": "map-july2",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-11-07_geolocated-june-buk-convoy-videos-in-russia.json b/bellingcat_geolocation_json/2014-11-07_geolocated-june-buk-convoy-videos-in-russia.json
new file mode 100644
index 0000000000000000000000000000000000000000..97e46e9b9b72b06874f31ae1aef38ff2ea13e5cb
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-11-07_geolocated-june-buk-convoy-videos-in-russia.json
@@ -0,0 +1,72 @@
+{
+ "slug": "geolocated-june-buk-convoy-videos-in-russia",
+ "title": "Geolocated June BUK convoy videos in Russia",
+ "url": "https://www.bellingcat.com/news/2014/11/07/geolocated-june-buk-convoy-videos-in-russia/",
+ "date": "2014-11-07T21:11:57+00:00",
+ "modified": "2014-11-08T23:53:38+00:00",
+ "author": "Veli-Pekka Kivimäki",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)In September, the discovery of six videos of a Buk convoy in Russia was reported on Bellingcat. During the research done by the Bellingcat team, another 10 videos have been discovered from June 23-25, which can be linked to the same convoy. As the basis for the geolocation work, the previous videos from June […]",
+ "content_html": "
In September, the discovery of six videos of a Buk convoy in Russia was reported on Bellingcat. During the research done by the Bellingcat team, another 10 videos have been discovered from June 23-25, which can be linked to the same convoy. As the basis for the geolocation work, the previous videos from June were utilized both as reference material, as well as to give indications of a possible route which would narrow down the search area.
\n
At the end of the post, there is a summary map and a table with links to all the locations.
\n
The first video examined here is one with the reported location of Troitskiy. The town can be found east and slightly north of Stary Oskol. One possible route from Kursk to Stary Oskol would take the convoy down P190 which passes through Troitskiy, making this a possible match. Looking at maps and overhead satellite imagery, it can be seen there are not too many intersections to check. One location in particular has matching features, and jumping into Street View confirms the match.
\n
\nLeft, video from Troitskiy. Right, Google Street View.
\n
The second video is an important one, because it provides a lot of detail on many of the convoy’s vehicles, including clear view of the license plates. The video’s description places it in Stary Oskol, and the video provides many landmarks to make the process of locating the video easier. A few such landmarks are the overpass, car dealerships and light rail track. These place the video on P188, which runs through the city.
\n
\nLeft, video from Stary Oskol. Right, Google Street View. \n
\n
Next, we have a video from YouTube, which doesn’t include location information. It also includes only a single Buk TEL (Transporter Erector Launcher) vehicle, so the link to the bigger convoy also needs to be established. This can be done by looking at the unique markings on the Buk TEL itself, as well as the trailer it’s on. The TEL has paint scratched off in a pattern that matches a TEL seen in our previous video. There are also missiles loaded on both units, the markings and sticker placement on trailer match, and based on frame-by-frame analysis of the video from 0:55 to 1:20 the license plate would appear to match as well.
\n
\nLeft, Buk TEL in Stary Oskol. Right, Buk TEL in YouTube video.
\n
As for the location, the video offers an important clue: a sign on the side of the road. The sign reads “Поповы Дворы”, Popovy Dvory. This can be found to be a location in the Kursk region, leading us to the exact location on road A144 / E38, as pictured below.
\n
\nAbove, screenshot from video. Below, Google Street View.
\n
Number four, we have a video again from Stary Oskol. This video is also along the light rail, which means this video locates along the same path as some previous videos.
\n
\nAbove, two screen shots from the video. Below, two Street View pictures from matching location.
\n
The fifth video has been deleted from VK since the Bellingcat team discovered it, but geolocation was done for it based on an archived copy. Based on previous experience, it was thought that the overall appearance of the roadways matches the Stary Oskol region, and further support this was the apparent rail track on the right of the image. Overhead satellite imagery was examined to find possible matches, and a suitable stretch of road was found just north of Neznamovo.
\n
\nAbove, two screen shots from deleted VK video. Below, two Street View pictures for comparison.
\n
Moving along to number six, we have another video from Stary Oskol, of which many copies have been circulating on VK. One of the videos is uploaded June 24th, meaning it’s the likely close to the source, if not the original. The vehicles in this video again match to the other videos of the June 23-25 convoy, down to the license plates. The video is filmed on the same major road from Stary Oskol, as many of the other videos.
\n
\nAbove, convoy in the video. Below, reference image from Street View.
\n
The 7th video helpful is helpfully titled “Трасса Старый Оскол – ОЭМК”, meaning the light rail track between Stary Oskol and OEMK steel works.
\n
\nAbove, convoy in the video. Below, reference image from Street View.
\n
Number eight is a video from a previously known location. It is showing the convoy vehicles at a turnaround in Alexeyevka, with a familiar building in background from the older Alexeyevka video. In the screen captures below, you can see the back of a DFDS Transport truck on the left, and the front of the same truck at the left edge of the frame in the right picture.
\n
\nLeft, new Alexeyevka video. Right, video reported in September.
\n
Number nine is another video reportedly from Alexeyevka. The video shows a 3-minute stretch of the same road, giving a good understanding of the overall surroundings, and what kind of road shape to look for in overhead satellite imagery. A suitable search area with matching landmarks was found south of the city, and again it was possible to confirm the location with Street View.
\n
\nLeft, new rural Alexeyevka video. Right, Street View.
",
+ "content_text": "Translations:\nEnglish (UK)\nIn September, the discovery of six videos of a Buk convoy in Russia was\nreported on Bellingcat\n. During the research done by the Bellingcat team, another 10 videos have been discovered from June 23-25, which can be linked to the same convoy. As the basis for the geolocation work, the previous videos from June were utilized both as reference material, as well as to give indications of a possible route which would narrow down the search area.\nAt the end of the post, there is a summary map and a table with links to all the locations.\nThe first video\nexamined here is one with the reported location of Troitskiy. The town can be found east and slightly north of Stary Oskol. One possible route from Kursk to Stary Oskol would take the convoy down P190 which passes through Troitskiy, making this a possible match. Looking at maps and overhead satellite imagery, it can be seen there are not too many intersections to check. One location in particular has matching features, and jumping into Street View confirms the match.\nLeft, video from Troitskiy. Right, Google Street View.\nThe second video\nis an important one, because it provides a lot of detail on many of the convoy’s vehicles, including clear view of the license plates. The video’s description places it in Stary Oskol, and the video provides many landmarks to make the process of locating the video easier. A few such landmarks are the overpass, car dealerships and light rail track. These place the video on P188, which runs through the city.\nLeft, video from Stary Oskol. Right, Google Street View.\nNext, we have\na video from YouTube\n, which doesn’t include location information. It also includes only a single Buk TEL (Transporter Erector Launcher) vehicle, so the link to the bigger convoy also needs to be established. This can be done by looking at the unique markings on the Buk TEL itself, as well as the trailer it’s on. The TEL has paint scratched off in a pattern that matches a TEL seen in our previous video. There are also missiles loaded on both units, the markings and sticker placement on trailer match, and based on frame-by-frame analysis of the video from 0:55 to 1:20 the license plate would appear to match as well.\nLeft, Buk TEL in Stary Oskol. Right, Buk TEL in YouTube video.\nAs for the location, the video offers an important clue: a sign on the side of the road. The sign reads “Поповы Дворы”, Popovy Dvory. This can be found to be a location in the Kursk region, leading us to the exact location on road A144 / E38, as pictured below.\nAbove, screenshot from video. Below, Google Street View.\nNumber four\n, we have a video again from Stary Oskol. This video is also along the light rail, which means this video locates along the same path as some previous videos.\nAbove, two screen shots from the video. Below, two Street View pictures from matching location.\nThe fifth video has been deleted from VK since the Bellingcat team discovered it, but geolocation was done for it based on an archived copy. Based on previous experience, it was thought that the overall appearance of the roadways matches the Stary Oskol region, and further support this was the apparent rail track on the right of the image. Overhead satellite imagery was examined to find possible matches, and a suitable stretch of road was found just north of Neznamovo.\nAbove, two screen shots from deleted VK video. Below, two Street View pictures for comparison.\nMoving along to\nnumber six\n, we have another video from Stary Oskol, of which many copies have been circulating on VK. One of the videos is uploaded June 24th, meaning it’s the likely close to the source, if not the original. The vehicles in this video again match to the other videos of the June 23-25 convoy, down to the license plates. The video is filmed on the same major road from Stary Oskol, as many of the other videos.\nAbove, convoy in the video. Below, reference image from Street View.\nThe 7th video\nhelpful is helpfully titled “Трасса Старый Оскол – ОЭМК”, meaning the light rail track between Stary Oskol and OEMK steel works.\nAbove, convoy in the video. Below, reference image from Street View.\nNumber eight\nis a video from a previously known location. It is showing the convoy vehicles at a turnaround in Alexeyevka, with a familiar building in background from the older Alexeyevka video. In the screen captures below, you can see the back of a DFDS Transport truck on the left, and the front of the same truck at the left edge of the frame in the right picture.\nLeft, new Alexeyevka video. Right, video reported in September.\nNumber nine\nis another video reportedly from Alexeyevka. The video shows a 3-minute stretch of the same road, giving a good understanding of the overall surroundings, and what kind of road shape to look for in overhead satellite imagery. A suitable search area with matching landmarks was found south of the city, and again it was possible to confirm the location with Street View.\nLeft, new rural Alexeyevka video. Right, Street View.\nThe tenth video\nis from Millerovo, and is\naddressed in a separate case study\n, walking through the geolocation process.\nSumming up all the locations, we get the following map:\n#\nLocation\nCoordinates\nBuk TELAR 3×2 in video?\n1\nTroitskiy\n51.3590587,37.500688\nYes, 00:24\n2\nStary Oskol\n51.3590587,37.500688\nYes, 02:01\n3\nPopovy Dvory\n51.6544589,36.7921105\nNo\n4\nStary Oskol\n51.3065105,37.9023863\nNo\n5\nStary Oskol\n51.2707467,37.9254813\nNo\n6\nStary Oskol\n51.3203988,37.886651\nNo\n7\nNeznamovo\n51.2441558,37.9365475\nNo\n8\nAlexeyevka\n50.624703,38.6503823\nYes, 6:22\n9\nAlexeyevka\n50.5831713,38.7146552\nYes, 0:52\n10\nMillerovo\n48.902343, 40.444658\nNo\nIn the follow-on post, we will examine a\nBuk convoy’s movements in Russia in July\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/01-troitsky.jpg",
+ "alt": "01-troitsky",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/02-staryoskol.jpg",
+ "alt": "02-staryoskol",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/03-tel-comparison.jpg",
+ "alt": "03-tel-comparison",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/04-popovy-dvory-sign.jpg",
+ "alt": "04-popovy-dvory-sign",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/05-stary-oskol-bldg.jpg",
+ "alt": "05-stary-oskol-bldg",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/06-stary-oskol-deleted.jpg",
+ "alt": "06-stary-oskol-deleted",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/07-stary-bridge.jpg",
+ "alt": "07-stary-bridge",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/08-stary-rail.jpg",
+ "alt": "08-stary-rail",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/09-alexeyevka.jpg",
+ "alt": "09-alexeyevka",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/10-alexeyevka-rural.jpg",
+ "alt": "10-alexeyevka-rural",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/map-june-mb.jpg",
+ "alt": "map-june-mb",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-11-08_geolocating-the-russian-buk-convoy-in-millerovo.json b/bellingcat_geolocation_json/2014-11-08_geolocating-the-russian-buk-convoy-in-millerovo.json
new file mode 100644
index 0000000000000000000000000000000000000000..c364e39efeb2e1a43bf17ec05c9bbcb9912bbbbd
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-11-08_geolocating-the-russian-buk-convoy-in-millerovo.json
@@ -0,0 +1,77 @@
+{
+ "slug": "geolocating-the-russian-buk-convoy-in-millerovo",
+ "title": "Geolocating the June Russian Buk Convoy in Millerovo",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/11/08/geolocating-the-russian-buk-convoy-in-millerovo/",
+ "date": "2014-11-08T23:52:44+00:00",
+ "modified": "",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)One of a series of videos that shows a convoy containing the Buk linked to the downing of MH17 was reportedly filmed near the town of Millerovo. The following demonstrates how it’s possible to use a combination of satellite map imagery and Google Street View imagery to find the exact location the video was […]",
+ "content_html": "
One of a series of videos that shows a convoy containing the Buk linked to the downing of MH17 was reportedly filmed near the town of Millerovo. The following demonstrates how it’s possible to use a combination of satellite map imagery and Google Street View imagery to find the exact location the video was filmed.
\n
The video was originally uploaded here, with the date and time on the camera shown to be 11:31am on June 25th.
\n
Based on details visible in the video it seems clear the video was filmed on this road running south of Millerovo towards the Ukrainian border 25km to the west, with the road in Ukraine continuing to rebel held Luhansk.
\n
Google Streetview in that area ends at the point where the video begins.
\n
There’s a number of elements of the Streetview image and video that matches. For example, here the position of large poles and the shape of the tree line matches
\n
\n
At the start of the video, on the right hand side of the road the small pole with a blue sign is also visible (left three images from the video, right hand image from Google Streetview)
\n
\n
Several elements visible in the Streetview image are also present in the satellite map imagery, and are visible in the video. The matches between the satellite imagery (from Yandax Maps) and the Streetview image is shown below
\n
\n
Green = The billboard that’s is just out of view at the start of the video.
\n
Red = Four large poles, either side of the road.
\n
Yellow = A large sign on the north side of the road.
\n
Pink = Two rows of smaller poles.
\n
8 seconds into the video the large sign on the north side of the road is visible, just before the large poles, and the two rows of smaller poles
\n
Slightly further along the road the large poles are visible, as are individual smaller poles
\n
After 6 of the smaller poles pass by next to the road, the tree line ends and opens up to a large field. It’s possible to count 6 of the small poles in the video and on the satellite map before the tree line ends.
\n
\n
It’s also possible to count the number of smaller poles the car passes, and based on this we can say the car stops at the end of the video at the 9th pole, shown below in red, with a sign on the right hand side of the road also visible, shown below in yellow
\n
\n
\n
Also visible are scattered trees to the left of the car visible on the satellite map imagery
\n
\n
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nOne of a series of videos that shows a convoy containing the Buk linked to the downing of MH17 was reportedly filmed near the town of Millerovo. The following demonstrates how it’s possible to use a combination of satellite map imagery and Google Street View imagery to find the exact location the video was filmed.\nThe video was originally uploaded\nhere\n, with the date and time on the camera shown to be 11:31am on June 25th.\nBased on details visible in the video it seems clear the video was filmed on\nthis road\nrunning south of Millerovo towards the Ukrainian border 25km to the west, with the road in Ukraine continuing to rebel held Luhansk.\nGoogle Streetview\nin that area ends at the point where the video begins.\nThere’s a number of elements of the Streetview image and video that matches. For example, here the position of large poles and the shape of the tree line matches\nAt the start of the video, on the right hand side of the road the small pole with a blue sign is also visible (left three images from the video, right hand image from Google Streetview)\nSeveral elements visible in the Streetview image are also present in the satellite map imagery, and are visible in the video. The matches between the satellite imagery (\nfrom Yandax Maps\n) and the Streetview image is shown below\nGreen = The billboard that’s is just out of view at the start of the video.\nRed = Four large poles, either side of the road.\nYellow = A large sign on the north side of the road.\nPink = Two rows of smaller poles.\n8 seconds into the video the large sign on the north side of the road is visible, just before the large poles, and the two rows of smaller poles\nSlightly further along the road the large poles are visible, as are individual smaller poles\nAfter 6 of the smaller poles pass by next to the road, the tree line ends and opens up to a large field. It’s possible to count 6 of the small poles in the video and on the satellite map before the tree line ends.\nIt’s also possible to count the number of smaller poles the car passes, and based on this we can say the car stops at the end of the video at the 9\nth\npole, shown below in red, with a sign on the right hand side of the road also visible, shown below in yellow\nAlso visible are scattered trees to the left of the car visible on the satellite map imagery",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/1.jpg",
+ "alt": "1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/2.jpg",
+ "alt": "2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/3.jpg",
+ "alt": "3",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/4.jpg",
+ "alt": "4",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/5.jpg",
+ "alt": "5",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/6.jpg",
+ "alt": "6",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/7.jpg",
+ "alt": "7",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/8.jpg",
+ "alt": "8",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/9.jpg",
+ "alt": "9",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/10.jpg",
+ "alt": "10",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/11.jpg",
+ "alt": "11",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/11/12.jpg",
+ "alt": "12",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2014-12-19_geolocating-tunisian-jihadists-in-raqqa.json b/bellingcat_geolocation_json/2014-12-19_geolocating-tunisian-jihadists-in-raqqa.json
new file mode 100644
index 0000000000000000000000000000000000000000..24449594ed3d4f4e1f1f446ec7100dd7d55bfd3c
--- /dev/null
+++ b/bellingcat_geolocation_json/2014-12-19_geolocating-tunisian-jihadists-in-raqqa.json
@@ -0,0 +1,97 @@
+{
+ "slug": "geolocating-tunisian-jihadists-in-raqqa",
+ "title": "Geolocating Tunisian Jihadists in Raqqa",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/12/19/geolocating-tunisian-jihadists-in-raqqa/",
+ "date": "2014-12-19T14:49:37+00:00",
+ "modified": "2014-12-19T20:19:48+00:00",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)This article is available in French here. Earlier this week a video was posted online showing Tunisian Jihadists claiming responsibility for the murders of two Tunisian secular politicians in 2013. It was claimed that the video was filmed in territory controlled by the Islamic State, and using a variety of geolocation techniques it is […]",
+ "content_html": "
Earlier this week a video was posted online showing Tunisian Jihadists claiming responsibility for the murders of two Tunisian secular politicians in 2013. It was claimed that the video was filmed in territory controlled by the Islamic State, and using a variety of geolocation techniques it is possible to find the position the video was filmed.
\n
The video is heavily edited, and includes the statements of the men filmed from multiple angles. 39 seconds into the video there’s a shot of a flag pole with buildings visible in the background
\n
\n
One building was of particular interest, the large red roofed building just behind the flag pole. The same building is also visible at the opening of the video (around 28 seconds onwards), when the four men sit down, revealing more details about the building.
\n
\n
Using simple picture editing software such as Paint.net it’s possible to create a composite image of that scene showing all the buildings in the background, providing a clearer view of the building.
\n
\n
Based on this we can see we are looking for a large building with a red roof, with rectangular windows on the top floor, and at least two floors with arched windows below that.
\n
We then began looking at satellite map imagery of the cities controlled by the Islamic State, starting with Raqqa. We quickly found one large red roofed building, the Faculty of Civil Engineering, and began to look for any images of the building. It was possible to find images of the building on Panoramio, and compare the photographs to the building visible in the video.
\n
\n
300 meters west of this building is a public square with a large flagpole visible, with the shadow cast by it and the flag visible in the below satellite map imagery.
\n
\n
Also from the start of the video is a short section where the men are shown entering the frame from the right, giving a different view from the filming location. It was possible to create another composite image of the buildings in the background using simple picture editing software.
\n
\n
South of the public square satellite map imagery shows a building that matches to what’s visible in the above image. In the below image, the top satellite image shows the red paint around the edge of the roof. In the bottom satellite image, taken from 2011, the north side of the building is visible, and the pattern of windows visible in the video match to that visible in the satellite map image.
\n
\n
To the south-west of the above building is another building visible in the composite image. A number of elements are a clear match, including the position of the windows on the front of the building, and the structure on the top of the building.
\n
\n
Also visible are the trees behind that building, and all of these elements suggest this location is an extremely strong match to what’s visible in the video.
\n
\n
It was also possible to estimate the position of the camera from the above image. As it’s possible to see the exact parts of the two buildings that line up with each other it’s also possible to draw a line between them which points towards the camera’s location. In the below image a line has been drawn between the areas on the building marked red and green, which corresponds with the position on the satellite map imagery.
\n
\n
Following the line northwards along the same heading it points towards the same location west of the Faculty of Civil Engineering that we found earlier. It seems based on this information it’s very likely the video was filmed somewhere in that location.
\n
\n
At this point it’s worth noting an object in the background in the video.
\n
\n
This balcony belongs to the Raqqa Governorate building, is exactly in the same spot as the line we drew from the two buildings in the south, and is shown in the below image.
\n
\n
Based on this information alone it seems very likely the video is filmed on the roof of this building. It’s also possible to find additional information pointing to this location. In the below image we can see the flagpole lines up between the windows 3rd and 4th from the right corner of the top floor of the Faculty of Civil Engineering building.
\n
\n
As with the other buildings it’s possible to draw a line from that position, through the flagpole, pointing to the location of the camera.
\n
\n
\n
This again points to the same location as before, and confirms the video is being filmed from the roof of the Raqqa Governorate building, currently occupied by the Islamic State. It’s also possible to establish this video was filmed after a specific date. In one section of the video an antennae is visible, one of a pair that were just east of the Governorate building
\n
\n
In late September one of the two antennae was destroyed in an air strike, so it’s possible to confirm that this video was filmed not online on the roof of the Raqqa Governorate building, but also after late September 2014.
\n
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nThis article is available in French\nhere\n.\nEarlier this week\na video\nwas posted online showing Tunisian Jihadists claiming responsibility for\nthe murders of two Tunisian secular politicians in 201\n3. It was claimed that the video was filmed in territory controlled by the Islamic State, and using a variety of geolocation techniques it is possible to find the position the video was filmed.\nThe video is heavily edited, and includes the statements of the men filmed from multiple angles. 39 seconds into the video there’s a shot of a flag pole with buildings visible in the background\nOne building was of particular interest, the large red roofed building just behind the flag pole. The same building is also visible at the opening of the video (around 28 seconds onwards), when the four men sit down, revealing more details about the building.\nUsing simple picture editing software such as\nPaint.net\nit’s possible to create a composite image of that scene showing all the buildings in the background, providing a clearer view of the building.\nBased on this we can see we are looking for a large building with a red roof, with rectangular windows on the top floor, and at least two floors with arched windows below that.\nWe then began looking at satellite map imagery of the cities controlled by the Islamic State, starting with Raqqa. We quickly found one large red roofed building,\nthe Faculty of Civil Engineering\n, and began to look for any images of the building. It was possible to\nfind images of the building on Panoramio\n, and compare the photographs to the building visible in the video.\n300 meters west of this building is a public square with a large flagpole visible, with the shadow cast by it and the flag visible in the below satellite map imagery.\nAlso from the start of the video is a short section where the men are shown entering the frame from the right, giving a different view from the filming location. It was possible to create another composite image of the buildings in the background using simple picture editing software.\nSouth of the public square satellite map imagery\nshows a building\nthat matches to what’s visible in the above image. In the below image, the top satellite image shows the red paint around the edge of the roof. In the bottom satellite image, taken from 2011, the north side of the building is visible, and the pattern of windows visible in the video match to that visible in the satellite map image.\nTo the south-west of the above building is\nanother building visible\nin the composite image. A number of elements are a clear match, including the position of the windows on the front of the building, and the structure on the top of the building.\nAlso visible are the trees behind that building, and all of these elements suggest this location is an extremely strong match to what’s visible in the video.\nIt was also possible to estimate the position of the camera from the above image. As it’s possible to see the exact parts of the two buildings that line up with each other it’s also possible to draw a line between them which points towards the camera’s location. In the below image a line has been drawn between the areas on the building marked red and green, which corresponds with the position on the satellite map imagery.\nFollowing the line northwards along the same heading it points towards the same location west of the Faculty of Civil Engineering that we found earlier. It seems based on this information it’s very likely the video was filmed somewhere in that location.\nAt this point it’s worth noting an object in the background in the video.\nThis balcony belongs to the\nRaqqa Governorate building\n, is exactly in the same spot as the line we drew from the two buildings in the south, and is shown in the below image.\nBased on this information alone it seems very likely the video is filmed on the roof of this building. It’s also possible to find additional information pointing to this location. In the below image we can see the flagpole lines up between the windows 3rd and 4th from the right corner of the top floor of the Faculty of Civil Engineering building.\nAs with the other buildings it’s possible to draw a line from that position, through the flagpole, pointing to the location of the camera.\nThis again points to the same location as before, and confirms the video is being filmed from the roof of the Raqqa Governorate building, currently occupied by the Islamic State. It’s also possible to establish this video was filmed after a specific date. In one section of the video an antennae is visible, one of a pair that were\njust east of the Governorate building\nIn late September one of the two antennae\nwas destroyed in an air strike\n, so it’s possible to confirm that this video was filmed not online on the roof of the Raqqa Governorate building, but also after late September 2014.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-1.jpg",
+ "alt": "Tunis 1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-2.jpg",
+ "alt": "Tunis 2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-3.png",
+ "alt": "Tunis 3",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-4.jpg",
+ "alt": "Tunis 4",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-6.jpg",
+ "alt": "Tunis 6",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-5.jpg",
+ "alt": "Tunis 5",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-7.jpg",
+ "alt": "Tunis 7",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-8.jpg",
+ "alt": "Tunis 8",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-9.jpg",
+ "alt": "Tunis 9",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-10.jpg",
+ "alt": "Tunis 10",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-11.jpg",
+ "alt": "Tunis 11",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-12.png",
+ "alt": "Tunis 12",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-13.png",
+ "alt": "Tunis 13",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-14.jpg",
+ "alt": "Tunis 14",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/Tunis-15.jpg",
+ "alt": "Tunis 15",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2014/12/screen_shot_2014-12-19_at_1.48.54_pm_1024.png",
+ "alt": "screen_shot_2014-12-19_at_1.48.54_pm_1024",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-03-03_stunt-geolocation-verifying-the-unverifiable.json b/bellingcat_geolocation_json/2015-03-03_stunt-geolocation-verifying-the-unverifiable.json
new file mode 100644
index 0000000000000000000000000000000000000000..fe6cc1176f18a39310c97a491199db083f0a13c9
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-03-03_stunt-geolocation-verifying-the-unverifiable.json
@@ -0,0 +1,52 @@
+{
+ "slug": "stunt-geolocation-verifying-the-unverifiable",
+ "title": "Stunt Geolocation - Verifying the Unverifiable",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/03/03/stunt-geolocation-verifying-the-unverifiable/",
+ "date": "2015-03-03T09:43:07+00:00",
+ "modified": "2015-03-03T17:15:22+00:00",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "When working on open source and social media investigation there's occasional images that at first seem as if they would be impossible to verify. The following picture, showing a 9S470 Self Propelled Command Post, was shared with us on Twitter.",
+ "content_html": "
When working on open source and social media investigation there’s occasional images that at first seem as if they would be impossible to verify. The following picture, showing a 9S470 Self Propelled Command Post, was shared with us on Twitter.
The image is significant because it shows what appeared to be one of the units that was part of the military convoy seen travelling through Russia in late June 2014 that contained the Buk missile launcher Bellingcat identified as also being seen in Ukraine on July 17th, and linked to the downing of MH17. In the Bellingcat report Origin of the Separatists’ Buk we were able to track the convoy using images from multiple sources as it travelled through Russia to the Ukrainian border.
\n
\n
At the very least this new image would add to the details we had gathered about the route of the convoy, and the original version of the image on VK.com already had location information tagged to it, indicating it was taken in Rossosh, Russia (50.202944, 39.510314). However, despite having co-ordinates it’s still important to match what’s visible in the image with images from the ground, to confirm that the co-ordinates provided are accurate. Google Streetview images of the area are available, but the photograph appears to provide little information that could be used to find its precise location.
\n
\n
However, there is in fact plenty of information in the image that can be used to match the location. First it is possible to establish the position of the camera in relation to the vehicle by examining the road markings. The vehicle is between two sets of dashed lines, the set that is furthest away extending further to the right than the set closest to the camera.
\n
In the Google Streetview imagery the same pattern of dashed lines can be seen, and based on the position of the dashed lines in relation to the solid lines the camera must be on the left side of the below image pointing across the road.
\n
\n
Street markings are not unique, but other elements of the photograph match to what’s visible in the Google Streetview imagery. Barely visible behind the tire of the vehicle is a white pole that’s also visible in the Google Streetview imagery.
\n
\n
Also visible in the top left corner of the photograph is a pole that also matches the Google Streetview imagery.
\n
Individually these features are not uncommon, but there is one element of the photograph that is very unusual, and confirms the exact location of the vehicle. A crack is visible in the photograph, running through the middle dashed line, with a patch of discolouration around it that matches to what’s visible in the Google Streetview imagery.
\n
\n
Unlike other matches in the photograph this pattern of cracks and discolouration would be something that’s very unusual, and combined with the other matching element would establish this as the exact location. This shows that even when a photograph appears to provide no useful information it’s still possible to find the exact location it was taken.
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nWhen working on open source and social media investigation there’s occasional images that at first seem as if they would be impossible to verify. The following picture, showing a 9S470 Self Propelled Command Post, was shared with us on Twitter.\n@vpkivimaki\nAlso found this June 24th photo from Rossosh (50.202944, 39.510314)\nhttps://t.co/uxQfC7qWn5\npic.twitter.com/KU9PS7OpL0\n— surpher (@5urpher)\nMarch 2, 2015\nThe image is significant because it shows what appeared to be one of the units that was part of the military convoy seen travelling through Russia in late June 2014 that contained the Buk missile launcher Bellingcat identified as also being seen in Ukraine on July 17th, and linked to the downing of MH17. In the Bellingcat report\nOrigin of the Separatists’ Buk\nwe were able to track the convoy using images from multiple sources as it travelled through Russia to the Ukrainian border.\nAt the very least this new image would add to the details we had gathered about the route of the convoy, and the\noriginal version of the image\non VK.com already had location information tagged to it, indicating it was taken in Rossosh, Russia (\n50.202944, 39.510314\n). However, despite having co-ordinates it’s still important to match what’s visible in the image with images from the ground, to confirm that the co-ordinates provided are accurate.\nGoogle Streetview images\nof the area are available, but the photograph appears to provide little information that could be used to find its precise location.\nHowever, there is in fact plenty of information in the image that can be used to match the location. First it is possible to establish the position of the camera in relation to the vehicle by examining the road markings. The vehicle is between two sets of dashed lines, the set that is furthest away extending further to the right than the set closest to the camera.\nIn the\nGoogle Streetview imagery\nthe same pattern of dashed lines can be seen, and based on the position of the dashed lines in relation to the solid lines the camera must be on the left side of the below image pointing across the road.\nStreet markings are not unique, but other elements of the photograph match to what’s visible in the Google Streetview imagery. Barely visible behind the tire of the vehicle is a white pole that’s also visible in the\nGoogle Streetview imagery\n.\nAlso visible in the top left corner of the photograph is a pole that also matches the\nGoogle Streetview imagery\n.\nIndividually these features are not uncommon, but there is one element of the photograph that is very unusual, and confirms the exact location of the vehicle. A crack is visible in the photograph, running through the middle dashed line, with a patch of discolouration around it that matches to what’s visible in the\nGoogle Streetview imagery\n.\nUnlike other matches in the photograph this pattern of cracks and discolouration would be something that’s very unusual, and combined with the other matching element would establish this as the exact location. This shows that even when a photograph appears to provide no useful information it’s still possible to find the exact location it was taken.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/03/23.jpg",
+ "alt": "23",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/03/G8NQGeq7u4Y.jpg",
+ "alt": "G8NQGeq7u4Y",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/03/dash.jpg",
+ "alt": "dash",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/03/SV-dash.jpg",
+ "alt": "SV dash",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/03/SV-comp-verge.jpg",
+ "alt": "SV comp verge",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/03/SV-poll-comp.jpg",
+ "alt": "SV poll comp",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/03/Crack.jpg",
+ "alt": "Crack",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-04-12_geolocating-the-walter-scott-shooting.json b/bellingcat_geolocation_json/2015-04-12_geolocating-the-walter-scott-shooting.json
new file mode 100644
index 0000000000000000000000000000000000000000..b1d6364cdca26b2478233db64e580fc1429a58af
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-04-12_geolocating-the-walter-scott-shooting.json
@@ -0,0 +1,97 @@
+{
+ "slug": "geolocating-the-walter-scott-shooting",
+ "title": "Geolocating the Walter Scott Shooting",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/04/12/geolocating-the-walter-scott-shooting/",
+ "date": "2015-04-12T09:09:40+00:00",
+ "modified": "2015-04-12T11:39:41+00:00",
+ "author": "Andrew Haggard",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)On April 4, 2015, a police officer named Michael Slager of the North Chaleston Police Department shot and killed Walter Scott as he attempted to flee from the officer. The case has made national headlines in the United States and has been picked up in international media as well. Since the shooting, several major […]",
+ "content_html": "
On April 4, 2015, a police officer named Michael Slager of the North Chaleston Police Department shot and killed Walter Scott as he attempted to flee from the officer. The case has made national headlines in the United States and has been picked up in international media as well.
\n
Since the shooting, several major events have taken place. Firstly, an eyewitness to the shooting came forward and provided video of the shooting to Mr. Scott’s family, which has been widely seen in the public. Secondly, the South Carolina State Law Enforcement Division (SLED), which is investigating the shooting, released dashboard camera video from Michael Slager’s police car. The final major developments are that Michael Slager has been fired from the police department, arrested, and charged with murder.
\n
Using the videos released by SLED and an eyewitness, it’s possible to determine the location of the shooting. The video below is from the police officer’s dashboard camera.
The police dashboard camera video reveals that Scott pulled into the parking lot of an Advance Auto Parts. A quick search of Advance Auto Parts reveals a couple of possible addresses within the North Charleston jurisdiction.
\n
\n
The location of the traffic stop occurred at Advanced Auto Parts store located at 1945 Remount Road in North Charleston.
\n
\n
From the video, we can tell that Walter Scott was driving Westbound on Remount Road/State Road S-10-13 since he makes a left into the Advance Auto Parts’ parking lot.
\n
After turning into the Advance Auto Parts’ parking lot from the street, Scott turns right and stops the vehicle.
\n
\n
Immediately, there are three things of note in this scene. Firstly and most obviously, the two white sheds in the background (red rectangle); secondly, the fences to the left of the white shed (yellow rectangle); and thirdly, the telephone pole (blue rectangle in front of the vehicle Scott was driving (green rectangle). As well, the brick building (pink arrow) next to the white sheds is worth noting.
\n
\n
All of these buildings can be found in satellite imagery or Google Street View. The green circle in the satellite overview shows the approximate location of the vehicle Scott was driving.
\n
\n
It’s also worth noting the flag of South Carolina is fluttering in the background. What’s more, it looks like the banner on the white shed says “Got Tint?”
\n
\n
Using Google’s Street View service and going in the eastbound direction down Remount Road, it’s possible to find the same flag fluttering in front of a window tinting business.
\n
\n
After exiting the vehicle for the second time, Scott runs from the car, heading left of the camera’s position, which would be south from the location of the car.
\n
\n
A man by the name of Feidin Santana happened to be walking by the location where Officer Michael Slager later shot and killed Walter Scott. Santana pulled out his phone and recorded Officer Slager shooting Walter Scott. Apparently, Santana stopped recording for several minutes and resumed recording as more officers arrived on scene. The videos is graphic in that it shows Walter Scott being shot and his body. Screen captures from these videos will be censored to not show Mr. Scott’s body. \n
\n
Several clues from Santana’s videos also help us to geolocate the near precise location of the shooting.
\n
Firstly, as Scott flees from and is shot by Slager, we can see a white building (dark blue rectangle), which is to the right of Santana and west of vehicle Scott was driving. We can also see there is a fence separating the Santana from Slager and Scott. You can also notice a red sign (pink square).
\n
\n
As well, we can also see another building (orange arrow) at another scene from the video that is to the north of Santana and west of Scott’s car. We can also clearly make out the pathway (red arrow) that links the white building from the previous screen capture with Craig Road. Finally, off to the right-hand side of the video is the Advance Auto Parts store (light blue arrow).
\n
\n
Again, all of these buildings and features can be spotted in Google Earth.
\n
\n
From Google Street View, it’s possible to view the pathway from Craig Road, which is perpendicular to Remount Road.
\n
\n
Based on signs posted at the entrance to the path, it appears the path is actually the driveway to Mega Pawn.
\n
\n
The sign seen behind the white building appears to be the LED sign for Mega Pawn, which can be seen from Rivers Avenue.
\n
\n
Based on all the video evidence available, it appears Scott died approximately 150 – 170 meters from the spot where he was pulled over by Slager. Approximately, 160 meters away from the location Scott’s death was North Charleston’s Fire Station Number Three. That station, like all other fire stations in North Charleston, does not have an ambulance. EMS services in both North Charleston and Charleston, South Carolina are provided by the county.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nOn April 4, 2015, a police officer named Michael Slager of the North Chaleston Police Department shot and killed Walter Scott as he attempted to flee from the officer. The case has made national headlines in the United States and has been picked up in international media as well.\nSince the shooting, several major events have taken place. Firstly, an eyewitness to the shooting came forward and provided video of the shooting to Mr. Scott’s family, which has been widely seen in the public. Secondly, the South Carolina State Law Enforcement Division (SLED), which is investigating the shooting, released dashboard camera video from Michael Slager’s police car. The final major developments are that Michael Slager has been fired from the police department, arrested, and charged with murder.\nUsing the videos released by SLED and an eyewitness, it’s possible to determine the location of the shooting. The video below is from the police officer’s dashboard camera.\nWalter Scott: Dash camera video\nfrom\nThe Post and Courier\non\nVimeo\n.\nThe police dashboard camera video reveals that Scott pulled into the parking lot of an Advance Auto Parts. A quick search of Advance Auto Parts reveals a couple of possible addresses within the North Charleston jurisdiction.\nThe location of the traffic stop occurred at Advanced Auto Parts store located at 1945 Remount Road in North Charleston.\nFrom the video, we can tell that Walter Scott was driving Westbound on Remount Road/State Road S-10-13 since he makes a left into the Advance Auto Parts’ parking lot.\nAfter turning into the Advance Auto Parts’ parking lot from the street, Scott turns right and stops the vehicle.\nImmediately, there are three things of note in this scene. Firstly and most obviously, the two white sheds in the background (red rectangle); secondly, the fences to the left of the white shed (yellow rectangle); and thirdly, the telephone pole (blue rectangle in front of the vehicle Scott was driving (green rectangle). As well, the brick building (pink arrow) next to the white sheds is worth noting.\nAll of these buildings can be found in satellite imagery or Google Street View. The green circle in the satellite overview shows the approximate location of the vehicle Scott was driving.\nIt’s also worth noting the flag of South Carolina is fluttering in the background. What’s more, it looks like the banner on the white shed says “Got Tint?”\nUsing Google’s Street View service and going in the eastbound direction down Remount Road, it’s possible to find the same flag fluttering in front of a window tinting business.\nAfter exiting the vehicle for the second time, Scott runs from the car, heading left of the camera’s position, which would be south from the location of the car.\nA man by the name of Feidin Santana happened to be walking by the location where Officer Michael Slager later shot and killed Walter Scott. Santana pulled out his phone and recorded Officer Slager shooting Walter Scott. Apparently, Santana stopped recording for several minutes and resumed recording as more officers arrived on scene. The videos is graphic in that it shows Walter Scott being shot and his body. Screen captures from these videos will be censored to not show Mr. Scott’s body.\nSeveral clues from Santana’s videos also help us to geolocate the near precise location of the shooting.\nFirstly, as Scott flees from and is shot by Slager, we can see a white building (dark blue rectangle), which is to the right of Santana and west of vehicle Scott was driving. We can also see there is a fence separating the Santana from Slager and Scott. You can also notice a red sign (pink square).\nAs well, we can also see another building (orange arrow) at another scene from the video that is to the north of Santana and west of Scott’s car. We can also clearly make out the pathway (red arrow) that links the white building from the previous screen capture with Craig Road. Finally, off to the right-hand side of the video is the Advance Auto Parts store (light blue arrow).\nAgain, all of these buildings and features can be spotted in Google Earth.\nFrom Google Street View, it’s possible to view the pathway from Craig Road, which is perpendicular to Remount Road.\nBased on signs posted at the entrance to the path, it appears the path is actually the driveway to Mega Pawn.\nThe sign seen behind the white building appears to be the LED sign for Mega Pawn, which can be seen from Rivers Avenue.\nBased on all the video evidence available, it appears Scott died approximately 150 – 170 meters from the spot where he was pulled over by Slager. Approximately, 160 meters away from the location Scott’s death was North Charleston’s Fire Station Number Three.\nThat station, like all other fire stations in North Charleston, does not have an ambulance\n. EMS services in both North Charleston and Charleston, South Carolina are provided by the county.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/1.jpg",
+ "alt": "1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/2.jpg",
+ "alt": "2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/3.jpg",
+ "alt": "3",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/4.jpg",
+ "alt": "4",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/5.jpg",
+ "alt": "5",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/6.jpg",
+ "alt": "6",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/7.jpg",
+ "alt": "7",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/8.jpg",
+ "alt": "8",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/9.jpg",
+ "alt": "9",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/10.jpg",
+ "alt": "10",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/11.jpg",
+ "alt": "11",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/12.jpg",
+ "alt": "12",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/13.jpg",
+ "alt": "13",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/14.jpg",
+ "alt": "14",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/15.jpg",
+ "alt": "15",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/04/16.jpg",
+ "alt": "16",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-05-21_air-rescue-in-ramadi-geolocating-made-easy-with-heads-up-display.json b/bellingcat_geolocation_json/2015-05-21_air-rescue-in-ramadi-geolocating-made-easy-with-heads-up-display.json
new file mode 100644
index 0000000000000000000000000000000000000000..48f949bbd3ac37c449d33dbde81cd6635b02ad2f
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-05-21_air-rescue-in-ramadi-geolocating-made-easy-with-heads-up-display.json
@@ -0,0 +1,87 @@
+{
+ "slug": "air-rescue-in-ramadi-geolocating-made-easy-with-heads-up-display",
+ "title": "Air Rescue in Ramadi: Geolocating Made Easy with Head-Up Display",
+ "url": "https://www.bellingcat.com/news/2015/05/21/air-rescue-in-ramadi-geolocating-made-easy-with-heads-up-display/",
+ "date": "2015-05-21T12:39:04+00:00",
+ "modified": "2015-05-21T17:49:50+00:00",
+ "author": "Nathan Patin",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK) The strategic Iraqi city of Ramadi has fallen to the Islamic State (IS), bringing the number of provincial capitals under the militant group’s control to three (the others being Mosul, Iraq and Raqqa, Syria). IS reportedly used the cover provided by a dust storm to gain a military advantage during the start of […]",
+ "content_html": "
The strategic Iraqi city of Ramadi has fallen to the Islamic State (IS), bringing the number of provincial capitals under the militant group’s control to three (the others being Mosul, Iraq and Raqqa, Syria).
\n
IS reportedly used the cover provided by a dust storm to gain a military advantage during the start of their assault last week; American warplanes were unable to provide air support for the Iraqi forces under attack owing to poor visibility. The momentum gained from this tactical maneuver culminated on Sunday with the militants overrunning the Anbar Operations Command and 8th Brigade Headquarters at Camp Ar-Ramadi, the last remaining government strongholds in the city.
\n
Iraqi security forces were seen fleeing the IS onslaught in dozens of Humvees and other military vehicles, scenes distressingly reminiscent of the fall of Mosul nearly a year ago. Commenting on the rout, Chairman of the Joint Chiefs of Staff General Martin Dempsey didn’t mince words, saying on Wednesday that Iraqi forces weren’t “driven from” Ramadi, they “drove out of Ramadi.”
\n
So, with yet another defeat at the hands of IS, it’s perhaps not surprising that the Iraqi Ministry of Defense (MoD) might want to put a good face on a situation that is nothing less than a serious setback. On May 19th, the Iraqi MoD posted to its YouTube channel footage purporting to be of a successful Iraqi Air Force rescue mission of Iraqi security forces trapped in Ramadi. While not exactly a propaganda coup, when your forces have once again melted away in the face of adversity, it’s arguably better than nothing.
\n\n
Interestingly, the MoD doesn’t redact any of the data on the head-up display (HUD) overlaid on the video. Here’s what these videos typically look like:
The HUD overlay in the Ramadi air rescue video, by contrast, is completely visible, save for the date and time in the top-left corner:
\n
\n
Even the date and time, however, can be seen on the screenshot in the Iraqi MoD’s press release:
\n
\n
The 18th of May was Monday, which is consistent with the reported timeline of events. (Corroborating the date of the rescue may not seem particularly important, but the Iraqi MoD has been known to stretch the truth to boost morale.)
\n
Because the HUD data isn’t redacted, it’s possible to geolocate exactly where the Iraqi security forces were rescued with the longitude and latitude visible in the bottom-right corner of the image. Because those coordinates refer to the target (“TGT”) in the HUD crosshairs, we have to wait for the security forces to come into view to determine their location.
\n
The first 30-odd seconds of the video are spent searching for the security forces until the camera zooms in on a group of individuals who signal the aircraft:
\n
\n
Now that the stranded security forces have been spotted, we can use the TGT coordinates (33°24’49N 43°20’47E) to determine their location. But before doing so, it’s helpful to get a better idea of what the location looks like from the video. At 2:46, the video zooms out to reveal that the security forces are situated in the corner of a facility with a number of large, circular structures:
\n
\n
Entering the aforementioned coordinates into Google Earth takes us to this location, which looks somewhat similar but lacks the circular structures seen in the video:
\n
\n
It turns out that this is because Google Earth’s imagery was taken in 2010. Bing, on the other hand, has more recent satellite imagery of the area:
\n
\n
Here the circular structures present in video are clearly visible, confirming that the TGT coordinates indeed mark the exact location where the rescue mission took place.
\n
It’s also interesting to note that the newer Bing imagery, which was taken in August 2013, can be seen side-by-side with Bing’s older imagery taken in August 2011, which is still more recent than the imagery used by Google Earth for this location:
\n
\n
Here’s the rendezvous location in relation to the rest of Ramadi:
\n
\n
The Iraqi MoD describes the rescue as taking place “in the district of the stadium”بمنطقة الملعب)). Indeed, that neighborhood is less than a mile from the site of the rescue operation.
\n
The video ends with understandably relieved Iraqis embracing one another and trotting down an airstrip. Once again, we can use the TGT coordinates (33°20’31N 43°35’58E) and Wikimapia to determine which airfield the rescued Iraqis were flown to.
\n
\n
As would be expected, the mission was conducted from al-Taqaddum Airbase, roughly 15 miles away:
\n
\n
Finally, the heartfelt coda on the runway provides the best view of the helicopters flown during the operation, which appear to be Mil Mi-17s.
\n
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nThe strategic Iraqi city of Ramadi has fallen to the Islamic State (IS), bringing the number of provincial capitals under the militant group’s control to three (the others being Mosul, Iraq and Raqqa, Syria).\nIS reportedly\nused\nthe cover provided by a dust storm to gain a military advantage during the start of their assault last week; American warplanes were unable to provide air support for the Iraqi forces under attack owing to poor visibility. The momentum gained from this tactical maneuver\nculminated\non Sunday with the militants overrunning the Anbar Operations Command and 8\nth\nBrigade Headquarters at Camp Ar-Ramadi, the last remaining government strongholds in the city.\nIraqi security forces were seen\nfleeing\nthe IS onslaught in dozens of Humvees and other military vehicles, scenes distressingly\nreminiscent\nof the fall of Mosul nearly a year ago. Commenting on the rout, Chairman of the Joint Chiefs of Staff General Martin Dempsey didn’t\nmince\nwords, saying on Wednesday that Iraqi forces weren’t “driven from” Ramadi, they “drove out of Ramadi.”\nSo, with yet another defeat at the hands of IS, it’s perhaps not surprising that the Iraqi Ministry of Defense (MoD) might want to put a good face on a situation that is nothing less than a serious setback. On May 19\nth\n, the Iraqi MoD posted to its YouTube channel footage purporting to be of a successful\nIraqi Air Force\nrescue mission of Iraqi security forces trapped in Ramadi. While not exactly a propaganda coup, when your forces have once again melted away in the face of adversity, it’s arguably better than nothing.\nInterestingly, the MoD doesn’t redact any of the data on the\nhead-up display\n(HUD) overlaid on the video. Here’s what these videos typically look like:\nUS Central Command,\nearlier this month\n:\nIraqi MoD,\nlast month\n:\nThe HUD overlay in the Ramadi air rescue video, by contrast, is completely visible, save for the date and time in the top-left corner:\nEven the date and time, however, can be seen on the screenshot in the Iraqi MoD’s\npress release\n:\nThe 18\nth\nof May was Monday, which is consistent with the reported timeline of events. (Corroborating the date of the rescue may not seem particularly important, but the Iraqi MoD has been known to\nstretch the truth\nto boost morale.)\nBecause the HUD data isn’t redacted, it’s possible to geolocate exactly where the Iraqi security forces were rescued with the longitude and latitude visible in the bottom-right corner of the image. Because those coordinates refer to the target (“TGT”) in the HUD crosshairs, we have to wait for the security forces to come into view to determine their location.\nThe first 30-odd seconds of the video are spent searching for the security forces until the camera zooms in on a group of individuals who signal the aircraft:\nNow that the stranded security forces have been spotted, we can use the TGT coordinates (33°24’49N 43°20’47E) to determine their location. But before doing so, it’s helpful to get a better idea of what the location looks like from the video. At 2:46, the video zooms out to reveal that the security forces are situated in the corner of a facility with a number of large, circular structures:\nEntering the aforementioned coordinates into Google Earth takes us to this location, which looks somewhat similar but lacks the circular structures seen in the video:\nIt turns out that this is because Google Earth’s imagery was taken in 2010. Bing, on the other hand, has more recent satellite imagery of the area:\nHere the circular structures present in video are clearly visible, confirming that the TGT coordinates indeed mark the exact location where the rescue mission took place.\nIt’s also interesting to note that the newer Bing imagery, which was\ntaken\nin August 2013, can be seen side-by-side with Bing’s older imagery taken in August 2011, which is still more recent than the imagery used by Google Earth for this location:\nHere’s the rendezvous location in relation to the rest of Ramadi:\nThe Iraqi MoD describes the rescue as taking place “in the district of the stadium”بمنطقة الملعب)\n). Indeed,\nthat neighborhood\nis less than a mile from the site of the rescue operation.\nThe video ends with understandably relieved Iraqis embracing one another and trotting down an airstrip. Once again, we can use the TGT coordinates (33°20’31N 43°35’58E) and Wikimapia to determine which airfield the rescued Iraqis were flown to.\nAs would be expected, the mission was conducted from\nal-Taqaddum Airbase\n, roughly 15 miles away:\nFinally, the heartfelt coda on the runway provides the best view of the helicopters flown during the operation, which appear to be\nMil Mi-17s\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/header.jpg",
+ "alt": "header",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/CENTCOM.jpg",
+ "alt": "CENTCOM",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/mod.jpg",
+ "alt": "mod",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/hud1.jpg",
+ "alt": "hud1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/date-and-time.jpg",
+ "alt": "date and time",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/signal.jpg",
+ "alt": "signal",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/zoomout.jpg",
+ "alt": "zoomout",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/GE.jpg",
+ "alt": "GE",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/Bing.jpg",
+ "alt": "Bing",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/bingdate.jpg",
+ "alt": "bingdate",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/relation-to-ramadi.jpg",
+ "alt": "relation to ramadi",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/happy.jpg",
+ "alt": "happy",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/taqaddum.jpg",
+ "alt": "taqaddum",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/05/mi-17.jpg",
+ "alt": "mi-17",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-05-26_mamai-geolocation.json b/bellingcat_geolocation_json/2015-05-26_mamai-geolocation.json
new file mode 100644
index 0000000000000000000000000000000000000000..cb5fd9cca31c6b39c6ae61d1b2f50e1ca493578c
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-05-26_mamai-geolocation.json
@@ -0,0 +1,57 @@
+{
+ "slug": "mamai-geolocation",
+ "title": "Confirmed: Active Russian Spetsnaz Serviceman Photographed in Luhansk",
+ "url": "https://www.bellingcat.com/news/europe/2015/05/26/mamai-geolocation/",
+ "date": "2015-05-26T13:12:33+00:00",
+ "modified": "2015-05-26T13:25:11+00:00",
+ "author": "Ruslan Leviev",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Ruslan Leviev and his team of investigators recently released an in-depth report on three Russian soldiers who died fighting in Ukraine. Leviev and his team members managed to track down the gravesites of each of these soldiers, all of which belonged to and were still active servicemen in the Russian 16th Separate Spetsnaz Brigade, military unit 54607. One […]",
+ "content_html": "
Ruslan Leviev and his team of investigators recently released an in-depth report on three Russian soldiers who died fighting in Ukraine. Leviev and his team members managed to track down the gravesites of each of these soldiers, all of which belonged to and were still active servicemen in the Russian 16th Separate Spetsnaz Brigade, military unit 54607. One of these soldiers was Timur “Mamai” Mamayusupov, photographed in May next to a distinct APC with shark teeth. This location was assumed to be Luhansk, but only now can this location be confirmed via geolocation. Below is a summary of the English translation of a report detailing the investigation from Leviev and his investigation team.
\n
Our investigation of three Spetsnaz servicemen killed in Ukraine featured several pieces of evidence that they really did visit Eastern Ukraine. A photo shows Ivan “Kardan” and Timur “Mamai” with weapons in a semi-ruined building, Kardan having white leg bands characteristic of rebels. “Mamai” was photographed close to two APCs, one of them bearing a distinct “shark” paint job.
However, until now we had not managed to find photographic or video evidence conclusively proving that the killed Spetsnaz servicemen were in the Donbass. So, we went back to Timur “Mamai”’s photo in front of the painted APC (we called “the shark APC”).
\n
In the comments to the translation of our investigation published on Bellingcat, a reader noted that Mamai’s photo was probably mirrored. This is visible by the APC’s paint job, as well as that in the photo Mamai apparently has his left hand on the trigger. We can also see a part of the inverted “P” (from the word “ЛНР” meaning “Luhansk People’s Republic”), confirming that the image was mirrored. Thus, the tree branch is actually to the left of Mamai, not the right. The below photograph of Mamai is properly mirrored, showing the original photograph with distinguishable landmarks:
\n
\n
After searching for photos of the Shark APC, it turned out that this vehicle had already been noticed by many Ukrainian and Western investigators. We found an array of photos that show some of the landmarks visible in the photo with Mamai, including this photograph found on VK, which was taken at the same location as the Mamai photo, with several landmarks visible in both images.
\n
\n
There are five landmarks that are visible in these two images; all of these places must be located in order to verify a geolocation:
\n
\n
A distinct light-yellow two-storey building with white shutters (in both photos, purple)
\n
A red fence (in both photos, red)
\n
A bent tree with distinct branch (in both photos, green)
\n
A tall chimney (only in Mamai photo, blue)
\n
A yellow marker on top of the red fence directly behind Mamai (only in Mamai photo, yellow)
\n
\n
All of these landmarks can be identified in photographs available in open source data.
\n
On April 8, a local tweeted a picture of a rebel base, giving the location as the backyard of Mafia restaurant on Oboronna street:
In April, a Ukrainian English-language website lugansk-news.com investigated the likely separatist base in Luhansk quite thoroughly, giving, in particular, this geolocation near Oboronna and Sasyury streets:
\n
Screenshot of satellite map geolocation from lugansk-news.com. Note that building #4 is not visible in the street-level photo, and that the building numbers are slightly off from the depicted buildings, not directly over them.
\n
The red fence, distinct tree, yellow marker, and yellow building are all visible in this photograph, showing a rebel base in Luhansk with numerous APCs.
\n
\n
But another landmark still remains–the tall chimney. A similar chimney is visible at a boiler near the location specified by lugansk-news.com, and in line with the previous geolocation:
After locating this chimney, we can now give exact coordinates for each of the landmarks visible in the photos:
\n
\n
Distinct light-yellow two-storey building: located at exactly 48.5604442, 39.3163428
\n
Red fence: located at exactly 48.560453, 39.316130
\n
Bent tree with distinct branch: shadow visible on current Yandex satellite imagery, located at exactly 48.560457, 39.316202
\n
Tall chimney: located at exactly 48.558031, 39.319053
\n
Yellow marker on top of the red fence directly behind Mamai: located on fence at exactly 48.560453, 39.316130
\n
\n
Camera perspective of Mamai photograph, with the white lines designating the line of sight, green circle showing the shadow of the bent tree, red circle the red fence, and purple the two-storey light-yellow building. The tall chimney is one block to the south of this site.
\n
This base, and the location of the Russian soldier “Mamai,” is near the “Gratsiya” factory and “Mafia” restaurant in central Luhansk at 48.5605241, 39.316147.
\n
\n
This article is a summary of a more detailed geolocation report found here in English translation. The authors and translators have included information for supporting their research through donations to their Yandex wallet at the end of the detailed article.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nRuslan Leviev and his team of investigators recently released an\nin-depth report\non three Russian soldiers who died fighting in Ukraine. Leviev and his team members managed to track down the gravesites of each of these soldiers, all of which belonged to and were still active servicemen in the Russian 16th Separate Spetsnaz Brigade, military unit 54607. One of these soldiers was Timur “Mamai” Mamayusupov, photographed in May next to a distinct APC with shark teeth. This location was assumed to be Luhansk, but only now can this location be confirmed via geolocation. Below is a summary of the English translation of\na report detailing the investigation\nfrom Leviev and his investigation team\n.\nOur\ninvestigation\nof three Spetsnaz servicemen killed in Ukraine featured several pieces of evidence that they really did visit Eastern Ukraine. A photo shows Ivan “Kardan” and Timur “Mamai” with weapons in a semi-ruined building, Kardan having white leg bands characteristic of rebels. “Mamai” was photographed close to two APCs, one of them bearing a distinct “shark” paint job.\n(original message\n|\noriginal photo\n|\nsaved copy\n).\nHowever, until now we had not managed to find photographic or video evidence conclusively proving that the killed Spetsnaz servicemen were in the Donbass. So, we went back to Timur “Mamai”’s photo in front of the painted APC (we called “the shark APC”).\nIn the comments to the translation of our investigation published on\nBellingcat\n, a reader noted that Mamai’s photo was probably mirrored. This is visible by the APC’s paint job, as well as that in the photo Mamai apparently has his left hand on the trigger. We can also see a part of the inverted “P” (from the word “ЛНР” meaning “Luhansk People’s Republic”), confirming that the image was mirrored. Thus, the tree branch is actually to the left of Mamai, not the right. The below photograph of Mamai is properly mirrored, showing the original photograph with distinguishable landmarks:\nAfter searching for photos of the Shark APC, it turned out that this vehicle had already been noticed by many Ukrainian and Western investigators. We found an array of photos that show some of the landmarks visible in the photo with Mamai, including\nthis photograph found on VK\n, which was taken at the same location as the Mamai photo, with several landmarks visible in both images.\nThere are five landmarks that are visible in these two images; all of these places must be located in order to verify a geolocation:\nA distinct light-yellow two-storey building with white shutters (in both photos,\npurple\n)\nA red fence (in both photos,\nred)\nA bent tree with distinct branch (in both photos,\ngreen\n)\nA tall chimney (only in Mamai photo,\nblue\n)\nA yellow marker on top of the red fence directly behind Mamai (only in Mamai photo,\nyellow\n)\nAll of these landmarks can be identified in photographs available in open source data.\nOn April 8, a local tweeted a picture of a rebel base, giving the location as the backyard of Mafia restaurant on Oboronna street:\nул Оборонная, во дворе \"Мафии\"\npic.twitter.com/PZKX1lNgxi\n— Voroshilovgrad (@UkroSpider)\nApril 8, 2015\nIn April, a Ukrainian English-language website lugansk-news.com\ninvestigated\nthe likely separatist base in Luhansk quite thoroughly, giving, in particular, this geolocation near Oboronna and Sasyury streets:\nScreenshot of satellite map geolocation from lugansk-news.com. Note that building #4 is not visible in the street-level photo, and that the building numbers are slightly off from the depicted buildings, not directly over them.\nThe red fence, distinct tree, yellow marker, and yellow building are all visible in this photograph, showing a rebel base in Luhansk with numerous APCs.\nBut another landmark still remains–the tall chimney. A similar chimney\nis visible\nat a boiler near the location specified by lugansk-news.com, and in line with the previous geolocation:\nScreenshot from\nYandex street-level imagery\nAfter locating this chimney, we can now give exact coordinates for each of the landmarks visible in the photos:\nDistinct light-yellow two-storey building: located at exactly 48.5604442, 39.3163428\nRed fence: located at exactly 48.560453, 39.316130\nBent tree with distinct branch: shadow visible on current Yandex satellite imagery, located at exactly 48.560457, 39.316202\nTall chimney: located at exactly 48.558031, 39.319053\nYellow marker on top of the red fence directly behind Mamai: located on fence at exactly 48.560453, 39.316130\nCamera perspective of Mamai photograph, with the white lines designating the line of sight, green circle showing the shadow of the bent tree, red circle the red fence, and purple the two-storey light-yellow building. The tall chimney is one block to the south of this site.\nThis base, and the location of the Russian soldier “Mamai,” is near the “Gratsiya” factory and “Mafia” restaurant in central Luhansk at 48.5605241, 39.316147.\nThis article is a summary of a more detailed geolocation report found\nhere\nin English translation. The authors and translators have included information for supporting their research through donations to their Yandex wallet at\nthe end of the detailed article\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/mourning.jpg",
+ "alt": "mourning",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/luh1.jpg",
+ "alt": "luh1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/GHMRwTU71H8.jpg",
+ "alt": "GHMRwTU71H8",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/base.jpg",
+ "alt": "Screenshot of satellite map geolocation from lugansk-news.com. Note that building #4 is not visible in the street-level photo, and that the building numbers are slightly off from the depicted buildings, not directly over them.",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/RussianArmy11.jpg",
+ "alt": "RussianArmy1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/chimney.jpg",
+ "alt": "chimney",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/geoloc_map.jpg",
+ "alt": "geoloc_map",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/cfsvpbuw0aaayyv__1_.jpg",
+ "alt": "cfsvpbuw0aaayyv__1_",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-05-28_geolocating-stanislav-tarasov.json b/bellingcat_geolocation_json/2015-05-28_geolocating-stanislav-tarasov.json
new file mode 100644
index 0000000000000000000000000000000000000000..c22cd3a7dde2446fccf5ca708fc6f90296da2eab
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-05-28_geolocating-stanislav-tarasov.json
@@ -0,0 +1,42 @@
+{
+ "slug": "geolocating-stanislav-tarasov",
+ "title": "Geolocating Stanislav Tarasov",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/05/28/geolocating-stanislav-tarasov/",
+ "date": "2015-05-28T19:43:52+00:00",
+ "modified": "2015-06-01T21:14:08+00:00",
+ "author": "Aric Toler",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)When searching for images near Pavlovka, a site of artillery attacks from Russia to Ukraine, we find a series of images from a soldier named Stanislav Tarasov. These photos can be found on various geomapping services for photos posted to Instagram, including Gramfeed. By looking at other images in his Instagram, one stands out, […]",
+ "content_html": "
When searching for images near Pavlovka, a site of artillery attacks from Russia to Ukraine, we find a series of images from a soldier named Stanislav Tarasov. These photos can be found on various geomapping services for photos posted to Instagram, including Gramfeed.
\n
By looking at other images in his Instagram, one stands out, showing the soldier holding his gun, facing a village in the background. This image has since been deleted from Instagram, but can still be found on Tarasov’s other social media accounts, including VK (archive) and Odnoklassniki.
\n
\n
When looking for suitable locations that fit this image around Pavlovka, numerous views facing the small village seem plausible, yet only one will fit all visible features. Using Google Earth, we can use the historical imagery that most closely matches the time of the photo (Summer 2014). Luckily, there are numerous satellite images of Pavlovka, Russia from August 2014, allowing us to closely match the satellite imagery with the ground-level view from Tarasov’s social media account.
\n
By changing the angle of the camera in Google earth, we are able to recreate the view of the image. By matching the roofs, path, and crop patterns, we can come closer to geolocating this photo. By facing east from the point at 47°56’10.33″N 39°50’2.55″E, the road, brush, and roof patterns match perfectly.
\n
\n
Only the top part of this path (marked with a blue arrow) is visible in the above satellite image, thus the road’s bend is not depicted.
\n
Many factors have changed in the satellite image since the photograph was taken—the wear on the fields with additional roads, some grass patterns, and so on. Additionally, some landmarks visible in the satellite photos, such as additional houses and what appears to be bales of hay, are not visible in the original photo due to the soldier obscuring the view. However, most importantly, the static landmarks match— an especially strong match with the angles of the roofs and the spacing between and sequencing of the houses matching. Due to this, along with the circumstantial evidence of this soldier being in Pavlovka at the time of the photograph’s upload, we can confidently say that this photograph was taken facing east towards the village of Pavlovka in the summer of 2014.
\n
Stanislav Tarasov was featured in Hiding in Plain Sight: Putin’s War in Ukraine, available here.
\n
Update:
\n
Comment “Markus W.” submitted the following images that further solidify the geolocation above, using extra landmarks visible in satellite imagery. With his permission, below are his findings:
\n
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nWhen searching for images near Pavlovka, a site of artillery attacks from Russia to Ukraine, we find a series of images from a soldier named Stanislav Tarasov. These photos can be found on various geomapping services for photos posted to Instagram, including Gramfeed.\nBy looking at other images in his Instagram, one stands out, showing the soldier holding his gun, facing a village in the background. This image has since been deleted from Instagram, but can still be found on Tarasov’s other social media accounts, including\nVK\n(\narchive\n) and\nOdnoklassniki\n.\nWhen looking for suitable locations that fit this image around Pavlovka, numerous views facing the small village seem plausible, yet only one will fit all visible features. Using Google Earth, we can use the historical imagery that most closely matches the time of the photo (Summer 2014). Luckily, there are numerous satellite images of Pavlovka, Russia from August 2014, allowing us to closely match the satellite imagery with the ground-level view from Tarasov’s social media account.\nBy changing the angle of the camera in Google earth, we are able to recreate the view of the image. By matching the roofs, path, and crop patterns, we can come closer to geolocating this photo. By facing east from the point at 47°56’10.33″N 39°50’2.55″E, the road, brush, and roof patterns match perfectly.\nOnly the top part of this path (marked with a blue arrow) is visible in the above satellite image, thus the road’s bend is not depicted.\nMany factors have changed in the satellite image since the photograph was taken—the wear on the fields with additional roads, some grass patterns, and so on. Additionally, some landmarks visible in the satellite photos, such as additional houses and what appears to be bales of hay, are not visible in the original photo due to the soldier obscuring the view. However, most importantly, the static landmarks match— an especially strong match with the angles of the roofs and the spacing between and sequencing of the houses matching. Due to this, along with the circumstantial evidence of this soldier being in Pavlovka at the time of the photograph’s upload, we can confidently say that this photograph was taken facing east towards the village of Pavlovka in the summer of 2014.\nStanislav Tarasov was featured in Hiding in Plain Sight: Putin’s War in Ukraine, available\nhere\n.\nUpdate:\nComment “Markus W.” submitted the following images that further solidify the geolocation above, using extra landmarks visible in satellite imagery.\nWith his permission\n, below are his findings:",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/pav1.jpg",
+ "alt": "pav1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/pav2.jpg",
+ "alt": "pav2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/pav31.jpg",
+ "alt": "pav3",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/uFA3scj.jpg",
+ "alt": "uFA3scj",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/3fgnLfz.jpg",
+ "alt": "3fgnLfz",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-05-28_russias-pantsir-s1s-geolocated-in-ukraine.json b/bellingcat_geolocation_json/2015-05-28_russias-pantsir-s1s-geolocated-in-ukraine.json
new file mode 100644
index 0000000000000000000000000000000000000000..f1252ddfc9df48c5ae7892f239570663864062af
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-05-28_russias-pantsir-s1s-geolocated-in-ukraine.json
@@ -0,0 +1,47 @@
+{
+ "slug": "russias-pantsir-s1s-geolocated-in-ukraine",
+ "title": "Russia's Pantsir-S1s Geolocated in Ukraine",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/05/28/russias-pantsir-s1s-geolocated-in-ukraine/",
+ "date": "2015-05-28T20:38:49+00:00",
+ "modified": "2015-05-29T07:15:25+00:00",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)During the conflict in Ukraine a number of Russian military vehicles not known to be in service in Ukraine have appeared under the control of separatists inside Ukraine. One of these systems is the Pantsir-S1 anti-aircraft system, with a number of sightings in separatist controlled Ukraine. Using geolocation it has been possible to establish […]",
+ "content_html": "
During the conflict in Ukraine a number of Russian military vehicles not known to be in service in Ukraine have appeared under the control of separatists inside Ukraine. One of these systems is the Pantsir-S1 anti-aircraft system, with a number of sightings in separatist controlled Ukraine. Using geolocation it has been possible to establish the exact location the images of the Pantsir-S1s have been captured, further confirming the appearance of these Russian system inside Ukraine.
\n
Luhansk
\n
The following video was shared online in February 2015 showing a Pantsir-S1 in separatist controlled Luhansk.
\n\n
Luhansk is well covered by Yandex Maps street view imagery, and the location the video was filmed can be found here. At the start of the video the large building on the left of the screen can also be found on the street view imagery.
\n
Shortly after that the same building can be seen in both the video and the street view imagery on the right side of the road.
\n
\n
Then on the left arched structures are also visible in both the video and street view imagery.
\n
\n
Those structures are then followed by two distinct buildings.
\n
\n
There are many other matches between the video and street view imagery, clearly establishing the exact location in separatist controlled Luhansk the Pantsir-S1 the video was filmed.
\n
Makeevka
\n
Another sighting of a Pantsir-S1 was shared on January 24th 2015, with a photograph showing it in what was claimed to be separatist controlled Makeevka.
\n
\n
The photograph was geolocated by the Conflict Report blog, with additional analysis done on Russian social media sites. This work established the location as 48.05830,38.01003, and it is possible to confirm the geolocation using the information provided by the earlier analysis.
\n
A number of structure are visible in the image that match the location, but one is of particular interest. The building, described as an unfinished shopping centre on Wikimapia, has a distinct shape and arrangement of windows, which is visible in the satellite imagery of the area.
\n
\n
Not only is the shape of the building a perfect match for what’s visible in the photograph, but the windows are also visible in the satellite imagery in the same positions as those on the building in the photograph. Along with other matching features detailed in earlier analysis of this photograph, the unusual design of the building visible in the photograph matching with the satellite imagery confirms that this is the correction location.
\n
These images were collected as part of the Bellingcat Ukraine Conflict Vehicle Tracking Project, a crowd sourced project aimed at collecting and organising images of military vehicles involved in the conflict with Ukraine.
\n
Pantsir-S1 anti-aircraft systems were featured in Hiding in Plain Sight: Putin’s War in Ukraine, available here.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nDuring the conflict in Ukraine a number of Russian military vehicles not known to be in service in Ukraine have appeared under the control of separatists inside Ukraine. One of these systems is the\nPantsir-S1 anti-aircraft system\n, with a number of sightings in separatist controlled Ukraine. Using geolocation it has been possible to establish the exact location the images of the Pantsir-S1s have been captured, further confirming the appearance of these Russian system inside Ukraine.\nLuhansk\nThe following video was shared online in February 2015 showing a Pantsir-S1 in separatist controlled Luhansk.\nLuhansk is well covered by Yandex Maps street view imagery, and the location the video was filmed can be found\nhere\n. At the start of the video the large building on the left of the screen can also be found on the street view imagery.\nShortly after that the same building can be seen in both the video and the street view imagery on the right side of the road.\nThen on the left arched structures are also visible in both the video and street view imagery.\nThose structures are then followed by two distinct buildings.\nThere are many other matches between the video and street view imagery, clearly establishing the exact location in separatist controlled Luhansk the Pantsir-S1 the video was filmed.\nMakeevka\nAnother sighting of a Pantsir-S1 was shared on January 24\nth\n2015, with a photograph showing it in what was claimed to be separatist controlled Makeevka.\nThe photograph was geolocated by the\nConflict Report blog\n, with additional analysis done on\nRussian social media sites\n. This work established the location as\n48.05830,38.01003\n, and it is possible to confirm the geolocation using the information provided by the earlier analysis.\nA number of structure are visible in the image that match the location, but one is of particular interest. The building, described as an unfinished shopping centre\non Wikimapia\n, has a distinct shape and arrangement of windows, which is visible in the satellite imagery of the area.\nNot only is the shape of the building a perfect match for what’s visible in the photograph, but the windows are also visible in the satellite imagery in the same positions as those on the building in the photograph. Along with other matching features detailed in earlier analysis of this photograph, the unusual design of the building visible in the photograph matching with the satellite imagery confirms that this is the correction location.\nThese images were collected as part of the\nBellingcat Ukraine Conflict Vehicle Tracking Project\n, a crowd sourced project aimed at collecting and organising images of military vehicles involved in the conflict with Ukraine.\nPantsir-S1 anti-aircraft systems were featured in Hiding in Plain Sight: Putin’s War in Ukraine, available\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/11.jpg",
+ "alt": "1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/2.jpg",
+ "alt": "2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/3.jpg",
+ "alt": "3",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/4.jpg",
+ "alt": "4",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/5.png",
+ "alt": "5",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/6.jpg",
+ "alt": "6",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-05-29_confirming-the-location-of-the-same-msta-s-in-russia-and-ukraine.json b/bellingcat_geolocation_json/2015-05-29_confirming-the-location-of-the-same-msta-s-in-russia-and-ukraine.json
new file mode 100644
index 0000000000000000000000000000000000000000..f522ffbd5f584e38b52d21e14f78713895537ad4
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-05-29_confirming-the-location-of-the-same-msta-s-in-russia-and-ukraine.json
@@ -0,0 +1,52 @@
+{
+ "slug": "confirming-the-location-of-the-same-msta-s-in-russia-and-ukraine",
+ "title": "Confirming the Location of the Same Msta-S in Russia and Ukraine",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/05/29/confirming-the-location-of-the-same-msta-s-in-russia-and-ukraine/",
+ "date": "2015-05-29T13:56:28+00:00",
+ "modified": "",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)During the conflict in Ukraine it has been possible to identify units sighted in Russia that have also been filmed and photographed in Ukraine. The following is an example of the same Msta-S self-propelled 152 mm howitzer filmed in both Ukraine and Russia. Rostov-na-Donu, Russia A video uploaded to YouTube on July 30 shows various pieces […]",
+ "content_html": "
During the conflict in Ukraine it has been possible to identify units sighted in Russia that have also been filmed and photographed in Ukraine. The following is an example of the same Msta-S self-propelled 152 mm howitzer filmed in both Ukraine and Russia.
\n
Rostov-na-Donu, Russia
\n
A video uploaded to YouTube on July 30 shows various pieces of Russian military equipment, including two Msta-S self-propelled howitzers.
\n
\n
The self-reported location of the video was in Rostov on the Taganrog highway, making that the first location we should check to either verify or disprove. The general location of the video can be established by the license plates visible in the photo—all of which end in either “61” or “161,” the regional plate numbers for the Rostov oblast.
\n
\n
The road has four lanes and has a distinct series of shops on one side of the road—a flower shop (ЦВЕТЫ), construction material shop (СТРОЙМАТЕРИАЛЫ), repair shop (РЕМОНТ), and hardware shop (КРЕПЕЖ), along with a lemonade / kvas kiosk kiosk on the corner.
\n
\n
By following the self-reported location (the highway that goes from Rostov to Taganrog) on Google street view, eventually we come to a location that shows the exact same series of shops, located at the intersection of Taganroskaya ulitsa (Taganrog Street) and Solovyinaya ulitsa. Even though some of the image is distorted or blurred out, this is clearly the same location:
\n
\n
Therefore, we can confidently say that this video was shot near the coordinates of 47.262757, 39.660493 in Rostov-na-Donu, Russia.
\n
Novoazovsk, Ukraine
\n
The same Msta-S appeared in a September video, filmed in Novoazovsk, Ukraine, driving along a road at 47.1275441, 38.0892229. While the video contains multiple shots filmed at different positions and angles, it’s possible to confirm the location of the road the Msta-S was driving along.
\n
\n
Signs visible in the September video are also visible in an earlier video filmed in 2012 showing an extremely lengthy section of the road. Due to the length of the video it is possible to match multiple features along the road, showing the camera is facing to the west. It’s also possible to establish the shot that shows the Msta-S also sighted in Russia was filmed from approximately the same location.
\n
In the 2012 video, just before the car reaches the signs filmed in the September 2014 video, a single story blue building is visible, opposite a junction, and with a pole in front of it. This matches with the shot in the September 2014 video showing the Msta-S previously sighted in Russia.
\n
\n
It’s possible to establish the location of the camera in each of those two shots as follows, therefore confirming the same Msta-S was seen in both Russia in July 2014 and later Ukraine in September 2014.
\n
\n
These images were collected as part of the Bellingcat Ukraine Conflict Vehicle Tracking Project, a crowd sourced project aimed at collecting and organising images of military vehicles involved in the conflict with Ukraine.
\n
This example is featured in Hiding in Plain Sight: Putin’s War in Ukraine, available here.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nDuring the conflict in Ukraine it has been possible to identify units sighted in Russia that have also been filmed and photographed in Ukraine. The following is an example of the same Msta-S self-propelled 152 mm howitzer filmed in both Ukraine and Russia.\nRostov-na-Donu, Russia\nA\nvideo\nuploaded to YouTube on July 30 shows various pieces of Russian military equipment, including two Msta-S self-propelled howitzers.\nThe self-reported location of the video was in Rostov on the Taganrog highway, making that the first location we should check to either verify or disprove. The general location of the video can be established by the license plates visible in the photo—all of which end in either “61” or “161,” the regional plate numbers for the Rostov oblast.\nThe road has four lanes and has a distinct series of shops on one side of the road—a flower shop (ЦВЕТЫ), construction material shop (СТРОЙМАТЕРИАЛЫ), repair shop (РЕМОНТ), and hardware shop (КРЕПЕЖ), along with a lemonade / kvas kiosk kiosk on the corner.\nBy following the self-reported location (the highway that goes from Rostov to Taganrog) on Google street view, eventually we come to a location that shows the exact same series of shops, located at the intersection of Taganroskaya ulitsa (Taganrog Street) and Solovyinaya ulitsa. Even though some of the image is distorted or blurred out, this is clearly the same location:\nTherefore, we can confidently say that this video was shot near the coordinates of 47.262757, 39.660493 in Rostov-na-Donu, Russia.\nNovoazovsk, Ukraine\nThe same Msta-S appeared in\na September video\n, filmed in Novoazovsk, Ukraine, driving along a road at\n47.1275441, 38.0892229\n. While the video contains multiple shots filmed at different positions and angles, it’s possible to confirm the location of the road the Msta-S was driving along.\nSigns visible in the September video are also visible in\nan earlier video filmed in 2012\nshowing an extremely lengthy section of the road. Due to the length of the video it is possible to match multiple features along the road, showing the camera is facing to the west. It’s also possible to establish the shot that shows the Msta-S also sighted in Russia was filmed from approximately the same location.\nIn the 2012 video, just before the car reaches the signs filmed in the September 2014 video, a single story blue building is visible, opposite a junction, and with a pole in front of it. This matches with the shot in the September 2014 video showing the Msta-S previously sighted in Russia.\nIt’s possible to establish the location of the camera in each of those two shots as follows, therefore confirming the same Msta-S was seen in both Russia in July 2014 and later Ukraine in September 2014.\nThese images were collected as part of the\nBellingcat Ukraine Conflict Vehicle Tracking Project\n, a crowd sourced project aimed at collecting and organising images of military vehicles involved in the conflict with Ukraine.\nThis example is featured in Hiding in Plain Sight: Putin’s War in Ukraine, available\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/m1.jpg",
+ "alt": "m1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/Plates1.jpg",
+ "alt": "Plates1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/m2.jpg",
+ "alt": "m2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/m3.jpg",
+ "alt": "m3",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/m4.jpg",
+ "alt": "m4",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/m5.jpg",
+ "alt": "m5",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/m6.jpg",
+ "alt": "m6",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-06-05_google-earth-image-verification.json b/bellingcat_geolocation_json/2015-06-05_google-earth-image-verification.json
new file mode 100644
index 0000000000000000000000000000000000000000..a999825e3e60eae8a0393caf04d87a335fd12e67
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-06-05_google-earth-image-verification.json
@@ -0,0 +1,97 @@
+{
+ "slug": "google-earth-image-verification",
+ "title": "Who to Trust, Google or the Russian MoD? A Guide to Verifying Google Earth Satellite Image Dates",
+ "url": "https://www.bellingcat.com/resources/how-tos/2015/06/05/google-earth-image-verification/",
+ "date": "2015-06-05T06:54:53+00:00",
+ "modified": "2022-01-18T15:49:19+00:00",
+ "author": "Aric Toler",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Этот пост также доступен на русском языке в переводе @Mortis_Banned Bellingcat recently published an in-depth investigation finding that the Russian Ministry of Defense (MoD) fabricated and misrepresented their satellite photographs in a July 21, 2014 press conference, immediately following the July 17 downing of MH17. The most obvious and compelling evidence of this investigation […]",
+ "content_html": "
Bellingcat recently published an in-depth investigation finding that the Russian Ministry of Defense (MoD) fabricated and misrepresented their satellite photographs in a July 21, 2014 press conference, immediately following the July 17 downing of MH17. The most obvious and compelling evidence of this investigation was in the clear differences in the landscape between Russian MoD satellite images–supposedly taken in mid-July–and the publicly available images taken in the week before and after the downing of MH17. While Bellingcat provided a guide for accessing this imagery in Google Earth, a question that has arisen is in the accuracy of these dates on Google Earth. As described in this guide, the findings of Bellingcat regarding the July 21 Russian MoD satellite images will be reaffirmed, along with a walk-through for anyone to verify Google Earth imagery via free and precisely dated image previews on Digital Globe.
\n
The Russian MoD satellite images under consideration show a Ukrainian base with Buk anti-aircraft systems at the coordinates of 48.098403, 37.754424, just north of Donetsk. While the Russian MoD claims that these photos are from mid-July, a comparison with Google Earth satellite photos dated in May, early July, and late July show that this cannot be true.
\n
These comparisons of the field and vegetation patterns to the immediate west and east of the Ukrainian base are damning, in that they clearly show that the Russian MoD satellite photos were taken in June, and not July. However, a question that has arisen is how to verify the exact dates of the Google Earth satellite photos. This verification can be done through checking the previews of imagery from Digital Globe, which supplies Google Earth imagery at a higher resolution for Google Satellite. This simple guide demonstrates how easily this imagery can be checked by anyone using Google Earth.
\n
Much of the confusion over the correct date of satellite photos is from how Google Earth handles the dating of their satellite imagery, where “strips” of images are overlaid over different captured days. The patchwork nature of the variant dated imagery is not visible unless you specifically turn it on with Google Earth, where it appears as such:
\n
\n
Each strip represents the most recent satellite image up to the date selected on Google Earth (in this case, it was April 26, 2015). We can determine the specific date of each strip by zooming in, which is the date determined in each Google Earth satellite snapshot in the Bellingcat report in the Russian MoD satellite photos. Instead of just relying on the Google Earth date, we can also check the publicly available image previews on Digital Earth to verify–without any doubt–the correct date of each satellite image.
\n
From the Digital Globe’s ImageFinder page, we can input coordinates to find the historical satellite imagery for a site.
\n
\n
After selecting “Enter Coordinates,” you can enter the coordinates of the base just north of Donetsk: 48.098403, 37.754424. I selected a 2 square-kilometer radius, but you can do any parameters you would like.
\n
\n
\n
After clicking “Continue,” something like this will show up, displaying the area that is selected by the selected parameters. Click “Search” to display all available imagery. Be prepared to wait a while–it can sometimes take a bit for a window to pop up after clicking “Search.”
\n
\n
A window will now pop up (check your ad blocking plug-ins and pop-up window settings!) showing all available imagery for the selected coordinates with your parameters. Click “Acquisition Date” (red box) to sort the available images by the date the images were taken.
\n
\n
From that, we get the following list of images taken in the past year with varying angles, cloud cover, image band type, and so on. Select “View” for your desired satellite image to find the free preview available for the image, so that we can verify that the Google Satellite image is the same as the Digital Globe image, thus verifying the capture date.
\n
\n
After clicking “View,” a window like this will come up, showing the satellite image along with its metadata. You want to re-size the image to “max available resolution,” and then save the image so that it is easier to examine the finer details. This particular image is for July 24, exactly a week after the shoot-down of MH17.
\n
\n
Be sure to save the image to examine it more closely.
\n
\n
Now, we can finally conduct a close comparison of the Digital Globe image and Google Earth satellite image to make sure they are the same, thus verifying the date of the Google Satellite image. The previews of the Digital Globe imagery will be a much lower resolution and quality than the Google Satellite version, but the key features that have changed since May and June will be visible.
\n
Google Earth Satellite, July 24, 2014 (UTC)
\n
Digital Globe, July 24, 2014, ID: 10300100333BC100
\n
We can do the same thing for July 2, 2014:
\n
\n
\n
And May 30, 2014:
\n
\n
\n
Though you may need to squint your eyes a bit, it is undeniable that the Google Satellite images are these same as the Digital Globe images taken on the same day. Thus, we can put to rest any notions that the dates in the Bellingcat report in the Russian Ministry of Defense were incorrect or estimated.
\n
Just to recap, take another look at red and yellow-boxed features of the vegetation and field in the the Russian Ministry of Defense’s satellite photo that was supposedly taken in mid-July, as compared with the (now double verified) dates of the Google Satellite images:
\n
\n
An example of incorrectly dated imagery used by the Russian Ministry of Defence
\n
Unless a patch of vegetation disappeared, reappeared, then disappeared again, the Russian Ministry of Defense falsely presented their satellite images. Their satellite images were taken in June, not on July, as they told the world while accusing Ukraine of shooting down MH17 just four days after the tragedy took place.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nЭтот пост также доступен на русском языке в переводе\n@Mortis_Banned\nBellingcat recently published an in-depth investigation finding that the Russian Ministry of Defense (MoD) fabricated and misrepresented their satellite photographs in a July 21, 2014 press conference, immediately following the July 17 downing of MH17. The most obvious and compelling evidence of this investigation was in the clear differences in the landscape between Russian MoD satellite images–supposedly taken in mid-July–and the publicly available images taken in the week before and after the downing of MH17. While Bellingcat provided a guide for accessing this imagery in Google Earth, a question that has arisen is in the accuracy of these dates on Google Earth. As described in this guide, the findings of Bellingcat regarding the July 21 Russian MoD satellite images will be reaffirmed, along with a walk-through for anyone to verify Google Earth imagery via free and precisely dated image previews on Digital Globe.\nThe Russian MoD satellite images under consideration show a Ukrainian base with Buk anti-aircraft systems at the coordinates of 48.098403, 37.754424, just north of Donetsk. While the Russian MoD claims that these photos are from mid-July, a comparison with Google Earth satellite photos dated in May, early July, and late July show that this cannot be true.\nThese comparisons of the field and vegetation patterns to the immediate west and east of the Ukrainian base are damning, in that they clearly show that the Russian MoD satellite photos were taken in June, and not July. However, a question that has arisen is how to verify the exact dates of the Google Earth satellite photos. This verification can be done through checking the previews of imagery from Digital Globe, which supplies Google Earth imagery at a higher resolution for Google Satellite.\nThis simple guide\ndemonstrates how easily this imagery can be checked by anyone using Google Earth.\nMuch of the confusion over the correct date of satellite photos is from how Google Earth handles the dating of their satellite imagery, where “strips” of images are overlaid over different captured days. The patchwork nature of the variant dated imagery is not visible unless you specifically turn it on with Google Earth, where it appears as such:\nEach strip represents the most recent satellite image up to the date selected on Google Earth (in this case, it was April 26, 2015). We can determine the specific date of each strip by zooming in, which is the date determined in each Google Earth satellite snapshot in the Bellingcat report in the Russian MoD satellite photos. Instead of just relying on the Google Earth date, we can also check the publicly available image previews on Digital Earth to verify–without any doubt–the correct date of each satellite image.\nFrom the Digital Globe’s ImageFinder page, we can input coordinates to find the historical satellite imagery for a site.\nAfter selecting “Enter Coordinates,” you can enter the coordinates of the base just north of Donetsk: 48.098403, 37.754424. I selected a 2 square-kilometer radius, but you can do any parameters you would like.\nAfter clicking “Continue,” something like this will show up, displaying the area that is selected by the selected parameters. Click “Search” to display all available imagery. Be prepared to wait a while–it can sometimes take a bit for a window to pop up after clicking “Search.”\nA window will now pop up (check your ad blocking plug-ins and pop-up window settings!) showing all available imagery for the selected coordinates with your parameters. Click “Acquisition Date” (red box) to sort the available images by the date the images were taken.\nFrom that, we get the following list of images taken in the past year with varying angles, cloud cover, image band type, and so on. Select “View” for your desired satellite image to find the free preview available for the image, so that we can verify that the Google Satellite image is the same as the Digital Globe image, thus verifying the capture date.\nAfter clicking “View,” a window like this will come up, showing the satellite image along with its metadata. You want to re-size the image to “max available resolution,” and then save the image so that it is easier to examine the finer details. This particular image is for July 24, exactly a week after the shoot-down of MH17.\nBe sure to save the image to examine it more closely.\nNow, we can finally conduct a close comparison of the Digital Globe image and Google Earth satellite image to make sure they are the same, thus verifying the date of the Google Satellite image. The previews of the Digital Globe imagery will be a much lower resolution and quality than the Google Satellite version, but the key features that have changed since May and June will be visible.\nGoogle Earth Satellite, July 24, 2014 (UTC)\nDigital Globe, July 24, 2014, ID: 10300100333BC100\nWe can do the same thing for July 2, 2014:\nAnd May 30, 2014:\nThough you may need to squint your eyes a bit, it is undeniable that the Google Satellite images are these same as the Digital Globe images taken on the same day. Thus, we can put to rest any notions that the dates in the Bellingcat report in the Russian Ministry of Defense were incorrect or estimated.\nJust to recap, take another look at red and yellow-boxed features of the vegetation and field in the the Russian Ministry of Defense’s satellite photo that was supposedly taken in mid-July, as compared with the (now double verified) dates of the Google Satellite images:\nAn example of incorrectly dated imagery used by the Russian Ministry of Defence\nUnless a patch of vegetation disappeared, reappeared, then disappeared again, the Russian Ministry of Defense falsely presented their satellite images. Their satellite images were taken in June, not on July, as they told the world while accusing Ukraine of shooting down MH17 just four days after the tragedy took place.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/strips.jpg",
+ "alt": "strips",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/globe.jpg",
+ "alt": "globe",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/coords.jpg",
+ "alt": "coords",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/globe_search.jpg",
+ "alt": "globe_search",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/globe_acquisition.jpg",
+ "alt": "globe_acquisition",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/globe_results.jpg",
+ "alt": "globe_results",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/globe_max.jpg",
+ "alt": "globe_max",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/globe_save.jpg",
+ "alt": "globe_save",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/ge_7.24.14.jpg",
+ "alt": "ge_7.24.14",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/globe_7242014.jpg",
+ "alt": "globe_7242014",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/ge_july2.jpg",
+ "alt": "ge_july2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/globe_july2.jpg",
+ "alt": "globe_july2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/ge_may30.jpg",
+ "alt": "ge_may30",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/globe_may30.jpg",
+ "alt": "globe_may30",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/05/Russia-MoD-2.jpg",
+ "alt": "Russia MoD 2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/06/MoD-1.jpg",
+ "alt": "An example of incorrectly dated imagery used by the Russian Ministry of Defence",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-07-25_searching-the-earth-essential-geolocation-tools-for-verification.json b/bellingcat_geolocation_json/2015-07-25_searching-the-earth-essential-geolocation-tools-for-verification.json
new file mode 100644
index 0000000000000000000000000000000000000000..4a2cccdc90fd7a0248c82f96aa0e055ea6e7bb5c
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-07-25_searching-the-earth-essential-geolocation-tools-for-verification.json
@@ -0,0 +1,42 @@
+{
+ "slug": "searching-the-earth-essential-geolocation-tools-for-verification",
+ "title": "Searching the Earth: Essential Geolocation Tools for Verification",
+ "url": "https://www.bellingcat.com/resources/how-tos/2015/07/25/searching-the-earth-essential-geolocation-tools-for-verification/",
+ "date": "2015-07-25T07:30:15+00:00",
+ "modified": "",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)As verification and open source investigation techniques and methodologies have been developed over the last few years various tools and platforms have been used to access the information that makes this kind of work possible. This includes pre-existing tools such as Google Earth and new tools developed over the last few years. In this […]",
+ "content_html": "
As verification and open source investigation techniques and methodologies have been developed over the last few years various tools and platforms have been used to access the information that makes this kind of work possible. This includes pre-existing tools such as Google Earth and new tools developed over the last few years. In this article we’ll examine some of these tools, and the different uses they have in the verification and investigation process.
\n
Googling Earth
\n
For many investigations Google Earth has become a key tool for verifying images and videos, playing a key role in the process of geolocation. Geolocation is a verification method where landmarks and features visible in photographs or videos can be compared to other images, frequently satellite imagery, to confirm the location an image was recorded. At its most basic, Google Earth allows access to satellite imagery from across the globe, a key resource in geolocation. It also has a number of other useful features that can also play a useful role in verification.
\n
One feature that is often overlooked is historical imagery, which can be easily found under the “view” menu. By turning this on the user is presented with a slider that allows them to change the displayed image to any historical image available on Google Earth. This is useful for a number of reasons; for example, the position of the satellite when it records an image might give you one view of a side of a building, and other images take at different times might give you a different view of the building, which is useful when you’re trying to match the features of a building in a photograph or video to a satellite image. The below example shows the same location on two different dates, providing a view of two different sides of the buildings.
\n
\n
\n
Another useful feature is the “Photos” layer, which displays photos from a number of sources, but most importantly Panoramio. Panoramio is a website where users can upload geotagged photographs, and these photographs are visible on Google Earth, making a useful source of ground level images of locations you might be interested in examining in more detail.
\n
The View from the Street
\n
Another source of ground level images are street view images available on Google Maps and Google Earth, and also on the Russian website Yandex Maps. While many people are familiar with Google Street View, Yandex Maps is less well known, but has its own collection of street view imagery, much of which is in locations not yet available on Google Street View.
\n
These images can be hugely useful for geolocation, but it’s also important to consider the date street view images were recorded. For example, in this recent piece, verifying the location where a MH17 related video was recorded, Yandex Maps imagery showed the location in question absent of the billboards that were visible in the video being geolocated. However, Google Earth satellite imagery showed the billboard were a recent addition to the area, and the Yandex Maps imagery was recorded before the billboards were constructed.
\n
\n
By comparing other images from Yandex Maps with other information about the area (for example, a recently constructed church), it was possible to verify the date of the Yandex Maps imagery. This demonstrates another overlooked facet of geolocation and verification, it’s not only the case that the image you’re investigating needs to be verified, but it’s also valuable to verify the reference image as well. Something like a geotag can be inaccurate or changed, so this is one of the things that needs to be verified as part of a robust verification process.
\n
A World of Geotags
\n
Recently new platforms have been developed that makes geotags even more useful as part of the verification process. Yomapic is a free website that allows for geographical searches for geotagged images of the popular Russian social media site VKontakte and Instagram. In the below example Yomapic displays images posted from the Islamic State controlled town of Raqqa in the last week.
\n
\n
An interesting feature of Yomapic is it’s possible to select an individual account by clicking on the username, displaying all images on the account, geotagged or not. The below image shows on Instagram user as they travel from the Seychelles to Kenya, then onto Turkey and finally Raqqa, Syria.
\n
\n
An alternative to Yomapic is EchoSec, which includes many more social media sites in its searches, including Twitter, Facebook, YouTube, Panoramio, and Flickr. It also has a date range search, which allows for searches around specific date ranges, which is very useful when investigating past events.
\n
While the obvious use for Yomapic and EchoSec is looking for posts related to specific events in an area, it’s also useful in another way. When geolocating images it might not be possible to find street view imagery or Panoramio photographs, but thanks to Yomapic and EchoSec it’s possible to see if anyone in the local area may have posted an Instagram photo or YouTube video showing the location you’re interested in. This adds to the range of option you have when searching for reference images as part of any geolocation, and by combining these tools it makes efforts to verify content far more effective.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nAs verification and open source investigation techniques and methodologies have been developed over the last few years various tools and platforms have been used to access the information that makes this kind of work possible. This includes pre-existing tools such as Google Earth and new tools developed over the last few years. In this article we’ll examine some of these tools, and the different uses they have in the verification and investigation process.\nGoogling Earth\nFor many investigations Google Earth has become a key tool for verifying images and videos, playing a key role in the process of geolocation. Geolocation is a verification method where landmarks and features visible in photographs or videos can be compared to other images, frequently satellite imagery, to confirm the location an image was recorded. At its most basic, Google Earth allows access to satellite imagery from across the globe, a key resource in geolocation. It also has a number of other useful features that can also play a useful role in verification.\nOne feature that is often overlooked is historical imagery, which can be easily found under the “view” menu. By turning this on the user is presented with a slider that allows them to change the displayed image to any historical image available on Google Earth. This is useful for a number of reasons; for example, the position of the satellite when it records an image might give you one view of a side of a building, and other images take at different times might give you a different view of the building, which is useful when you’re trying to match the features of a building in a photograph or video to a satellite image. The below example shows the same location on two different dates, providing a view of two different sides of the buildings.\nAnother useful feature is the “Photos” layer, which displays photos from a number of sources, but most importantly\nPanoramio\n. Panoramio is a website where users can upload geotagged photographs, and these photographs are visible on Google Earth, making a useful source of ground level images of locations you might be interested in examining in more detail.\nThe View from the Street\nAnother source of ground level images are street view images available on Google Maps and Google Earth, and also on the Russian website\nYandex Maps\n. While many people are familiar with Google Street View, Yandex Maps is less well known, but has its own collection of street view imagery, much of which is in locations not yet available on Google Street View.\nThese images can be hugely useful for geolocation, but it’s also important to consider the date street view images were recorded. For example, in\nthis recent piece\n, verifying the location where a MH17 related video was recorded, Yandex Maps imagery showed the location in question absent of the billboards that were visible in the video being geolocated. However, Google Earth satellite imagery showed the billboard were a recent addition to the area, and the Yandex Maps imagery was recorded before the billboards were constructed.\nBy comparing other images from Yandex Maps with other information about the area (for example, a recently constructed church), it was possible to verify the date of the Yandex Maps imagery. This demonstrates another overlooked facet of geolocation and verification, it’s not only the case that the image you’re investigating needs to be verified, but it’s also valuable to verify the reference image as well. Something like a geotag can be inaccurate or changed, so this is one of the things that needs to be verified as part of a robust verification process.\nA World of Geotags\nRecently new platforms have been developed that makes geotags even more useful as part of the verification process.\nYomapic\nis a free website that allows for geographical searches for geotagged images of the popular Russian social media site\nVKontakte\nand Instagram. In the below example Yomapic displays images posted from the Islamic State controlled town of Raqqa in the last week.\nAn interesting feature of Yomapic is it’s possible to select an individual account by clicking on the username, displaying all images on the account, geotagged or not. The below image shows on Instagram user as they travel from the Seychelles to Kenya, then onto Turkey and finally Raqqa, Syria.\nAn alternative to Yomapic is\nEchoSec\n, which includes many more social media sites in its searches, including Twitter, Facebook, YouTube, Panoramio, and Flickr. It also has a date range search, which allows for searches around specific date ranges, which is very useful when investigating past events.\nWhile the obvious use for Yomapic and EchoSec is looking for posts related to specific events in an area, it’s also useful in another way. When geolocating images it might not be possible to find street view imagery or Panoramio photographs, but thanks to Yomapic and EchoSec it’s possible to see if anyone in the local area may have posted an Instagram photo or YouTube video showing the location you’re interested in. This adds to the range of option you have when searching for reference images as part of any geolocation, and by combining these tools it makes efforts to verify content far more effective.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/07/11.jpg",
+ "alt": "1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/07/21.jpg",
+ "alt": "2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/07/31.jpg",
+ "alt": "3",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/07/4.jpg",
+ "alt": "4",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/07/51.jpg",
+ "alt": "5",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-08-07_shadow-of-a-doubt.json b/bellingcat_geolocation_json/2015-08-07_shadow-of-a-doubt.json
new file mode 100644
index 0000000000000000000000000000000000000000..ccd4def1d8e925df9e50e89387a8294010ba132c
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-08-07_shadow-of-a-doubt.json
@@ -0,0 +1,47 @@
+{
+ "slug": "shadow-of-a-doubt",
+ "title": "Shadow of a Doubt: Crowdsourcing Time Verification of the MH17 Missile Launch Photo",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/08/07/shadow-of-a-doubt/",
+ "date": "2015-08-07T12:53:32+00:00",
+ "modified": "2015-08-08T13:29:00+00:00",
+ "author": "Aric Toler",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)While much open source investigation is done through crowd sourcing, the administrator of the MH17 webtalk.ru forum has taken a unique approach: a single-blind poll to collaboratively determine the timing of a photograph. While this experiment does not have the scientific rigor of a pharmaceutical trial, the methodology of administrator “bootblack” is worthwhile of some examination. […]",
+ "content_html": "
While much open source investigation is done through crowd sourcing, the administrator of the MH17 webtalk.ru forum has taken a unique approach: a single-blind poll to collaboratively determine the timing of a photograph. While this experiment does not have the scientific rigor of a pharmaceutical trial, the methodology of administrator “bootblack” is worthwhile of some examination.
\n
The photograph in question is the infamous MH17 launch image, which was posted by @WowihaY after he received it from a friend shortly after the downing of MH17 (4:20pm local time). The photograph shows a missile contrail south of Snizhne and southeast of the photographer’s location. Significantly, the EXIF data on this photograph shows that it was taken–according to the camera’s internal clock–at 4:25pm local time, July 17, 2014. This time places the photograph at just five minutes after the downing of MH17; however, the internal clock’s time was likely not perfectly synced and accurate the same way a cell phone would be. The circumstances of this photograph have been hotly debated online, even though the photograph has been verified as true by multiple international experts. The Dutch criminal investigation into the downing of MH17 have interviewed the photographer, and are taking the photograph into account in its investigation. Needless to say, this image is a very important piece of evidence into the tragedy’s investigation, and a more accurate time-frame for the photograph (other than the camera’s internal clock) would further confirm its value.
\n
The brightened version of the smoke trail photograph. The original photograph was darker, while this one was retouched by the photographer to better depict the smoke trail in question.
\n
Bootblack, the administrator of the popular MH17 forum of webtalk.ru, acquired numerous photographs from a local of the shadow on a particular rooftop in Torez. The administrator also received the exact time each photograph was taken. This roof in question is the same one that appears in the photograph above (bottom-left/center, to the left of a tall pole). The purpose of Bootblack’s experiment is to crowdsource the opinions of the forum to select which of the photographs most accurately depicts the angle of the rooftop shadow when compared to the July 17 launch photograph. As anyone who has ever seen a sundial will know, the time can be determined by a shadow once various factors (angle of object casting shadow, azimuth, etc) are determined.
\n
The administrator received the time of each photograph, but chose not to share them with the forum, hoping for unbiased assessments of which photograph was the most accurate. Theoretically, with some adjustments with the azimuth and the exact conditions recorded for the day of the MH17 shootdown, we would be able to know–down to a very narrow time-frame–when the launch photo was taken without relying on the reading of a camera’s internal clock in the EXIF data. We probably cannot get the exact second of the photograph with this shadow angle method, but we would be able to refute an argument that the photograph was actually taken in the morning or early afternoon.
\n
The house and shadow in question from the launch photo
\n
Bootblack set out the following procedures for the thread:
\n
1) In a joint effort (poll and comments) we will select a photo with the shadow that matches the shadow in the [photographer’s] photo. If someone’s eye fixes upon an intermediate position, just tell me.
\n
2) Knowing the recorded time of the selected photo, we will calculate the azimuth of the shadow. The EXIF has already been deleted to eliminate subjectivity.
\n
3) We will recalculate the obtained azimuth to July 17, 2014 and from that we will calculate the exact time the photographer shot the shadow.
\n
One of the photographs provided by the forum administrator
\n
For three pages, the forum users debated which of the twelve provided photographs best matched the shadow on the launch photo. Different users, including the administrator, threw their lots in to different ones, with various creative uses of overlaying lines over the given photographs to judge the correct angles:
\n
\n
\n
\n
\n
Eventually, the administrator revealed the time of the photograph numbered 4328, which was in between the two leading vote-getters (4325 and 4330). This photograph was taken at August 1, 2015 at 4:35:46pm with the azimuth of 260.52, which corresponds to a July 17, 2014 time of 4:21:29pm. The administrator later calculated what he saw as a more accurate time based on the line alignment in the two photographs: 4:22:40pm on July 17, 2014.
\n
If the results of the “shadow poll” experiment are trusted, the camera’s internal clock was about 3 minutes fast. It is difficult to conclusively say the exact minute–let alone the second–that the launch photograph was taken due to the margin of error in the shadow exercise and the impreciseness of the camera’s internal clock. However, this exercise in tandem with the self-reported EXIF data demonstrates that the launch photograph was indeed taken soon after 4:20pm, the time of the MH17 downing.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nWhile much open source investigation is done through crowd sourcing, the administrator of the MH17 webtalk.ru forum has taken a unique approach:\na single-blind poll to collaboratively determine the\ntiming of a photograph\n. While this experiment does not have the scientific rigor of a pharmaceutical trial, the methodology of administrator “bootblack” is worthwhile of some examination.\nThe photograph in question is the infamous MH17 launch image, which was posted by\n@Wowiha\nY\nafter he received it from a friend shortly after the downing of MH17 (4:20pm local time). The photograph shows a missile contrail south of Snizhne and southeast of the photographer’s location. Significantly, the EXIF data on this photograph shows that it was taken–according to the camera’s internal clock–at 4:25pm local time, July 17, 2014. This time places the photograph at just five minutes after the downing of MH17; however, the internal clock’s time was likely not perfectly synced and accurate the same way a cell phone would be. The circumstances of this photograph have been hotly debated online, even though the photograph has been verified as true\nby multiple international experts\n. The Dutch criminal investigation into the downing of MH17 have interviewed the photographer, and are taking the photograph into account in its investigation. Needless to say, this image is a very important piece of evidence into the tragedy’s investigation, and a more accurate time-frame for the photograph (other than the camera’s internal clock) would further confirm its value.\nThe brightened version of the smoke trail photograph. The original photograph was darker, while this one was retouched by the photographer to better depict the smoke trail in question.\nBootblack, the administrator of the popular MH17 forum of webtalk.ru, acquired numerous photographs from a local of the shadow on a particular rooftop in Torez. The administrator also received the exact time each photograph was taken. This roof in question is the same one that appears in the photograph above (bottom-left/center, to the left of a tall pole). The purpose of Bootblack’s experiment is to crowdsource the opinions of the forum to select which of the photographs most accurately depicts the angle of the rooftop shadow when compared to the July 17 launch photograph. As anyone who has ever seen a sundial will know, the time can be determined by a shadow once various factors (angle of object casting shadow, azimuth, etc) are determined.\nThe administrator received the time of each photograph, but chose not to share them with the forum, hoping for unbiased assessments of which photograph was the most accurate. Theoretically, with some adjustments with the azimuth and the exact conditions recorded for the day of the MH17 shootdown, we would be able to know–down to a very narrow time-frame–when the launch photo was taken without relying on the reading of a camera’s internal clock in the EXIF data. We probably cannot get the exact second of the photograph with this shadow angle method, but we would be able to refute an argument that the photograph was actually taken in the morning or early afternoon.\nThe house and shadow in question from the launch photo\nBootblack set out the following procedures for the thread:\n1) In a joint effort (poll and comments) we will select a photo with the shadow that matches the shadow in the [photographer’s] photo. If someone’s eye fixes upon an intermediate position, just tell me.\n2) Knowing the recorded time of the selected photo, we will calculate the azimuth of the shadow. The EXIF has already been deleted to eliminate subjectivity.\n3) We will recalculate the obtained azimuth to July 17, 2014 and from that we will calculate the exact time the photographer shot the shadow.\nOne of the photographs provided by the forum administrator\nFor three pages, the forum users debated which of the twelve provided photographs best matched the shadow on the launch photo. Different users, including the administrator, threw their lots in to different ones, with various creative uses of overlaying lines over the given photographs to judge the correct angles:\nEventually, the administrator revealed the time of the photograph numbered 4328, which was in between the two leading vote-getters (4325 and 4330). This photograph was taken at August 1, 2015 at 4:35:46pm with the azimuth of 260.52, which corresponds to a July 17, 2014 time of 4:21:29pm. The administrator later calculated what he saw as a more accurate time based on the line alignment in the two photographs: 4:22:40pm on July 17, 2014.\nIf the results of the “shadow poll” experiment are trusted, the camera’s internal clock was about 3 minutes fast. It is difficult to conclusively say the exact minute–let alone the second–that the launch photograph was taken due to the margin of error in the shadow exercise and the impreciseness of the camera’s internal clock. However, this exercise in tandem with the self-reported EXIF data demonstrates that the launch photograph was indeed taken soon after 4:20pm, the time of the MH17 downing.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/07/BswldgYIUAAV7A_.jpg",
+ "alt": "BswldgYIUAAV7A_",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/08/house.jpg",
+ "alt": "house",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/08/DSCN4325.jpg",
+ "alt": "One of the photographs provided by the forum administrator",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/08/8-5-2015-6-54-14-PM.jpg",
+ "alt": "8-5-2015 6-54-14 PM",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/08/8-6-2015-5-07-14-PM.jpg",
+ "alt": "8-6-2015 5-07-14 PM",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/08/8-5-2015-6-52-46-PM.jpg",
+ "alt": "8-5-2015 6-52-46 PM",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-09-22_geolocating-russian-military-aircraft-in-syria.json b/bellingcat_geolocation_json/2015-09-22_geolocating-russian-military-aircraft-in-syria.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4633a2b6ddc5175b1b5f279142d66cd1d50b5c7
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-09-22_geolocating-russian-military-aircraft-in-syria.json
@@ -0,0 +1,47 @@
+{
+ "slug": "geolocating-russian-military-aircraft-in-syria",
+ "title": "Geolocating Russian Military Aircraft in Syria",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/09/22/geolocating-russian-military-aircraft-in-syria/",
+ "date": "2015-09-22T12:01:14+00:00",
+ "modified": "2015-09-22T12:05:48+00:00",
+ "author": "Eliot Higgins",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Over the last week there’s been a great deal of interest in the movement of Russian military equipment and troops to Syria, with images published by AllSource Analysis of al-Assad International Airport showing the arrival of military aircraft supporting Russia’s growing presence. Today a photograph was shared on Twitter showing what claimed to be the aircraft visible […]",
+ "content_html": "
Over the last week there’s been a great deal of interest in the movement of Russian military equipment and troops to Syria, with images published by AllSource Analysis of al-Assad International Airport showing the arrival of military aircraft supporting Russia’s growing presence.
\n
\n
Today a photograph was shared on Twitter showing what claimed to be the aircraft visible in the above satellite image.
\n
\n
The immediate question is whether or not this image can be verified as being taken at al-Assad International Airport, something that can be easily answered using Google Earth and the original satellite image provided by AllSource Analysis. The first question to answer is where is the camera located? We can see the aircraft are on the left of the photograph, and the white stripes are directly in front of the camera. It’s quite easy to establish that the only location that matches in the satellite image is on the left hand side of the image, with the likely camera position marked in the below image in red.
\n
\n
It’s also possible to identify the four aircraft closest to the camera as Su-30Ms, as shown in the satellite imagery, and the remaining aircraft, while not completely clear, fit with the shape of the SU-25s, also seen in the satellite imagery. It’s also possible to the see distinct markings in front of the Su-30Ms on the photograph, as shown below.
\n
\n
While not visible on the AllSource Analysis image, a radar is visible on Google Earth satellite imagery that’s also visible in the photograph.
\n
\n
Finally it’s possible to use the ground eye view in Google Earth to see that the terrain in the area matches the photograph, with the hills in the background of the photograph clearly visible on Google Earth.
\n
\n
Using these few simple geolocation techniques it is possible to confirm that this photograph shows the recent Russian arrivals at al-Assad International Airport.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nOver the last week there’s been a great deal of interest in the movement of Russian military equipment and troops to Syria, with images published by AllSource Analysis of al-Assad International Airport showing the arrival of military aircraft supporting Russia’s growing presence.\nToday a photograph was\nshared on Twitter\nshowing what claimed to be the aircraft visible in the above satellite image.\nThe immediate question is whether or not this image can be verified as being taken at al-Assad International Airport, something that can be easily answered using Google Earth and the original satellite image provided by AllSource Analysis. The first question to answer is where is the camera located? We can see the aircraft are on the left of the photograph, and the white stripes are directly in front of the camera. It’s quite easy to establish that the only location that matches in the satellite image is on the left hand side of the image, with the likely camera position marked in the below image in red.\nIt’s also possible to identify the four aircraft closest to the camera as Su-30Ms, as shown in the satellite imagery, and the remaining aircraft, while not completely clear, fit with the shape of the SU-25s, also seen in the satellite imagery. It’s also possible to the see distinct markings in front of the Su-30Ms on the photograph, as shown below.\nWhile not visible on the AllSource Analysis image, a radar is visible on Google Earth satellite imagery that’s also visible in the photograph.\nFinally it’s possible to use the ground eye view in Google Earth to see that the terrain in the area matches the photograph, with the hills in the background of the photograph clearly visible on Google Earth.\nUsing these few simple geolocation techniques it is possible to confirm that this photograph shows the recent Russian arrivals at al-Assad International Airport.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/09/1.jpg",
+ "alt": "1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/09/2.jpg",
+ "alt": "2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/09/3.jpg",
+ "alt": "3",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/09/4.jpg",
+ "alt": "4",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/09/7.jpg",
+ "alt": "7",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/09/8.jpg",
+ "alt": "8",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2015-11-11_geolocation-once-again-disproves-russias-targeting-claims-in-syria.json b/bellingcat_geolocation_json/2015-11-11_geolocation-once-again-disproves-russias-targeting-claims-in-syria.json
new file mode 100644
index 0000000000000000000000000000000000000000..ddfcbccd1a14377922ab240214d016e84d2ef426
--- /dev/null
+++ b/bellingcat_geolocation_json/2015-11-11_geolocation-once-again-disproves-russias-targeting-claims-in-syria.json
@@ -0,0 +1,47 @@
+{
+ "slug": "geolocation-once-again-disproves-russias-targeting-claims-in-syria",
+ "title": "Geolocation Once Again Disproves Russia's Targeting Claims in Syria",
+ "url": "https://www.bellingcat.com/news/2015/11/11/geolocation-once-again-disproves-russias-targeting-claims-in-syria/",
+ "date": "2015-11-11T07:18:54+00:00",
+ "modified": "",
+ "author": "Nathan Patin",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)In just three days, Russian aircraft have carried out 137 sorties, targeting 448 “terrorist objects” throughout Syria, according to a Russian Ministry of Defense (MoD) press briefing held by spokesperson Igor Konashenkov on 9 November 2015. Three airstrikes previously posted on the MoD’s YouTube page were singled out by Konashenkov for closer examination. Given the Russian MoD’s well-documented propensity for fabricating information, it’s […]",
+ "content_html": "
In just three days, Russian aircraft have carried out 137 sorties, targeting 448 “terrorist objects” throughout Syria, according to a Russian Ministry of Defense (MoD) press briefing held by spokesperson Igor Konashenkov on 9 November 2015. Three airstrikes previously posted on the MoD’s YouTube page were singled out by Konashenkov for closer examination. Given the Russian MoD’s well-documentedpropensityfor fabricatinginformation, it’s a worthwhile endeavor to verify the Russian MoD’s version of events.
\n
The first video purports to show an airstrike on an “ammunition depot” in the governorate of Raqqa. According to Konashenkov, the depot contained “anti-tank systems, grenade launchers, and small arms,” and heavy vehicles were seen being loaded with weapons, which were “delivered to terrorist groups of the region.” Highlighting that the bombed depot was indeed located in Raqqa, the MoD presentation transitions to the video by having it zoom out from the approximate location where the strike allegedly took place:
\n
\n
In this instance, the MoD implies that the strike occurred near al-Tabqa in the west of Raqqa Governorate. However, as has been the case dozens of times before, Russia incorrectly described the video’s location in both the press conference and the video itself. The airstrike was geolocated 115 kilometers (km) away near al-Bab, which is in the governorate of Aleppo:
\n
\nIt’s unclear why the Russian MoD would claim that it was targeting “terrorists” in Raqqa when al-Bab is controlled by the Islamic State (IS) and has been since early 2014.
\n
The second video is described as showing an airstrike against an IS training camp in Idlib Governorate. The difficulty here is that there hasn’t been an IS presence in Idlib since January 2014, when they were expelled by a coalition of rebel groups. Nonetheless, a Su-34 fighter jet “performed a pinpoint strike against the ammunition and weapon depot,” according to Konashenkov. The attack was geolocated in Idlib Governorate as claimed by the Russian MoD:
\n
\n
Curiously, however, the video in the MoD press conference zooms out from an area south of Ma’saran (or “Масеран” on the map in the Russian MoD presentation), roughly 20 km away from the actual site of the attack:
\n
\n
Finally, the third video that Konashenkov discusses also claims to show an attack on IS in Idlib, this time against a “strong point” containing a both a communication and command center and an ammunition depot. Geolocation of the strike places the airstrike not in Idlib but in Aleppo Governorate, near Sabiqiyah, where IS has no presence:
\n
\n
It’s possible that the airstrike was carried out in order to aid Assad’s forces fighting rebels in the area. The location of the video on the map in the Russian MoD presentation is also incorrect, presenting the attack as having taken place southwest of Ariha in Idlib Governorate, when, in fact, the attack took place in another governorate entirely, just less than 70 km away.
\n
\n
Russia’s persistent inaccuracy regarding the identities of the groups it’s targeting and the locations where it’s targeting them is all the more odd given that the Russian MoD has stated that it only carries out airstrikes when “the data on the engaged objects has been checked through several channels.” Konashenkov claims that targeting and verification information comes from the Syrian government, the Baghdad information center, and “representatives of the Syrian opposition,” though the latter source is particularly dubious. Taking the MoD spokesman at his word means that, even after triple-checking its targets before it destroys them, Russian airstrikes have still managed to exact an immense toll on civilians.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nIn just three days, Russian aircraft have carried out 137 sorties, targeting 448 “terrorist objects” throughout Syria, according to a Russian Ministry of Defense (MoD)\npress briefing\nheld by spokesperson\nIgor Konashenkov\non 9 November 2015. Three airstrikes previously posted on the MoD’s YouTube page were singled out by Konashenkov for closer examination. Given the Russian MoD’s\nwell-documented\npropensity\nfor fabricating\ninformation\n, it’s a worthwhile endeavor to verify the Russian MoD’s version of events.\nThe\nfirst video\npurports to show an airstrike on an “ammunition depot” in the governorate of Raqqa. According to Konashenkov, the depot contained “anti-tank systems, grenade launchers, and small arms,” and heavy vehicles were seen being loaded with weapons, which were “delivered to terrorist groups of the region.” Highlighting that the bombed depot was indeed located in Raqqa, the MoD presentation transitions to the video by having it zoom out from the approximate location where the strike allegedly took place:\nIn this instance, the MoD implies that the strike occurred near al-Tabqa in the west of Raqqa Governorate. However, as has been the case\ndozens of times before\n, Russia incorrectly described the video’s location in both the press conference and the video itself. The airstrike was\ngeolocated\n115 kilometers (km) away near al-Bab, which is in the governorate of Aleppo:\nIt’s unclear why the Russian MoD would claim that it was targeting “terrorists” in Raqqa when al-Bab is controlled by the Islamic State (IS) and has been since\nearly 2014\n.\nThe\nsecond video\nis described as showing an airstrike against an IS training camp in Idlib Governorate. The difficulty here is that there hasn’t been an IS presence in Idlib since\nJanuary 2014\n, when they were expelled by a coalition of rebel groups. Nonetheless, a Su-34 fighter jet “performed a pinpoint strike against the ammunition and weapon depot,” according to Konashenkov. The attack was\ngeolocated\nin Idlib Governorate as claimed by the Russian MoD:\nCuriously, however, the video in the MoD press conference zooms out from an area south of Ma’saran (or “Масеран” on the map in the Russian MoD presentation), roughly 20 km away from the actual site of the attack:\nFinally, the\nthird video\nthat Konashenkov discusses also claims to show an attack on IS in Idlib, this time against a “strong point” containing a both a communication and command center and an ammunition depot.\nGeolocation\nof the strike places the airstrike not in Idlib but in Aleppo Governorate, near Sabiqiyah, where IS has no presence:\nIt’s possible that the airstrike was\ncarried out in order to aid Assad’s forces\nfighting rebels in the area. The location of the video on the map in the Russian MoD presentation is also incorrect, presenting the attack as having taken place southwest of Ariha in Idlib Governorate, when, in fact, the attack took place in another governorate entirely, just less than 70 km away.\nRussia’s persistent inaccuracy regarding the identities of the groups it’s targeting and the locations where it’s targeting them is all the more odd given that the Russian MoD has\nstated\nthat it only carries out airstrikes when “the data on the engaged objects has been checked through several channels.” Konashenkov claims that targeting and verification information comes from the Syrian government, the Baghdad information center, and “representatives of the Syrian opposition,” though the latter source is particularly\ndubious\n. Taking the MoD spokesman at his word means that, even after triple-checking its targets before it destroys them, Russian airstrikes have still managed to exact an\nimmense toll\non\ncivilians\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/11/2015-11-10-16_27_35.gif",
+ "alt": "2015-11-10 16_27_35",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/11/video-1.jpg",
+ "alt": "video 1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/11/Video-21.jpg",
+ "alt": "Video 2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/11/ge2.jpg",
+ "alt": "ge2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/11/Video-31.jpg",
+ "alt": "Video 3",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2015/11/ge3.jpg",
+ "alt": "ge3",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2016-02-25_geolocating-mexican-sicarios-in-chihuahua.json b/bellingcat_geolocation_json/2016-02-25_geolocating-mexican-sicarios-in-chihuahua.json
new file mode 100644
index 0000000000000000000000000000000000000000..daed95ef75834dd9c082acf565a5948b2cc23766
--- /dev/null
+++ b/bellingcat_geolocation_json/2016-02-25_geolocating-mexican-sicarios-in-chihuahua.json
@@ -0,0 +1,62 @@
+{
+ "slug": "geolocating-mexican-sicarios-in-chihuahua",
+ "title": "Geolocating Mexican Sicarios in Chihuahua",
+ "url": "https://www.bellingcat.com/news/americas/2016/02/25/geolocating-mexican-sicarios-in-chihuahua/",
+ "date": "2016-02-25T21:12:50+00:00",
+ "modified": "2016-02-27T19:57:46+00:00",
+ "author": "Bellingcat Investigation Team",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "On April 08 2015, the person behind the “La Polaka” Youtube profile posted this video (YouTube). It features an unidentified cameraman walking alongside a resting convoy of about 15 vehicles and 50 heavily armed men wearing assault gears, military fatigues or civilian clothing. As the cameraman is cheering the column, radio communications can be heard between a group leader and “Meny” – likely an advanced element of the commando, scouting for possible enemies or security forces. One can hear the cameraman repeating “Animo, sicarios” which can be translated as “let’s go get them, hitmen”. The scene was reportedly filmed during late fall 2014, prior to clashes with a rival gang. Based on what can be observed in the video, it is possible to pinpoint the exact location of the deadly convoy.",
+ "content_html": "
On April 08 2015, the person behind the “La Polaka” Youtube profile posted this video (YouTube). It features an unidentified cameraman walking alongside a resting convoy of about 15 vehicles and 50 heavily armed men wearing assault gears, military fatigues or civilian clothing. As the cameraman is cheering the column, radio communications can be heard between a group leader and “Meny” – likely an advanced element of the commando, scouting for possible enemies or security forces. One can hear the cameraman repeating “Animo, sicarios” which can be translated as “let’s go get them, hitmen”. The scene was reportedly filmed during late fall 2014, prior to clashes with a rival gang. Based on what can be observed in the video, it is possible to pinpoint the exact location of the deadly convoy.
\n
Locating the video
\n
“La Polaka” mentions that the video took place in Chihuahua and indications for the towns of Guachochi and Creel can be seen at 02:00 in the video.
\n
\n
The most direct route connecting Guachochi and Creel is a paved road with a length of 158 kilometres. Considering that the indication for Creel points towards the left side of the screen and that Guachochi is on the opposite side, it can be deducted that the filming location is on the Western side of the main road linking both towns.
\n
\n
By looking at the featured environment, it appears that the location is at a T-junction, with the convoy resting on a paved secondary route connecting to the Guachochi-Creel main road. As there are very few similar sites along the way, the terrain at 27.29842, -107.5267 offers a similar landscape to what can be observed in the video.
\n
\n
During the first minute of the video, the cameraman is filming the tail end of the convoy. One can see that the road is going into a steep curve (1) and that rock walls made from excavation (2) are on both sides of the road. A serie of building (3-4-5) and vegetation (6) can also be perceived on top of the small cliff on the right-hand side of the road from the cameraman point of view – which is conform with the features of the commercial imagery.
\n
\n
As the cameraman is moving towards the head of the convoy, the left side of the road is open ground (7) while the right side presents more dense vegetation (8). Considering that the vehicles in the convoy have an average length of six meters, with an average five meters of spacing between them, the resting convoy would occupy approximately 160 meters, which is consistent with what can be measured on Google Map.
\n
\n
At the end of the video, the lead elements of the convoy are seen next to a T-junction at the main Guachochi-Creel road. At the intersection, there are a road sign (9) and a building of rectangular shape (10). However, the billboard (11), also located on the far side of the road, is not present in any commercial imagery available. According to Google Earth, the last satellite images accessible of the area were taken in June 2010, four years before the video was taken. This certainly explains why the billboard can be seen in the video, but is missing from the satellite photo.
\n
\n
Analysis
\n
Online medias have widely discussed this video. Although no one has been able to positively identify to which group the convoy of sicarios belongs to, they have been linked to different clashes that occurred around the time the video was recorded. Proceso specified that a couple of vehicles seen in the video were later found on the location of a November 2014 violent shootout in Samachique, Chihuahua, in which dozens of people were killed. InSight Crime further indicates that some of the sicarios pictured in the video were found dead on the site of a March 2015 gunfight in a ranch close to this area.
\n
During fall 2014, armed factions of the Sinaloa Cartel (Gente Nueva) and the Juarez Cartel (La Linea) were still fighting in Southern Chihuahua over trafficking routes – with the Sinaloa Cartel having the upper hand. Experts also suspect that the fighting was linked to control over resource extraction. In the recent years, drug trafficking organizations in Mexico have started to diversify their business activities and turned their attention to mining and energy industries. It is thus believed that the commando outlined in the video belongs to either the Gente Nueva or the Juarez Cartel.
\n
Gente Nueva’s soldiers originate from the Mexican security forces and are considered to be an “elite force” within the Sinaloa Cartel. They are known to operate mostly in military uniforms but in several cases, they have been seen wearing civilian clothing.
\n
La Linea is also composed of ex-members of the Mexican security forces but tend to wear civilian clothes, with only a few individuals dressing in military fashion. In 2015, La Linea reportedly separated from the Juarez Cartel and formed the Nuevo Cártel de Juarez, thus indicating the influence of the group in the area.
\n
Nevertheless, these images offer a rare insight into the crude world of armed enforcers and are a cold reminder of the level of freedom and firepower these groups still retain in some rural parts of Mexico.
\n
Since Felipe Calderón became President of Mexico in 2006, drug-related violence skyrocketed in the country. Following his decision to deploy 50, 000 soldiers by 2012 on the Mexican streets, the number of homicides tripled from 10, 452 in 2006 to 27, 213 in 2011. As Mexico and the US went after heads of cartels, they atomised the main ones into smaller entities that grew larger. Higher rates of crime were thus the direct outcome of inter- and intra-cartel gunfights as well as battles between Mexican security forces and drug trafficking organizations. The new President, Enrique Peña Nieto, has vowed to not repeat the mistakes of his predecessor. Yet events such as the 2015 Hollywood-style escape of El Chapo and the disappearance of 43 students in the city of Iguala back in 2014 have shown the extent to which cartels have penetrated the Mexican government at all levels. The only positive note was the rapid recapture of El Chapo, which has been hailed as a victory for the Mexican fight against drug trafficking organizations. This view however underestimates the rise of new and smaller cartels, such as the Jalisco New Generation Cartel, which according to experts are the future of drug trafficking in opposition to large and centralized groups like the Sinaloa cartel.
",
+ "content_text": "Translations:\nEnglish (UK)\nBackground information\nOn April 08 2015, the person behind the “La Polaka” Youtube profile posted this video (\nYouTube\n). It features an unidentified cameraman walking alongside a resting convoy of about 15 vehicles and 50 heavily armed men wearing assault gears, military fatigues or civilian clothing. As the cameraman is cheering the column, radio communications can be heard between a group leader and “Meny” – likely an advanced element of the commando, scouting for possible enemies or security forces. One can hear the cameraman repeating “Animo, sicarios” which can be translated as “let’s go get them, hitmen”. The scene was reportedly filmed during late fall 2014, prior to clashes with a rival gang. Based on what can be observed in the video, it is possible to pinpoint the exact location of the deadly convoy.\nLocating the video\n“La Polaka” mentions that the video took place in Chihuahua and indications for the towns of Guachochi and Creel can be seen at 02:00 in the video.\nThe most direct route connecting Guachochi and Creel is a paved road with a length of 158 kilometres. Considering that the indication for Creel points towards the left side of the screen and that Guachochi is on the opposite side, it can be deducted that the filming location is on the Western side of the main road linking both towns.\nBy looking at the featured environment, it appears that the location is at a T-junction, with the convoy resting on a paved secondary route connecting to the Guachochi-Creel main road. As there are very few similar sites along the way, the terrain at 27.29842, -107.5267 offers a similar landscape to what can be observed in the video.\nDuring the first minute of the video, the cameraman is filming the tail end of the convoy. One can see that the road is going into a steep curve (1) and that rock walls made from excavation (2) are on both sides of the road. A serie of building (3-4-5) and vegetation (6) can also be perceived on top of the small cliff on the right-hand side of the road from the cameraman point of view – which is conform with the features of the commercial imagery.\nAs the cameraman is moving towards the head of the convoy, the left side of the road is open ground (7) while the right side presents more dense vegetation (8). Considering that the vehicles in the convoy have an average length of six meters, with an average five meters of spacing between them, the resting convoy would occupy approximately 160 meters, which is consistent with what can be measured on Google Map.\nAt the end of the video, the lead elements of the convoy are seen next to a T-junction at the main Guachochi-Creel road. At the intersection, there are a road sign (9) and a building of rectangular shape (10). However, the billboard (11), also located on the far side of the road, is not present in any commercial imagery available. According to Google Earth, the last satellite images accessible of the area were taken in June 2010, four years before the video was taken. This certainly explains why the billboard can be seen in the video, but is missing from the satellite photo.\nAnalysis\nOnline medias have widely discussed this video. Although no one has been able to positively identify to which group the convoy of sicarios belongs to, they have been linked to different clashes that occurred around the time the video was recorded.\nProceso specified\nthat a couple of vehicles seen in the video were later found on the location of a November 2014 violent shootout in Samachique, Chihuahua, in which dozens of people were killed.\nInSight Crime further indicates\nthat some of the sicarios pictured in the video were found dead on the site of a March 2015 gunfight in a ranch close to this area.\nDuring fall 2014, armed factions of the Sinaloa Cartel (Gente Nueva) and the Juarez Cartel (La Linea) were still fighting in Southern Chihuahua over trafficking routes – with the Sinaloa Cartel having the upper hand. Experts also suspect that the fighting was linked to control over resource extraction. In the recent years, drug trafficking organizations in Mexico have started to diversify their business activities and turned their attention to mining and energy industries. It is thus believed that the commando outlined in the video belongs to either the Gente Nueva or the Juarez Cartel.\nGente Nueva’s soldiers originate from the Mexican security forces and are considered to be an “elite force” within the Sinaloa Cartel. They are known to operate mostly in military uniforms but in several cases, they have been seen wearing civilian clothing.\nLa Linea is also composed of ex-members of the Mexican security forces but tend to wear civilian clothes, with only a few individuals dressing in military fashion. In 2015, La Linea reportedly separated from the Juarez Cartel and formed the Nuevo Cártel de Juarez, thus indicating the influence of the group in the area.\nNevertheless, these images offer a rare insight into the crude world of armed enforcers and are a cold reminder of the level of freedom and firepower these groups still retain in some rural parts of Mexico.\nSince Felipe Calderón became President of Mexico in 2006, drug-related violence\nskyrocketed in the country\n. Following his decision to deploy 50, 000 soldiers by 2012 on the Mexican streets, the number of homicides tripled from 10, 452 in 2006 to 27, 213 in 2011. As Mexico and the US went after heads of cartels, they atomised the main ones into smaller entities that grew larger. Higher rates of crime were thus the direct outcome of inter- and intra-cartel gunfights as well as battles between Mexican security forces and drug trafficking organizations. The new President, Enrique Peña Nieto, has vowed to not repeat the mistakes of his predecessor. Yet events such as the 2015 Hollywood-style escape of El Chapo and the disappearance of 43 students in the city of Iguala back in 2014 have shown the extent to which cartels have penetrated the Mexican government at all levels. The only positive note was the rapid recapture of El Chapo, which has been hailed as a victory for the Mexican fight against drug trafficking organizations. This view however underestimates the rise of new and smaller cartels, such as the Jalisco New Generation Cartel, which according to experts are the future of drug trafficking in opposition to large and centralized groups like the Sinaloa cartel.\nSources\nhttp://www.insightcrime.org/news-analysis/new-video-appears-of-hit-squad-in-mine-rich-mexico-region\nhttp://www.jornada.unam.mx/2014/11/04/politica/010n2pol\nhttp://www.proceso.com.mx/?p=400706\nhttp://archive.elpasotimes.com/news/ci_27916069/video-narco-gunmen-convoy-was-filmed-last-year/\nhttps://www.youtube.com/watch?v=QQ2j5oeHsd4\nhttp://www.lapolaka.com/animo-sicarios/\nLa Linea\nhttp://www.narcoviolencia.com.mx/2015/06/fotos-al-igual-que-los-zetas-y-el-cjng.html\nGente Nueva\nhttp://www.narcoviolencia.com.mx/2015/02/la-gente-nueva-sus-origenes-del.html\nBy Vince. Edited by\nAliaume\nLeroy",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/02/1-2.jpg",
+ "alt": "1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/02/2-2.jpg",
+ "alt": "2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/02/3-2.jpg",
+ "alt": "3",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/02/4-2.jpg",
+ "alt": "4",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/02/5-2.jpg",
+ "alt": "5",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/02/6-1.jpg",
+ "alt": "6",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/02/7-1.jpg",
+ "alt": "7",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/02/8-1.jpg",
+ "alt": "8",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/02/9.1.jpg",
+ "alt": "9.1",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2016-10-21_geolocation-turkish-airstrikes-ypg-positions.json b/bellingcat_geolocation_json/2016-10-21_geolocation-turkish-airstrikes-ypg-positions.json
new file mode 100644
index 0000000000000000000000000000000000000000..953a60547d220095079b9f2e8d7ffde6fd6ed1d7
--- /dev/null
+++ b/bellingcat_geolocation_json/2016-10-21_geolocation-turkish-airstrikes-ypg-positions.json
@@ -0,0 +1,52 @@
+{
+ "slug": "geolocation-turkish-airstrikes-ypg-positions",
+ "title": "Geolocation of Turkish Airstrikes on YPG Positions",
+ "url": "https://www.bellingcat.com/news/2016/10/21/geolocation-turkish-airstrikes-ypg-positions/",
+ "date": "2016-10-21T12:52:25+00:00",
+ "modified": "2016-10-23T12:50:30+00:00",
+ "author": "Christiaan Triebert",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)On the night from October 19 to October 20, the Turkish Air Force bombed positions allegedly of the People’s Protection Units (Kurdish: Yekîneyên Parastina Gel, YPG) in Syria’s Aleppo governorate. This is a brief overview of open source information on the incident. Geolocation In total, the Turkish Air Force dropped 26 bombs on 18 […]",
+ "content_html": "
On the night from October 19 to October 20, the Turkish Air Force bombed positions allegedly of the People’s Protection Units (Kurdish: Yekîneyên Parastina Gel, YPG) in Syria’s Aleppo governorate. This is a brief overview of open source information on the incident. \n
\n
Geolocation
\n
In total, the Turkish Air Force dropped 26 bombs on 18 different targets, including nine buildings used as shelters, and an arsenal, according to Turkey’s state-run press service Anadolu Agency, citing a military source. A 1-minute and 56-second video showing six different strikes, all of which have been geolocated by Twitter users [tweet, website].
\n
Five out of six strikes have been geolocated, as shown on the map below.
\n
\n
Strike 1 (0:00-0:17)
\n
The first airstrike has been geolocated by Twitter user @was_algonquin to 36.390068, 37.253568 (Wikimapia).
\n
\n
Strike 2 (0:17-0:38)
\n
The second strike has been also geolocated by @was_algonquin to 36.39277, 37.25621 (Wikimapia).
\n
\n
Strike 3 (0:38-0:56)
\n
The third airstrike has again been geolocated by @was_algonquin to 36.399254, 37.256873 (Wikimapia).
\n
\n
Strike 4 (0:56-1:15)
\n
The fourth airstrike has not been geolocated yet.
\n
Strike 5 (1:15-1:36)
\n
The airstrike has been geolocated by Twitter user @fevciinzii was geolocated to 36.411509, 37.2348118 (Wikimapia).
\n
\n
Strike 6 (1:36-1:56)
\n
The sixth strike was also geolocated by @fevciinzii to 36.4051194, 37.2297049 (Wikimapia).
\n
\n
Mapped on a base layer of syria.liveuamap.com, the airstrikes were indeed in YPG-held territory, close to the frontline with Turkish-backed Free Syrian Army (FSA) fighters. The locations of the airstrikes are marked in pink.
\n
\n
The location of the airstrikes is very close near the village of Hassadjek, the location where Russia accused Belgium of killing six civilians due to airstrikes on October 18, 2016. Hassadjek was reportedly captured by the YPG on September 26, 2016.
\n
Information related the incident was “handed over to the Russian Reconciliation Center in Syria by the local council of the Syrian city of Afrin,” Russia Todaywrote. The Russian Ministry of Defence said neither Syrian nor Russian jets were carrying out airstrikes in the area.
\n
However, Belgium has accused the Russian MoD of faking data about the airstrike (Novaya Gazeta). Belgian Defence Minister Steven Vandeput tweeted that it is “Russian disinformation” [archived]:
\n
\n
Casualties
\n
The Turkish military said it killed between 160 and 200 YPG fighters, according to Anadolu Agency.
\n
However, senior YPG-commander Mahmoud Barkhadan told The Associated Press that the death toll was much lower, putting it at no more than ten casualties and around twenty injured. Mr. Barkhadan said that Turkish tanks had been shelling their positions since October 20 and were joined by military aircraft overnight.
\n
Context
\n
Both Turkey and the YPG are fighting against so-called Islamic State (IS) in Syria. But despite their common goal, Turkey perceives the YPG, and it’s political arm PYD as well as the broader coalition it fronts, the SDF, as extensions of the Kurdistan Workers’ Party (Kurdish: Partiya Karkerên Kurdistanê, PKK). Turkey views YPG and SDF territorial gains long its border with Syria as a national seucrity threat.
\n
This has caused friction between Turkey and its NATO ally the United States (US), as the latter has provided military support to the SDF/YPG. In other words, NATO-member “Turkey Bombs U.S.-Backed Kurds in Syria’s Aleppo Province” (NBC News). However, not much support has been provided to the YPG in Afrin.
\n
Turkey has started a cross-border operation with allied groups on August 24, code-named by Turkey as Operation Euphrates Shield.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nOn the night from October 19 to October 20, the Turkish Air Force bombed positions allegedly of the People’s Protection Units (Kurdish:\nYekîneyên Parastina Gel\n, YPG) in Syria’s Aleppo governorate. This is a brief overview of open source information on the incident.\nGeolocation\nIn total, the Turkish Air Force dropped 26 bombs on 18 different targets, including nine buildings used as shelters, and an arsenal, according to Turkey’s state-run press service\nAnadolu Agency\n, citing a military source. A\n1-minute and 56-second video showing six different strikes, all of which have been geolocated by Twitter users [\ntweet\n,\nwebsite\n].\nFive out of six strikes have been geolocated, as shown on the map below.\nStrike 1 (0:00-0:17)\nThe first airstrike has been geolocated by Twitter user\n@was_algonquin\nto 36.390068, 37.253568 (\nWikimapia\n).\nStrike 2 (0:17-0:38)\nThe second strike has been also geolocated by\n@was_algonquin\nto 36.39277, 37.25621 (\nWikimapia\n).\nStrike 3 (0:38-0:56)\nThe third airstrike has again been geolocated by\n@was_algonquin\nto 36.399254, 37.256873 (\nWikimapia\n).\nStrike 4 (0:56-1:15)\nThe fourth airstrike has not been geolocated yet.\nStrike 5 (1:15-1:36)\nThe airstrike has been geolocated by Twitter user\n@fevciinzii\nwas geolocated to 36.411509, 37.2348118 (\nWikimapia\n).\nStrike 6 (1:36-1:56)\nThe sixth strike was also geolocated by\n@fevciinzii\nto 36.4051194, 37.2297049 (\nWikimapia\n).\nMapped on a base layer of\nsyria.liveuamap.com\n, the airstrikes were indeed in YPG-held territory, close to the frontline with Turkish-backed Free Syrian Army (FSA) fighters. The locations of the airstrikes are marked in pink.\nThe location of the airstrikes is very close near the village of Hassadjek, the location where Russia accused Belgium of killing six civilians due to airstrikes on October 18, 2016. Hassadjek was\nreportedly\ncaptured by the YPG on September 26, 2016.\nInformation related the incident was “handed over to the Russian Reconciliation Center in Syria by the local council of the Syrian city of Afrin,”\nRussia Today\nwrote.\nThe Russian Ministry of Defence said neither Syrian nor Russian jets were carrying out airstrikes in the area\n.\nHowever, Belgium has accused the Russian MoD of faking data about the airstrike (\nNovaya Gazeta\n). Belgian Defence Minister Steven Vandeput\ntweeted\nthat it is “Russian disinformation” [\narchived\n]:\nCasualties\nThe Turkish military said it killed between 160 and 200 YPG fighters, according to\nAnadolu Agency\n.\nHowever, senior YPG-commander Mahmoud Barkhadan told\nThe Associated Press\nthat the death toll was much lower, putting it at no more than ten casualties and around twenty injured. Mr. Barkhadan said that Turkish tanks had been shelling their positions since October 20 and were joined by military aircraft overnight.\nContext\nBoth Turkey and the YPG are fighting against so-called Islamic State (IS) in Syria. But despite their common goal, Turkey perceives the YPG, and it’s political arm PYD as well as the broader coalition it fronts, the SDF, as extensions of the Kurdistan Workers’ Party (Kurdish:\nPartiya Karkerên Kurdistanê\n, PKK). Turkey views YPG and SDF territorial gains long its border with Syria as a national seucrity threat.\nThis has caused friction between Turkey and its NATO ally the United States (US), as the latter has provided military support to the SDF/YPG. In other words, NATO-member “Turkey Bombs U.S.-Backed Kurds in Syria’s Aleppo Province” (\nNBC News\n). However, not much support has been provided to the YPG in Afrin.\nTurkey has started a cross-border operation with allied groups on August 24, code-named by Turkey as Operation Euphrates Shield.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/10/Strike-1-1200x472.jpg",
+ "alt": "strike-1",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/10/strike-2-1.jpg",
+ "alt": "strike-2",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/10/Strike-3-1200x387.jpg",
+ "alt": "strike-3",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/10/strike-5-1200x487.jpg",
+ "alt": "strike-5",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/10/strike-6-1200x498.png",
+ "alt": "strike-6",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/10/overview.png",
+ "alt": "overview",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2016/10/vandeput-1.png",
+ "alt": "vandeput",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2017-04-04_photographs-ukrainian-buks-geolocated-nearly-three-years-later.json b/bellingcat_geolocation_json/2017-04-04_photographs-ukrainian-buks-geolocated-nearly-three-years-later.json
new file mode 100644
index 0000000000000000000000000000000000000000..840adb8d84102813446181be881281b3e52841d1
--- /dev/null
+++ b/bellingcat_geolocation_json/2017-04-04_photographs-ukrainian-buks-geolocated-nearly-three-years-later.json
@@ -0,0 +1,52 @@
+{
+ "slug": "photographs-ukrainian-buks-geolocated-nearly-three-years-later",
+ "title": "Photographs of Ukrainian Buks Geolocated in Poltava Oblast, Nearly Three Years Later",
+ "url": "https://www.bellingcat.com/news/europe/2017/04/04/photographs-ukrainian-buks-geolocated-nearly-three-years-later/",
+ "date": "2017-04-04T18:41:32+00:00",
+ "modified": "2022-01-18T15:52:09+00:00",
+ "author": "Bellingcat Investigation Team",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)On July 15, 2014, a chaplain for the Ukrainian Armed Forces named O. (short for “otets,” or “father”) Gredil Mykhailo uploaded a handful of photographs showing a Ukrainian Buk missile launcher in a field that appeared to function as a Ukrainian military camp. We can now say that this Buk missile launcher was not located […]",
+ "content_html": "
On July 15, 2014, a chaplain for the Ukrainian Armed Forces named O. (short for “otets,” or “father”) Gredil Mykhailo uploaded a handful of photographs showing a Ukrainian Buk missile launcher in a field that appeared to function as a Ukrainian military camp. We can now say that this Buk missile launcher was not located in eastern Ukraine, but instead in the Poltava Oblast in central Ukraine.
\n
The photographs would not be noteworthy if it were not for two things: the fact that Mykhailo uploaded the photographs two days before the downing of Malaysian Airlines Flight 17 (July 17, 2014), and that he included geotags indicating that the photographs were taken near Donetsk in eastern Ukraine. The apparent location of these photographs seemed to contradict statements from the Ukrainian government claiming that there were no operational Buks in the zone of the “Anti-Terrorist Operation,” or the “ATO zone.” These photographs gained minor popularity in some circles online after a user uploaded a series of photographs to an imgur album showing saved copies of Mykhailo’s photographs, including a geotagged photograph of the Buk missile launcher. One of these photographs shows Buk missile launcher 210 of the second battalion of the 223rd Anti-Aircraft Regiment, which is based in Stryi in western Ukraine. The photograph was geotagged in the Yasinuvatsky rayon (district), north of Donetsk. Another photograph, uploaded to the popular Russian forum Yaplakal, shows Mykhailo inside a control unit, geotagged in the Perevalsky rayon near Debaltseve.
\n
However, after nearly three years, the photographs have been geolocated thanks to members of the Smolensk Forum. Mykhailo’s photographs were not actually taken in eastern Ukraine, but instead on the Myrhorod Air Base in the Poltava Oblast, hundreds of kilometers from the front lines of the war.
\n
The most notable photograph uploaded by Mykhailo, showing Buk missile launcher 210 with an unusual paint job typical for Buks of the 223rd Anti-Aircraft Regiment. Source (geotag removed)
\n
Geotag for the photograph with the Buk missile launcher, indicating the location of the Yasinuvatsky area, north of Donetsk.
\n
According to the imgur post, Mykhailo replaced the geotags with a location near Moscow, Russia after people found out about some of his photographs geotagged in the ATO zone. He also claimed that the previous images were photoshopped. However, Mykhailo’s own reaction of 2 August 2014 does not mention this, but only that the location is hundreds of kilometers from Donetsk.
\n
An article and photographs of May 10, 2014 on Mykhailo’s church website make clear that he visited a military camp before in May 2014. Two photographs in the article show Buk missile launcher 231 and Buk missile loader 222, which belong to the 2nd battalion of the 223rd Anti-Aircraft Regiment. One of the photographs shows, apparently, an airfield landing strip. Comparing May and July 2014 photographs of Mykhailo’s visits to the air defense base makes clear that he visited the same military camp in the area.
\n
Comparison between May and July 2014 photographs of Mykhailo’s visit to a military camp.
\n
Another comparison between May and July 2014 photographs of Mykhailo’s visit to a military camp, with the May 2014 photograph showing what seems to be an airfield landing strip.
\n
Geolocation
\n
User “Vam” from the Smolensk Forum began reexamining the location of the “chaplain photos” and also noticed similarities between the July 2014 photographs posted by Mykhailo and May 2014 photographs on Mykhailo’s church website. The same user later noticed an article on the site “Catholic Columnist” published on July 14, 2014 about a visit from Mykhailo to air defense troops. In this article, it’s described how the chaplain brought supplies to soldiers near Poltava before moving to the ATO zone, further south and east. This is a likely explanation for why the geotags were set for the Yasinuvatsky and Perevalsky areas, as Mykhailo may have uploaded the photographs of the air defense military camp a a few days later while traveling in the ATO zone.
\n
Users “Andrew Andreev” and “Vam” from the Smolensk Forum then connected all the dots and figured out the location — Myrhorod, in the Poltava Oblast, located in between Kyiv and Donetsk. A grain elevator in the distance of Mykhailo’s July 15, 2014 photograph showing Buk missile launcher 210 was geolocated to 49,940772, 33.607480.
\n
Furthermore, an April 2014 video of military equipment–including Buk vehicles–being transported in Myrhorod in April 2014 was found. This video includes Buk missile launchers 210 and 231, along with Buk missile loader 222 of the 2nd battalion of the 223rd Anti-Aircraft Regiment–visible in Mykhailo’s May and July 2014 photographs.
\n\n
Finally, and most importantly, the military camp that includes the Buks were found on satellite imagery, at 49.919630, 33.629099.
\n
Buks and other military equipment on August 14, 2014 (Google Earth)
\n
The satellite imagery of August 14, 2014 does not show the Buk missile launchers very clearly, but satellite imagery from August 17, 2015 and February 21, 2016 show the same military camp, including Buk missile launchers, more clearly. The missiles mounted to the missile launchers are clearly visible with this higher-quality satellite imagery.
\n
Comparison between August 14, 2014 and February 21, 2016 satellite imagery, showing the same military camp, including Buk missile launchers loaded with missiles.
Several structures visible in Mykhailo’s May and July 2014 photographs include: a grain elevator (top left), a chimney (second from left), another grain elevator and two other buildings (second from right) and two tall apartment buildings (right).
\n
Another question answered
\n
With this photograph geolocated, another question regarding the location of Buk missile launchers at the time of the downing of MH17 has been answered. Previously, footage aired on Ukrainian Military TV the day before the downing was a significant question, but we were able to geolocate this footage in March 2016 in the article “Ukrainian Buk Footage from July 16, 2014 Geolocated.” Now, with the geolocation of the Mykhailo photographs, we see another instance in which Ukrainian Buks thought to be within range of downing MH17 were, in reality, quite far from the MH17 crash site, let alone the front lines. Still, after nearly three years, the only operational Buk missile launcher observed within firing range of MH17 on July 17, 2014 was the Russian Buk 332, from the Kursk-based 53rd Anti-Aircraft Missile Brigade.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nOn July 15, 2014, a chaplain for the Ukrainian Armed Forces named O. (short for “otets,” or “father”) Gredil Mykhailo uploaded a handful of photographs showing a Ukrainian Buk missile launcher in a field that appeared to function as a Ukrainian military camp. We can now say that\nthis Buk missile launcher was not located in eastern Ukraine\n, but instead in the Poltava Oblast in central Ukraine.\nThe photographs would not be noteworthy if it were not for two things: the fact that Mykhailo uploaded the photographs two days before the downing of Malaysian Airlines Flight 17 (July 17, 2014), and that he included geotags indicating that the photographs were taken near Donetsk in eastern Ukraine. The apparent location of these photographs seemed to contradict statements from the Ukrainian government claiming that there were no operational Buks in the zone of the “Anti-Terrorist Operation,” or the “ATO zone.” These photographs gained minor popularity in some circles online after a user uploaded a series of photographs to\nan imgur album\nshowing saved copies of Mykhailo’s photographs, including a geotagged photograph of the Buk missile launcher. One of these photographs shows Buk missile launcher 210 of the second battalion of the 223\nrd\nAnti-Aircraft Regiment, which is based in Stryi in western Ukraine. The photograph was geotagged in the Yasinuvatsky rayon (district), north of Donetsk. Another photograph, uploaded to the popular Russian forum\nYaplakal\n, shows Mykhailo inside a control unit, geotagged in the Perevalsky rayon near Debaltseve.\nHowever, after nearly three years, the photographs have been geolocated thanks to members of the\nSmolensk Forum\n. Mykhailo’s photographs were not actually taken in eastern Ukraine, but instead on the Myrhorod Air Base in the Poltava Oblast, hundreds of kilometers from the front lines of the war.\nThe most notable photograph uploaded by Mykhailo, showing Buk missile launcher 210 with an unusual paint job typical for Buks of the 223rd Anti-Aircraft Regiment.\nSource (geotag removed)\nGeotag for the photograph with the Buk missile launcher, indicating the location of the Yasinuvatsky area, north of Donetsk.\nAccording to the\nimgur post\n, Mykhailo replaced the geotags with a location near Moscow, Russia after people found out about some of his photographs geotagged in the ATO zone. He also claimed that the previous images were photoshopped. However, Mykhailo’s own reaction of\n2 August 2014\ndoes not mention this, but only that the location is hundreds of kilometers from Donetsk.\nAn article and photographs of\nMay 10, 2014\non Mykhailo’s church website make clear that he visited a military camp before in May 2014. Two photographs in the article show Buk missile launcher 231 and Buk missile loader 222, which belong to the 2\nnd\nbattalion of the 223\nrd\nAnti-Aircraft Regiment. One of the photographs shows, apparently, an airfield landing strip. Comparing May and July 2014 photographs of Mykhailo’s visits to the air defense base makes clear that he visited the same military camp in the area.\nComparison between May and July 2014 photographs of Mykhailo’s visit to a military camp.\nAnother comparison between May and July 2014 photographs of Mykhailo’s visit to a military camp, with the May 2014 photograph showing what seems to be an airfield landing strip.\nGeolocation\nUser “Vam” from the Smolensk Forum\nbegan reexamining the location of the “chaplain photos”\nand also\nnoticed similarities\nbetween the July 2014 photographs posted by Mykhailo and May 2014 photographs on Mykhailo’s church website. The same user later noticed an article on the site “\nCatholic Columnist\n” published on July 14, 2014 about a visit from Mykhailo to air defense troops. In this article, it’s described how the chaplain brought supplies to soldiers near Poltava before moving to the ATO zone, further south and east. This is a likely explanation for why the geotags were set for the Yasinuvatsky and Perevalsky areas, as Mykhailo may have uploaded the photographs of the air defense military camp a a few days later while traveling in the ATO zone.\nUsers “Andrew Andreev” and “Vam” from the Smolensk Forum then connected all the dots and figured out the location — Myrhorod, in the Poltava Oblast, located in between Kyiv and Donetsk. A grain elevator in the distance of Mykhailo’s July 15, 2014 photograph showing Buk missile launcher 210 was geolocated to\n49,940772, 33.607480\n.\nFurthermore, an April 2014 video of military equipment–including Buk vehicles–being transported in Myrhorod in April 2014 was found. This video includes Buk missile launchers 210 and 231, along with Buk missile loader 222 of the 2nd battalion of the 223rd Anti-Aircraft Regiment–visible in Mykhailo’s May and July 2014 photographs.\nFinally, and most importantly, the military camp that includes the Buks were found on satellite imagery, at\n49.919630, 33.629099\n.\nBuks and other military equipment on August 14, 2014 (Google Earth)\nThe satellite imagery of August 14, 2014 does not show the Buk missile launchers very clearly, but satellite imagery from August 17, 2015 and February 21, 2016 show the same military camp, including Buk missile launchers, more clearly. The missiles mounted to the missile launchers are clearly visible with this higher-quality satellite imagery.\nComparison between August 14, 2014 and February 21, 2016 satellite imagery, showing the same military camp, including Buk missile launchers loaded with missiles.\nAside from the\ngrain elevator\nfrom a company named\n‘Myrhorod feed mill’\n, several other structures visible in Mykhailo’s May and July 2014 photographs can be found in satellite imagery. A\nchimney\nof a factory named\n‘Armaprom’\n, is visible\nhere\n, another\ngrain elevator\nand two other buildings of a company named\n‘Mirgorodsky elevator’\n, are visible\nhere\n, and two high\napartment buildings\n, are visible\nhere\n.\nSeveral structures visible in Mykhailo’s May and July 2014 photographs include: a grain elevator (top left), a chimney (second from left), another grain elevator and two other buildings (second from right) and two tall apartment buildings (right).\nAnother question answered\nWith this photograph geolocated, another question regarding the location of Buk missile launchers at the time of the downing of MH17 has been answered. Previously, footage aired on Ukrainian Military TV the day before the downing was a significant question, but we were able to geolocate this footage in March 2016 in the article “\nUkrainian Buk Footage from July 16, 2014 Geolocated\n.” Now, with the geolocation of the Mykhailo photographs, we see another instance in which Ukrainian Buks thought to be within range of downing MH17 were, in reality, quite far from the MH17 crash site, let alone the front lines. Still, after nearly three years, the only\noperational\nBuk missile launcher observed within firing range of MH17 on July 17, 2014 was the Russian Buk 332, from the Kursk-based 53rd Anti-Aircraft Missile Brigade.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/03/0mvEb4QXPjw.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/03/lRNxJUF.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/comp_mykhailo.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/mykhailo_comp2.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/2014buksmykhiailo.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/comb-1-1200x446.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/features_mykhailo.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2017-05-05_geolocated-coalition-videos.json b/bellingcat_geolocation_json/2017-05-05_geolocated-coalition-videos.json
new file mode 100644
index 0000000000000000000000000000000000000000..aeb2c803ca9b4a66c1c1e4f5215247f3d67f4d0f
--- /dev/null
+++ b/bellingcat_geolocation_json/2017-05-05_geolocated-coalition-videos.json
@@ -0,0 +1,92 @@
+{
+ "slug": "geolocated-coalition-videos",
+ "title": "Crowdsourced Geolocation and Analysis of Coalition Airstrike Videos from Syria and Iraq",
+ "url": "https://www.bellingcat.com/news/middle-east/2017/05/05/geolocated-coalition-videos/",
+ "date": "2017-05-05T13:12:16+00:00",
+ "modified": "2017-07-10T17:33:33+00:00",
+ "author": "Christiaan Triebert",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)A version of this article was also published on Airwars. The international Coalition fighting the so-called Islamic State (IS) has published dozens of videos purportedly showing airstrikes against targets of the group in Syria and Iraq through the video-sharing social medium YouTube. Bellingcat has crowdsourced the geolocation of these videos, and this article serves […]",
+ "content_html": "
A version of this article was also published on Airwars.
\n
The international Coalition fighting the so-called Islamic State (IS) has published dozens of videos purportedly showing airstrikes against targets of the group in Syria and Iraq through the video-sharing social medium YouTube. Bellingcat has crowdsourced the geolocation of these videos, and this article serves as an analysis of those geolocated airstrike videos.
\n
Content
\n\n
Introduction
\n
Background Information on ‘The Coalition’
\n
Geolocation and ‘Strike Regions’
\n
Claimed Targets: ‘Vehicles’, ‘Factories’ and ‘Fighting Positions’
\n
Mapping the Coalition’s Region Labels
\n
Highlighted Incidents
\n\n
Introduction
\n
In reaction to the rapid territorial gains made by the so-called Islamic State (IS) in the first half of 2014, accompanied by the reported human rights abuses and the fear of further conflict spill-overs, the United States (US) started to intervene militarily against the group in Syria and Iraq. In October of that same year, the US started coordinating the military efforts of a dozen of allied countries under the banner of the Combined Joint Task Force (CJTF; often referred to as ‘the Coalition’) engaging in Operation Inherent Resolve (OIR). Today, May 5, 2017, marks the 1,001 day of the campaign.
\n
While earlier videos were posted by the US Central Command (CENTCOM), the first video depicting an airstrike was uploaded to the Coalition’s official YouTube channel on April 11, 2015. Over 200 airstrike videos followed in the nearly two years since, which Bellingcat has aimed to geolocated by using the collaborative crowdsourcing platform Check developed by Meedan, a company that builds digital tools for global journalism and translation.
\n
A typical airstrike video uploaded by the Coalition to their YouTube channel. This airstrike video, allegedly on “three ISIL buildings near Ramadi, Iraq” is one of the very few videos that is still accessible via the YouTube channel.
\n
The Coalition is composed of several countries, but by far the majority (around 68% as of April 24, 2017) of OIR’s airstrikes have been conducted by the US Air Force, as the following tables by monitoring organisation Airwars show. The tables are estimates of US and allied strikes in both Iraq and Syria.
\n
It is important to note that Coalition airstrike videos are also disseminated through other channels, such as the Ministries of Defence of Coalition members, including the Brits, the French, the Jordanians, and the Iraqis. Besides, videos have been uploaded by YouTube users seemingly affiliated with US Air Force’s F/A-18 Hornets and Rockwell B-1 Lancer aircraft conducting bombing raids above Syria and Iraq. These airstrike videos were not published previously elsewhere to our knowledge. While we have crowdsourcing projects running for those specific MoD videos as well, they are not included in this analysis.
\n
The number of airstrikes by the US and its allies, accurate as of April 24, 2017. Courtesy of Airwars.
\n
The estimated number of airstrikes in Iraq. The chart on the left shows the US and allied strikes in Iraq, while the right-hand chart gives a breakdown of allied-only strikes. All data is drawn by Airwars from official military reports of the Coalition (CJTF-OIR), France, the UK, Belgium, Denmark, the Netherlands, Australia, and Canada. Courtesy of Airwars.
\n
In recent weeks, official videos showing airstrikes against targets of the so-called Islamic State (IS) in Syria and Iraq have been removed from the public YouTube channel of the Coalition. In a written statement to Bellingcat, Coalition says it has removed the videos from YouTube and now posts higher-quality videos to DVIDS for “greater transparency and increased availability”. However, an initial assessment appears to show that not all videos have been migrated. Coalition officials have also given a different account to Airwars in the past as to why the videos were removed, suggesting their presence on the official YouTube channel no longer matched strategic goals. Airwars is permanently archiving all known Coalition and CENTCOM videos issued since August 2014, to ensure their continied availability.
\n
Given the recent removal of these videos from YouTube, Bellingcat presents its analysis of all those geolocated YouTube videos. We will continue to monitor, analyse, and geolocate the videos published to DVIDS. You can join the effort via Check‘s Bellingcat page, where you can register an account.
\n
Overview of Geolocations
\n
So far, 67 percent of the airstrikes shown in the Coalition airstrike videos have been successfully geolocated. You can access all our data, which will be updated as soon as there are new geolocations, including from DVIDS HUB, on Silk.
\n
\n
The term ‘airstrike’ is somewhat imprecise, but in this analysis we consider ‘one airstrike’ as one or more multiple airstrikes at one location. Geolocation is the determination of the exact locations that are being targeted by airstrikes in the videos. Most of the videos have been geolocated by Twitter users @obretix and @miminas999, as well as the author of this article.
\n
By far most of the strikes shown in videos uploaded to YouTube (as of April 28, 2017) were geolocated to Iraq.
\n
By far most of the strikes shown in videos uploaded to YouTube (as of April 28, 2017) were geolocated to Iraq.
\n
Broken down by provinces, the highest number of airstrikes were geolocated to the Iraqi governorates of Nineveh and Anbar, followed by the Syrian governorate of Aleppo, as of April 28, 2017.
\n
Most airstrike shown in videos were geolocated to the Iraqi governorates of Nineveh and Anbar, followed by the Syrian governorate of Aleppo, as of April 28, 2017.
\n
Exploring the Coalition’s Location Accuracy
\n
Generally, the Coalition gives an indication of a geographical location by labelling the videos “near [location X]”. There are only a handful of videos that do not contain the word “near” but simply mention a location. This analysis considers “near” as being within a 10 km range of the claimed location and is labelled as “accurate” when it falls within that range. Of all geolocated videos, 68 percent was labelled as accurate. The video outside of the 10 km range went up to around 93 km of the claimed location, and for one video no location approximation was given.
\n
Many of the videos with a significant distance from the claimed location are oil-related facilities that are indeed ‘near’ Deir ez-Zor or Al-Bukamal, such as an oil separation facility at the Al-Ahmar oil field. With the locations being in the desert with no other locations near, these location claims can still be considerate as relatively accurate.
\n
However, there are other incidents that appear to be less concisely located. Perhaps the most interesting incident of all the geolocated videos was a strike on an IS “concealed tactical vehicle” that was claimed to have been conducted on March 23, 2015, which was labelled as “near Al Hawl”, a town in north-eastern Syria. However, the targeted building has been successfully geolocated to a building in Jayar Ghalfas, a town in northwest Iraq.
\n
A screenshot from a Coalition video claiming to show an airstrike on an IS ‘concealed tactical vehicle’ near Al-Hawl, a town in eastern Syria. The building was geolocated to Jayar Ghalfas, a town in northwest Iraq, as the Microsoft Bing satellite imagery (36.137411, 41.297414) on the right shows. The location is around 30 km southwest of Al-Hawl.
\n
Though this video was labelled as being in a different country than where it actually took place, it is still relatively near Al-Hawl — around 30 km away.
\n
When Col. Steve Warren, at the time the Coalition’s spokesperson, gave an “Ask Me Anything” on the social media and news aggregation website Reddit, the author asked him about this particular incident. Col. Warren replied that this was “an administrative error that it’s listed as Syria rather than Iraq”, explaining that Al-Hawl in Syria “was the nearest identifiable city to the strike.”
\n
The reply by Col. Warren, the Coalition’s spokesperson on the question why it was labelled near to a town in Syria but showed a location in Iraq.
This strike was not an official Coalition attack – and was instead the US unilaterally targeting alleged al-Qaida fighters. The US carries out nearly all of the alliance’s anti-IS bombings in Syria, and military assets can be used for both campaigns.
\n
When asked for clarification about this incident, a CENTCOM spokesperson told Bellingcat that they “don’t mean to cause any confusion. Different internal reports may have listed this differently.”
\n
Either way, the Coalition thus seems to use a limited number of labels for their targeted location areas. The “Al Hawl, Syria” label was probably closer than their nearest other Iraqi location label, “Sinjar”, around 50 km northeast of Jayar Ghalfas.
\n
The Coalition’s ‘Region Labels’
\n
Which region does the Coalition use to label one airstrike as “near Mosul” but the other one as “near Al Hawl”? To get a better insight in which regions are used by the Coalition, the geolocator @obretix mapped all geolocated airstrike videos and then used a Voronoi diagram which is a partitioning of a plane into regions based on distance to points in a specific subset of the plane. In this case, the points are thus all “near” locations mentioned by the Coalition. The geolocations are then corresponding to a region that is closer than any other point.
\n\n
As the following image shows, Al-Hawl is indeed the closest location to the target struck by the Coalition (circled in red) in the number of areas the Coalition has used.
\n
An excerpt from a Voronoi diagram of an impression of the regions used by the Coalition, based on the geolocated videos. The geolocated airstrike of March 23, 2016 that was labelled near ‘Al Hawl’ is circled in red. Map by ‘obretix’.
\n
There are more interesting insights the maps reveals. While the Coalition does use a label for “near Kubaysah”, a city in Iraq, some strikes within the city’s perimeters were labelled as “near Hit” — a city nearby but still less accurate than using a label there was for that city.
\n
A detailed view on the Kubaysah/Hit region on the Voronoi diagram map, showing that two videos labelled as “near Hit” were in fact closer to Kubaysah, which also has its own ‘location label’.
\n
Another example of remarkable region labels is the use of hamlets, such as “near Washiya” and “near Sultan Abdullah” – places with only a few houses, but close to respectively Aleppo and Mosul. It is worth mentioning that “near Aleppo” is not used in any of the YouTube videos, while “near Washiya” has been twice for a target only a double dozen kilometres away from Aleppo city. Why would these small hamlets be used as a region label, while the case of Jayar Ghalfas could not get its own region label? Is this intentional? This is a question that remains unanswered.
\n
What is clear is that it the publicly provided locations by the Coalition for its airstrikes in Iraq and Syria may be off as much as 90+ kilometres, making it more difficult for outside monitors to verify and report on the civilian death toll of the air war.
\n
In its Transparency Audit of the Coalition, published in December 2016, Airwars noted problems with the public reporting process. Difficult to navigate internal logs “in turn led to quite vague military reporting”. Locations, then, could only be taken as approximate. One CENTCOM senior official explained the situation in some detail:
\n
“When the aircrew come back [from a strike mission], as you drill into a geographic location, some of those areas have towns that consist of three or four people. So typically what’s going to be in the strike log is going to be the largest city nearby. And they’ll annotate, ‘Conducted a strike near Mosul.’ In fact it’s going to be some small town that’s 23 clicks [kilometres] outside of Mosul. If they put that on the strike log, once it goes through the ‘Enterprise’ [slang for the Combined Air Operations Centre] no one knows where that is. Officials were keen to stress that if an incident was being investigated, “we do have the ability to go back and drill down into the detail.”
\n
Overall, all location claims were all within a 100 km distance of the claimed location, and all of these claims were relatively accurate as to the location it referred to — unlike the Russian airstrike videos, which were in some cases massively inaccurate.
\n
Claimed Targets: ‘Vehicles’, ‘Factories’ and ‘Fighting Positions’
\n
With regards to the targets, the most used targets mentioned can be categorised as “fighting positions”, followed by vehicles and facilities for improvised explosive devices (IEDs) or suicide vehicle borne IEDs (SVBIEDs).
\n
\n
On the left, a still from a Coalition video showing an airstrike on a IS ‘homemade explosives cache near Bayji, Iraq’ allegedly on October 14, 2015. The location shown in the video, which was uploaded two days later, is around 10 km away from Bayji at 35.052918, 43.552203.
\n
Most of the targets are classified as ‘fighting positions’ by the Coalition. These are usually buildings that are, reportedly, used by IS militants as positions in battle. An example can be seen in the following airstrikes on buildings in Manbij, reportedly on July 21 and 24, 2016, and geolocated by @obretix to 36.522428, 37.962093 and 36.517496, 37.955025, respectively. The strikes were in support of the Syrian Democratic Forces (SDF) during the Manbij offensive, codenamed Operation Marty and Commander Faysal Abu Layla by the group. The offensive lasted almost three months and resulted in a major SDF victory on August 27, 2016.
\n
On the left, a still from a Coalition video showing an airstrike on a IS ‘fighting position near Manbij, Syria’ allegedly on July 21, 2016. The location shown in the video was indeed geolocated to Manbij at 36.517496, 37.955025.
\n
On the left, a still from a Coalition video showing an airstrike on a IS ‘fighting position near Hasakah, Syria’ allegedly on May 18, 2015. The location shown in the video was geolocated to Abu Tinah, around 30 km from Hasakah, at 36.602831, 40.3928205.
\n
Vehicles are the second-most common target mentioned in the Coalition airstrike videos. One of such vehicles is an IS “technical” in Hīt reportedly on May 30, 2016, geolocated to the outskirts of the town. The date is worth mentioning, as the 2016 Hīt offensive (Operation Desert Lynx) launched by the Iraqi Government claimed the town and its surrounding were already recaptured back in April 2016.
\n
On the left, a still from a Coalition video showing an airstrike on a IS ‘”technical” vehicle near the town of Hit, Iraq’ allegedly on May 30, 2016. The location shown in the video is indeed near Hit, at 33.627078, 42.856765.
\n
Targets related to IEDs and VBIEDs combined also account for a higher number airstrike videos, and such alleged targets in Mosul have been discussed before in the Bellingcat article “The Battle for Mosul: A View from Space before the Operation”.
\n
It is worth mentioning that all of these targets in the Coalition’s videos are “clean” targets like vehicles, factories and fighting positions. It almost looks like video game, just like IS’s propaganda videos of suicide vehicle-borne improvised explosive devices (SVBIEDs). The Coalition’s videos appear only to showcase the precision and efficiency of Coalition bombs and missiles, and rarely show people, let alone victims.
\n
Civilian Casualties
\n
It is important to mention that civilian casualties take place in many of these videos – even the US has admitted to unintentionally killing civilians. Some videos, such as the airstrikes on the University of Mosul in March 2016, may in fact show the airstrikes that caused significant civilian casualties.
\n
Perhaps the most striking example of a video showing potentially significant civilian casualties came from an airstrike on September 20-21, 2015, targeting an IS “VBIED network” according to the Coalition. The video was deleted after questioning, but archived and re-uploaded by others, including investigative journalist Azmat Khan.
\n\n
But was this really a “VBIED network”? Under the original upload, a commenter starting posting that the houses shown were his family’s residence in Mosul.
\n
“I will NEVER forget my innocent and dear cousins who died in this pointless airstrike. Do you really know who these people were? They were innocent and happy family members of mine.”
\n
Days after the strike, Dr Zareena Grewal, a relative living in the US wrote in the New York Times that four family members had died in the strike. On April 2, 2017 – 588 days later – the Coalition finally admitted that it indeed bombed a family home which they confused for an IS headquarters and VBIED facility.
\n
“The case was brought to our attention by the media and we discovered the oversight, relooked [at] the case based on the information provided by the journalist and family, which confirmed the 2015 assessment,” Colonel Joe Scrocca, Director of Public Affairs for the Coalition told Airwars.
\n
Even though the published strike video actually depicted the killing of a family, it remained – wrongly captioned – on the official Coalition YouTube channel for more than a year.
\n
Airwars lowest estimate of civilian casualties killed by the Coalition is 3,111, as of April 25, 2017. To read more about linking Coalition airstrike videos with IS’ videos, please see the article “The Bombers and the Bombed: ” published by Hate Speech International. The approximate locations of these civilian casualties are mapped by the author for Airwars.
\n
\n\n
The author can be contacted at [email protected]. Parts of this article were based on the author’s dissertation “A Tale of Two Air Wars” which was submitted in 2016 to complete the Conflict, Security & Development Master of Arts programme at King’s College London. The author would like to thank @obretix, Aric Toler, Samuel Oakford, and Chris J. Woods for their feedback.
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nA version of this article was also published on\nAirwars\n.\nThe international Coalition fighting the so-called Islamic State (IS) has published dozens of videos purportedly showing airstrikes against targets of the group in Syria and Iraq through the video-sharing social medium YouTube.\nBellingcat\nhas crowdsourced the geolocation of these videos, and this article serves as an analysis of those geolocated airstrike videos.\nContent\nIntroduction\nBackground Information on ‘The Coalition’\nGeolocation and ‘Strike Regions’\nClaimed Targets: ‘Vehicles’, ‘Factories’ and ‘Fighting Positions’\nMapping the Coalition’s Region Labels\nHighlighted Incidents\nIntroduction\nIn reaction to the rapid territorial gains made by the so-called Islamic State (IS) in the first half of 2014, accompanied by the reported human rights abuses and the fear of further conflict spill-overs, the United States (US) started to intervene militarily against the group in Syria and Iraq. In October of that same year, the US started coordinating the military efforts of a dozen of allied countries under the banner of the Combined Joint Task Force (CJTF; often referred to as ‘the Coalition’) engaging in Operation Inherent Resolve (OIR). Today, May 5, 2017, marks the 1,001 day of the campaign.\nWhile earlier videos were posted by the US Central Command (CENTCOM), the first video depicting an airstrike was uploaded to the\nCoalition’s official YouTube channel\non April 11, 2015. Over 200 airstrike videos followed in the nearly two years since, which Bellingcat has aimed to geolocated by using the collaborative crowdsourcing platform\nCheck\ndeveloped by\nMeedan\n, a company that builds digital tools for global journalism and translation.\nA typical airstrike video uploaded by the Coalition to their YouTube channel. This airstrike video, allegedly on\n“\nthree ISIL buildings near Ramadi, Iraq\n”\nis one of the very few videos that is still accessible via the YouTube channel.\nThe Coalition is composed of several countries, but by far the majority (around 68% as of April 24, 2017) of OIR’s airstrikes have been conducted by the US Air Force, as the following tables by monitoring organisation\nAirwars\nshow. The tables are estimates of US and allied strikes in both Iraq and Syria.\nIt is important to note that Coalition airstrike videos are also disseminated through other channels, such as the Ministries of Defence of Coalition members, including the Brits, the French, the Jordanians, and the Iraqis. Besides, videos have been uploaded by YouTube users seemingly affiliated with US Air Force’s\nF/A-18 Hornets\nand\nRockwell B-1 Lancer\naircraft conducting bombing raids above Syria and Iraq. These airstrike videos were not published previously elsewhere to our knowledge. While we have crowdsourcing projects running for those specific MoD videos as well, they are not included in this analysis.\nThe number of airstrikes by the US and its allies, accurate as of April 24, 2017. Courtesy of Airwars.\nThe estimated number of airstrikes in Iraq. The chart on the left shows the US and allied strikes in Iraq, while the right-hand chart gives a breakdown of allied-only strikes. All data is drawn by Airwars from official military reports of the Coalition (CJTF-OIR), France, the UK, Belgium, Denmark, the Netherlands, Australia, and Canada. Courtesy of Airwars.\nIn recent weeks, official videos showing airstrikes against targets of the so-called Islamic State (IS) in Syria and Iraq have been removed from the public YouTube channel of the Coalition. In a written statement to Bellingcat, Coalition says it has removed the videos from YouTube and\nnow posts higher-quality videos to DVIDS\nfor “greater transparency and increased availability”. However, an initial assessment appears to show that not all videos have been migrated. Coalition officials have also given a different account to Airwars in the past as to why the videos were removed, suggesting their presence on the official YouTube channel no longer matched strategic goals. Airwars is permanently archiving all known Coalition and CENTCOM videos issued since August 2014, to ensure their continied availability.\nGiven the recent removal of these videos from YouTube, Bellingcat presents its analysis of all those geolocated YouTube videos. We will continue to monitor, analyse, and geolocate the videos published to DVIDS. You can join the effort via\nCheck\n‘s Bellingcat page, where you can register an account.\nOverview of Geolocations\nSo far, 67 percent of the airstrikes shown in the Coalition airstrike videos have been successfully geolocated. You can access all our data, which will be updated as soon as there are new geolocations, including from DVIDS HUB,\non Silk\n.\nThe term ‘airstrike’ is somewhat imprecise, but in this analysis we consider ‘one airstrike’ as one or more multiple airstrikes at one location. Geolocation is the determination of the exact locations that are being targeted by airstrikes in the videos. Most of the videos have been geolocated by Twitter users\n@obretix\nand\n@miminas999\n, as well as the author of this article.\nBy far most of the strikes shown in videos uploaded to YouTube (as of April 28, 2017) were geolocated to Iraq.\nBy far most of the strikes shown in videos uploaded to YouTube (as of April 28, 2017) were geolocated to Iraq.\nBroken down by provinces, the highest number of airstrikes were geolocated to the Iraqi governorates of Nineveh and Anbar, followed by the Syrian governorate of Aleppo, as of April 28, 2017.\nMost airstrike shown in videos were geolocated to the Iraqi governorates of Nineveh and Anbar, followed by the Syrian governorate of Aleppo, as of April 28, 2017.\nExploring the Coalition’s Location Accuracy\nGenerally, the Coalition gives an indication of a geographical location by labelling the videos “near [location X]”. There are only a handful of videos that do not contain the word “near” but simply mention a location. This analysis considers “near” as being within a 10 km range of the claimed location and is labelled as “accurate” when it falls within that range. Of all geolocated videos, 68 percent was labelled as accurate. The video outside of the 10 km range went up to around 93 km of the claimed location, and for one video no location approximation was given.\nMany of the videos with a significant distance from the claimed location are oil-related facilities that are indeed ‘near’ Deir ez-Zor or Al-Bukamal, such as an oil separation facility at the Al-Ahmar oil field. With the locations being in the desert with no other locations near, these location claims can still be considerate as relatively accurate.\nHowever, there are other incidents that appear to be less concisely located. Perhaps the most interesting incident of all the geolocated videos was a strike on an IS “concealed tactical vehicle” that was claimed to have been conducted on March 23, 2015, which was labelled as “near Al Hawl”, a town in north-eastern Syria. However, the targeted building has been successfully geolocated to a building in Jayar Ghalfas, a town in northwest Iraq.\nA screenshot from a Coalition video claiming to show an airstrike on an IS ‘concealed tactical vehicle’ near Al-Hawl, a town in eastern Syria. The building was geolocated to Jayar Ghalfas, a town in northwest Iraq, as the Microsoft Bing satellite imagery (36.137411, 41.297414) on the right shows. The location is around 30 km southwest of Al-Hawl.\nThough this video was labelled as being in a different country than where it actually took place, it is still relatively near Al-Hawl — around 30 km away.\nWhen Col. Steve Warren, at the time the Coalition’s spokesperson,\ngave an “Ask Me Anything”\non the social media and news aggregation website Reddit, the author asked him about this particular incident. Col. Warren replied that this was “an administrative error that it’s listed as Syria rather than Iraq”, explaining that Al-Hawl in Syria “was the nearest identifiable city to the strike.”\nThe reply by Col. Warren, the Coalition’s spokesperson on the question why it was labelled near to a town in Syria but showed a location in Iraq.\nMore recently, the US erred in its labelling once more, when\na controversial strike on a group of individuals gathered in a mosque in Al-Jinah\n, Syria, was initially labelled as being in the Idlib governorate. While close, the building was actually in the nearby Aleppo governorate.\nThis strike was not an official Coalition attack – and was instead the US unilaterally targeting alleged al-Qaida fighters. The US carries out nearly all of the alliance’s anti-IS bombings in Syria, and military assets can be used for both campaigns.\nWhen asked for clarification about this incident, a CENTCOM spokesperson told Bellingcat that they “don’t mean to cause any confusion. Different internal reports may have listed this differently.”\nEither way, the Coalition thus seems to use a limited number of labels for their targeted location areas. The “Al Hawl, Syria” label was probably closer than their nearest other Iraqi location label, “Sinjar”, around 50 km northeast of Jayar Ghalfas.\nThe Coalition’s ‘Region Labels’\nWhich region does the Coalition use to label one airstrike as “near Mosul” but the other one as “near Al Hawl”? To get a better insight in which regions are used by the Coalition, the geolocator @obretix mapped all geolocated airstrike videos and then used a Voronoi diagram which is a partitioning of a plane into regions based on distance to points in a specific subset of the plane. In this case, the points are thus all “near” locations mentioned by the Coalition. The geolocations are then corresponding to a region that is closer than any other point.\nAs the following image shows, Al-Hawl is indeed the closest location to the target struck by the Coalition (circled in red) in the number of areas the Coalition has used.\nAn excerpt from a Voronoi diagram of an impression of the regions used by the Coalition, based on the geolocated videos. The geolocated airstrike of March 23, 2016 that was labelled near ‘Al Hawl’ is circled in red. Map by ‘obretix’.\nThere are more interesting insights the maps reveals. While the Coalition does use a label for “near Kubaysah”, a city in Iraq, some strikes within the city’s perimeters were labelled as “near Hit” — a city nearby but still less accurate than using a label there was for that city.\nA detailed view on the Kubaysah/Hit region on the Voronoi diagram map, showing that two videos labelled as “near Hit” were in fact closer to Kubaysah, which also has its own ‘location label’.\nAnother example of remarkable region labels is the use of hamlets, such as “near Washiya” and “near Sultan Abdullah” – places with only a few houses, but close to respectively Aleppo and Mosul. It is worth mentioning that “near Aleppo” is not used in any of the YouTube videos, while “near Washiya” has been twice for a target only a double dozen kilometres away from Aleppo city. Why would these small hamlets be used as a region label, while the case of Jayar Ghalfas could not get its own region label? Is this intentional? This is a question that remains unanswered.\nWhat is clear is that it the publicly provided locations by the Coalition for its airstrikes in Iraq and Syria may be off as much as 90+ kilometres, making it more difficult for outside monitors to verify and report on the civilian death toll of the air war.\nIn its Transparency Audit of the Coalition, published in December 2016, Airwars noted problems with the public reporting process. Difficult to navigate internal logs “in turn led to quite vague military reporting”. Locations, then, could only be taken as approximate. One CENTCOM senior official explained the situation in some detail:\n“When the aircrew come back [from a strike mission], as you drill into a geographic location, some of those areas have towns that consist of three or four people. So typically what’s going to be in the strike log is going to be the largest city nearby. And they’ll annotate, ‘Conducted a strike near Mosul.’ In fact it’s going to be some small town that’s 23 clicks [kilometres] outside of Mosul. If they put that on the strike log, once it goes through the ‘Enterprise’ [slang for the Combined Air Operations Centre] no one knows where that is. Officials were keen to stress that if an incident was being investigated, “we do have the ability to go back and drill down into the detail.”\nOverall, all location claims were all within a 100 km distance of the claimed location, and all of these claims were relatively accurate as to the location it referred to — unlike the Russian airstrike videos, which were\nin some cases massively inaccurate\n.\nClaimed Targets: ‘Vehicles’, ‘Factories’ and ‘Fighting Positions’\nWith regards to the targets, the most used targets mentioned can be categorised as “fighting positions”, followed by vehicles and facilities for improvised explosive devices (IEDs) or suicide vehicle borne IEDs (SVBIEDs).\nOn the left, a still from a Coalition video showing an airstrike on a IS ‘homemade explosives cache near Bayji, Iraq’ allegedly on October 14, 2015. The location shown in the video, which was uploaded two days later, is around 10 km away from Bayji at\n35.052918, 43.552203\n.\nMost of the targets are classified as ‘fighting positions’ by the Coalition. These are usually buildings that are, reportedly, used by IS militants as positions in battle. An example can be seen in the following airstrikes on buildings in Manbij, reportedly on July 21 and 24, 2016, and geolocated by @obretix to\n36.522428, 37.962093\nand\n36.517496, 37.955025\n, respectively. The strikes were in support of the Syrian Democratic Forces (SDF) during the Manbij offensive, codenamed Operation Marty and Commander Faysal Abu Layla by the group. The offensive lasted almost three months and resulted in a major SDF victory on August 27, 2016.\nOn the left, a still from a Coalition video showing an airstrike on a IS ‘fighting position near Manbij, Syria’ allegedly on July 21, 2016. The location shown in the video was indeed geolocated to Manbij at\n36.517496, 37.955025\n.\nOn the left, a still from a Coalition video showing an airstrike on a IS ‘fighting position near Hasakah, Syria’ allegedly on May 18, 2015. The location shown in the video was geolocated to Abu Tinah, around 30 km from Hasakah, at\n36.602831, 40.3928205\n.\nVehicles are the second-most common target mentioned in the Coalition airstrike videos. One of such vehicles is an IS “technical” in Hīt reportedly on May 30, 2016,\ngeolocated to the outskirts of the town\n. The date is worth mentioning, as the 2016 Hīt offensive (Operation Desert Lynx) launched by the Iraqi Government claimed the town and its surrounding were already recaptured back in April 2016.\nOn the left, a still from a Coalition video showing an airstrike on a IS ‘”technical” vehicle near the town of Hit, Iraq’ allegedly on May 30, 2016. The location shown in the video is indeed near Hit, at\n33.627078, 42.856765\n.\nTargets related to IEDs and VBIEDs combined also account for a higher number airstrike videos, and such alleged targets in Mosul have been discussed before in the Bellingcat article\n“\nThe Battle for Mosul: A View from Space before the Operation\n”.\nIt is worth mentioning that all of these targets in the Coalition’s videos are “clean” targets like vehicles, factories and fighting positions. It almost looks like video game, just like\nIS’s propaganda videos of suicide vehicle-borne improvised explosive devices\n(SVBIEDs). The Coalition’s videos appear only to showcase\nthe precision and efficiency\nof Coalition bombs and missiles, and rarely show people, let alone victims.\nCivilian Casualties\nIt is important to mention that civilian casualties take place in many of these videos – even the US has admitted to unintentionally killing civilians. Some videos, such as the airstrikes on the University of Mosul in March 2016, may in fact show the airstrikes that caused significant civilian casualties.\nPerhaps the most striking example of a video showing potentially significant civilian casualties came from an airstrike on September 20-21, 2015, targeting an IS “VBIED network” according to the Coalition. The video was deleted after questioning, but\narchived\nand re-uploaded by others, including investigative journalist Azmat Khan.\nBut was this really a “VBIED network”? Under the original upload, a commenter starting posting that the houses shown were his family’s residence in Mosul.\n“I will NEVER forget my innocent and dear cousins who died in this pointless airstrike. Do you really know who these people were? They were innocent and happy family members of mine.”\nDays after the strike, Dr Zareena Grewal, a relative living in the US\nwrote in the New York Times\nthat four family members had died in the strike. On April 2, 2017 – 588 days later – the Coalition\nfinally admitted\nthat it indeed bombed a family home which they confused for an IS headquarters and VBIED facility.\n“The case was brought to our attention by the media and we discovered the oversight, relooked [at] the case based on the information provided by the journalist and family, which confirmed the 2015 assessment,” Colonel Joe Scrocca, Director of Public Affairs for the Coalition told\nAirwars\n.\nEven though the published strike video actually depicted the killing of a family, it remained – wrongly captioned – on the official Coalition YouTube channel for more than a year.\nAirwars lowest estimate of civilian casualties killed by the Coalition is 3,111, as of April 25, 2017. To read more about linking Coalition airstrike videos with IS’ videos, please see the article “\nThe Bombers and the Bombed:\n” published by\nHate Speech International\n. The approximate locations of these civilian casualties are\nmapped\nby the author for Airwars.\nThe author can be contacted at\n[email protected]\n. Parts of this article were based on the author’s dissertation “A Tale of Two Air Wars” which was submitted in 2016 to complete the Conflict, Security & Development Master of Arts programme at King’s College London. The author would like to thank @obretix, Aric Toler, Samuel Oakford, and Chris J. Woods for their feedback.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/YT-interface-1-1200x884.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/number-of-airstrikes-per-1.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/estimate-1-1200x584.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/05/locations-1200x554.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/geolocated-airstrkes-per-country.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/closest-location-governorate.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/checkcoalition-figure-3-1-1200x426.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/checkcoalition-figure-4-1-1200x141.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/checkcoalition-figure-5-1.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/checkcoalition-figure-6-1.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/example-1.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/manbij-exmaple.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/fighting-position-example-2.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/vehicle-example.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/04/Untitled-1.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2017-09-04_geolocating-libyas-social-media-executioner.json b/bellingcat_geolocation_json/2017-09-04_geolocating-libyas-social-media-executioner.json
new file mode 100644
index 0000000000000000000000000000000000000000..cf73148e1b29937e289955e630593a3c99cdb98f
--- /dev/null
+++ b/bellingcat_geolocation_json/2017-09-04_geolocating-libyas-social-media-executioner.json
@@ -0,0 +1,92 @@
+{
+ "slug": "geolocating-libyas-social-media-executioner",
+ "title": "Geolocating Libya's Social Media Executioner",
+ "url": "https://www.bellingcat.com/news/middle-east/2017/09/04/geolocating-libyas-social-media-executioner/",
+ "date": "2017-09-04T13:40:30+00:00",
+ "modified": "2017-10-03T15:06:36+00:00",
+ "author": "Christiaan Triebert",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)Warning: this article contains graphic material. Update: Incident One, Four and Five have been geolocated. In Short: The International Criminal Court issued its first ever arrest warrant solely based on social media evidence; This warrant is a significant development and aligns the Court with the realities of today’s conflict zones; All seven videos […]",
+ "content_html": "
Warning: this article contains graphic material. Update: Incident One, Four and Five have been geolocated.
\n
In Short:
\n
\n
The International Criminal Court issued its first ever arrest warrant solely based on social media evidence;
\n
This warrant is a significant development and aligns the Court with the realities of today’s conflict zones;
\n
All seven videos on which Libyan commander Werfalli is charged with murder as a war crime can be found online;
\n
With the help of the verification platform Check, Bellingcat is crowdsourcing the geolocation of the locations shown in these videos. Three out of seven videos have been geolocated so far.
\n
Open source information is a potential goldmine of evidence, though there are potential challenges with using this information in a legal setting, which will be discussed in relation to the Werfalli case.
\n
\n
Introduction
\n
On August 15, 2017, the International Criminal Court (ICC) issued its first ever arrest warrant solely based on evidence collected on social media.
\n
The arrest warrant accuses Mahmoud Mustafa Busayf Al-Werfalli, an alleged commander of the Al-Saiqa Brigade in Libya, of mass executions in or near Benghazi.
\n
The use of social media evidence is a significant development, as it aligns the ICC with the realities of many of today’s conflict zones. It also allows anyone with an Internet connection to access the evidence themselves, as it is all open source information – the kind of information we always base our investigations on at Bellingcat. For that reason, this article examines the social media evidence used in the case in further depth.
\n
Firstly, a contextual background will be given to Werfalli and where the alleged incidents took place related to so-called “Operation Dignity” in and around the Libyan city of Benghazi.
\n
Secondly, the seven incidents based on seven separate videos will be examined in further detail. We have tried to extract as much information as possible out of the videos, and asking our readers to help geolocate the execution site. (Update: So far, already two videos have been geolocated.)
\n
Thirdly, the social media evidence of this specific arrest warrant will be discussed in the light of the promise and peril of open source evidence. There is no doubt that open source information is a huge potential source for legal evidence, but what are the methodological challenges in assessing and eventually using these materials?
\n
Contextual Background: Operation Dignity
\n
Since a bloody uprising started against the government of then-President Muammar Qaddafi in 2011, Libya has been in a state of unabated violence and conflict. After the relatively quick fall of Qaddafi, the country has been divided into rival political and armed groups. Opposing government and parliaments roughly divide Libya in an eastern and western side.
\n
Libya is roughly divided into an eastern and western section, as this LiveUAmap of approximate territorial control as of August 31, 2017, shows. The LNA-controlled areas are in red, the Libyan Dawn-controlled areas in blue, and the pink area represents the territory held by Tuareg militias. The black strip at Benghazi are jihadist groups. This is an oversimplified map of territorial control, as it does not show Amazigh (Tuareg) or Toubou controlled areas.
\n
The Government of National Accord (GNA) is seated in the western part of Libya. As a rival of the eastern-based House of Representatives, the GNA has the armed support of several groups, most notably the Libya Dawn group – a coalition of pro-Islamist militias which attacked Tripoli International Airport in the summer of 2014 and has controlled many of Libya’s coastal cities since. The GNA are the internationally recognised authorities of Libya, and has received outspoken support from a number of Western countries, including the United States (US), the United Kingdom, and France.
\n
The eastern part of Libya, and a small area west of Tripoli, are controlled by a coalition of armed groups named the Libyan National Army (LNA). This force of former army units, ex-revolutionary groups and tribal coalitions is under the command of General Khalifa Haftar, a dual citizen of Libya and the US. The eastern-based parliament is known as the House of Representatives.
\n
Parts of Benghazi, Libya’s second-most populous city on the eastern Mediterranean coast, have been under control of a coalition of Islamist groups since the 2011 revolution. This coalition of Islamist groups became known as the Benghazi Revolutionary Shura Council (BRSC), which includes groups like Ansar Al-Sharia, the February 17 Revolutionary Martyrs’ Brigade, the Rafallah Al-Sahati militia, and the eastern Libya Shield Brigade.
\n
In May 2014, General Haftar launched “Operation Dignity” against the BRSC in and around Benghazi. The operation continued until at least March 18, 2017. One of the groups taking part in the anti-BRSC offensive were the Libyan Special Forces, colloquially known as the Al-Saiqa Brigade.
\n
The Saiqa Brigade was originally part of Qaddafi’s army, but defected in an early stage of the 2011 uprising. Being active especially in western Benghazi in recent years, the group reportedly consists of around 5,000 soldiers. Werfalli, the subject of the ICC arrest warrant, is said to be a Saiqa Brigade commander since at least 2015.
\n
During the periodic fighting in and around Benghazi, Werfalli is said to have been directly responsible for the execution of 33 persons who were either civilians or persons hors de combat. “There is no information in the evidence to show that they have been afforded a trial by a legitimate court, whether military or otherwise”, the arrest warrant states. There was thus an alleged denial of fair trial for the killed individuals – making them unlawful.
\n
Werfalli announced his resignation on May 15 in a video published on YouTube, though the reason why is unclear. There has been chatter that it may have to do with an attack on a police station in Benghazi by the Saiqa Brigade that left one officer dead.
\n
On May 15, 2017, Werfalli announced his resignation. The reason as to why the Saiqa Brigae commander resigned unclear.
\n
Following the ICC arrest warrant, the LNA arrested Werfalli, stating he was under investigation by a military prosecutor. The LNA said that it is ready “to cooperate with [the ICC] in informing you of the result and course of the judicial case”. The LNA statement did not mention whether it would hand over Werfalli to the ICC.
\n
However, a spokesperson for the Saiqa Brigades dismissed the ICC arrest warrant, saying that the ICC should rather arrest the opponents of the armed group, Al Jazeera reported.
\n
The Footage: Seven Incidents in Seven Videos
\n
Al-Werfalli is charged with murder as a war crime under Article 8(2)(c)(i) of the Rome Statute, the treaty that established the ICC, which is the “Violence to life and person, in particular murder of all kinds, mutilation, cruel treatment and torture”.
\n
The charge is based on seven incidents shown in seven separate social media videos. All of these videos have been found online. Each incident and video will be presented and discussed below, following the order of the arrest warrant.
Mr Al-Werfalli, wearing camouflage trousers and carrying a weapon, is seen in a video footage to stand near a hooded, unidentified person, who is moving around an open dirt area with his arms in the air. Mr Al-Werfalli is heard saying “Put your hands up! Put your hands up! Put your hands up!”. After that, he shoots the hooded person with his left hand a number of times and the person falls on the ground. Mr Al-Werfalli approaches the body, shoots again the body on the ground for a number of times and states “You have been misled by he who did you harm. You have been misled by Satan”. The video depicting this incident was posted on Facebook on 3 June 2016.
\n
\n
The first video is still available online on Facebook, with a duration of 2:16. It has been watched almost 200,000 times as of writing this article, and shared over 5,000 times. The logo and the name of the Saiqa Brigades (Arabic: القوات الخاصة الليبية) is visible throughout the video.
\n
A still of the video showing Werfalli, wearing camouflage trousers and carrying what appears to be a Russian-made PK machine gun, moments before he shoots an unidentified person.
\n
The source is a Facebook profile under the name Mohamed Al-Ghali with Facebook identification number 100012315483959 and username hamoalgali01. It is unclear whether this profile is the original source of the video. [Update: Al-Ghali has reached out to Bellingcat to specify that he was not the original source of the video]
\n
The accompanying text with the video claims to show the execution of a Syrian national who is a member of the so-called Islamic State (IS). In the comments section, the user uploaded a photo seemingly of the executed individual at the same scene.
\n
\n
The upload time of the video to Facebook is June 4, 2016, at 1:12 am local time in Benghazi. This means there is an inconsistency with the arrest warrant’s claim that it was uploaded on June 3, though this may have to do with the time Facebook displays as your local time zone based on your computer time. If you figure out what time settings your computer is in, there are plenty of useful online tools to convert different time zones.
\n
\n
It is unclear when this video was recorded, though it must have been before June 4, 2017.
\n
The exact location of this execution site is at 32.014221, 20.113838 (Google Maps, Wikimapia), an orchard in between the Benghazi suburbs of Al-Hawari and Al-Qawarishah.
\n
A stitched panorama of still from the execution video gives a wider perspective of the area, which is useful for the geolocation process. Visual clues are the buildings on the left and right edge of the photos, the orchard, the two lines of trees next to what appears to be a narrow road, and an electricity pylon in the background.
\n
The video was geolocated by Bellingcat’s “Daniel”, who first made a map of what was visible in the video.
\n
\n
Bellingcat’s Timmi Allen created the following image showing some of the elements of the video that match the satellite imagery of the area.
Mr Al-Werfalli, wearing camouflage trousers and a black t-shirt with the logo of the Al-Saiqa Brigade, and carrying a weapon, is seen in a video footage shooting with his left hand three male figures in the head. The three persons are kneeling in front of a wall with their arms tied behind their backs. After the bodies of the three men fall, Mr Al-Werfalli shoots at the bodies a few more times. Two of the victims are alleged to have been revolutionaries in Benghazi. The video depicting this incident was posted on social media on 20 March 2017.
\n
\n
The second video is available via YouTube as a 18-second clip, uploaded by a user named “Hassan Mahmoud” (Arabic: حسن محمود). It has been watched over 10,000 times at the time of writing this article.
\n
Using Amnesty International’s YouTube Data Viewer, the exact upload time to YouTube can established at March 21, 2017, at 4:09 am local time in Benghazi. The date is inconsistent with the time mentioned by the ICC warrant, though this may have to do with the display date of YouTube and the actual uploading time (hence the importance of Amnesty’s tool).
\n
It is unclear when this video was recorded.
\n
The exact location of this street has not been found yet, but you can help trying to find the exact location. Hints and possible locations can be posted on Check, where there is a specific entry for the geolocation of this video.
Mr Al-Werfalli, wearing camouflage trousers and a black t-shirt with the logo of the Al-Saiqa Brigade, and carrying a weapon, is seen in a video footage in a room, with others being present as well. He is standing next to an unidentified man in a white t-shirt, who is kneeling bare feet, with his hands behind his head. Mr Al-Werfalli shoots the person with his left hand in the head and continues to do so after the man falls down. The men around Mr Al-Werfalli cheer in approval and another unidentified man from the group comes forward and shoots the victim two more times. The videos depicting this incident were posted on social media on 7 and 8 May 2017.
\n
\n
The third video is still available online on Facebook, and its duration is 1:17. It has been watched over 32,000 times and shared 242 times at the time of writing.
\n
The source is a Facebook profile going by name Bin Mansour with Facebook identification number 100012980119097. It is unclear whether this profile is the original uploader of the video.
\n
The accompanying text with the video claims to show the execution of “one of the Kharijites” by Werfalli.
\n
The upload time of the video to Facebook is May 7, 2017, at 5:12 pm local time in Benghazi, which is consistent with the time mentioned in the ICC warrant.
\n
It is unclear when the video was recorded.
\n
The exact location of this house has not been found yet, and may be close to impossible without insider knowledge. You can help trying to find the exact location. Hints and possible locations can be posted on Check.
Mr Al-Werfalli is seen in a video footage in an outdoor setting together with two other men who carry weapons and whose heads are covered. The three men stand behind two other men, who are dressed in camouflage and are kneeling bare feet on the ground. The two men are claimed to be, the first, a member of the BRSC, and the second, possibly a member of Ansar al-Sharia. The two men can be seen earlier in the video footage being detained in a cage. Mr Al-Werfalli is seen walking away from the kneeling men, while the other two men with firearms walk towards them. Mr Al-Werfalli has his left hand in the air and sweeps it down towards the ground in a manner that suggests that he is ordering them to proceed with the execution. The two men shoot the kneeling persons who fall on the ground. The video depicting this incident was posted on social media on or about 22 May 2017.
\n
\n
The fourth video is also still available online on Facebook, with a duration of 2:47. It has been watched almost 37,000 times and shared 340 times as of writing this article. Only the last 9 seconds of the video show the execution, and the rest of the video shows interviews with the two individuals who are claimed to be members of the BRSC and Ansar Al-Sharia.
\n
The source is the same Facebook profile as the one that posted the video showing Incident One. It is unclear whether this profile is the original source of the video.
\n
The accompanying text with the video claims to show “the execution of the death penalty” (Arabic: تنفيذ حكم الإعدام) in Benghazi of a BRSC member and an Ansar Al-Sharia member.
\n
The upload time of the video to Facebook is May 22, 2017, at 11:27 am local time in Benghazi. This date is consistent with the date mentioned by the ICC.
\n
In the fourth video, Werfalli is seen in an outdoor setting with two other men wearing balaclavas. They stand behind two other men who are claimed to be a member of the BRSC and Ansar Al-Sharia. Werfalli orders the two man wearing balaclavas to proceed with the execution of the two other men.
\n
The execution site has been geolocated by Twitter-user @5urpher to a compound in Al-Hawari, a southern suburb of Benghazi. The coordinates are 32.046755, 20.116750 (Google Maps, Wikimapia).
\n
A satellite image of Google Earth of the compound in southern Benghazi where the execution shown in the Incident Four video took place.
\n
It is unclear when this video was recorded, though once it is geolocated it may be able to determine an approximate time of the day based on the clearly visible shadows.
Mr Al-Werfalli, wearing a full camouflage uniform, is seen in a video footage in an outdoor setting together with five men wearing hoods and camouflage trousers and holding firearms. Those five hooded, armed men stand behind four other hooded men, who are kneeling barefoot on the ground with their hands behind their backs. They point their firearms at each kneeling person’s head. Mr Al-Werfalli raises his left hand in the air and sweeps it down towards the ground in a manner that suggests that he is ordering them to proceed with the execution. The five men shoot the four kneeling persons, who fall on the ground. The video depicting this incident was posted on social media on 9 June 2017.
\n
\n
The fifth video can be found online on Facebook, uploaded by someone using the username “aalathry1” with Facebook identification number 1797850107. The execution takes place in the last 47 seconds of the video.
\n
\n
The video was uploaded on June 9, 2017 at 2:29 pm local time in Benghazi. This date is consistent with the ICC warrant.
\n
The exact date when this execution took place is unknown.
\n
The exact location of this execution site has been geolocated by Twitter-user @5urpher to a military compound between Benghazi and the airport. The coordinates are 32.122988, 20.164492. The location was found by using the lines of electricity of OpenStreetMap in Benghazi.
\n
The location of Incident Five (yellow pin) in Google Earth, with an overlay of electricity pylons (black squares).
Mr Al-Werfalli is seen in a video footage in an area of desert together with two other men who are holding firearms. They stand behind two other persons, who are kneeling on the ground with their hands behind their back. Mr Al-Werfalli is seen speaking into the camera and then raising his left hand in the air and sweeping it down towards the ground in a manner that suggests that he is ordering the two men to proceed with the execution. The men shoot the persons kneeling, who fall on the ground. The video depicting this incident was posted on social media on 19 June 2017.
\n
\n
The sixth video is available on both Facebook and YouTube with a duration of 59 seconds. The video has been watched 80,000 times on Facebook and 18,000 times on YouTube.
\n
The Facebook source is a page – not an account – with username “Benghazi2”.
\n
The YouTube source is a channel named “Libya Al-Mostakbal”, which has over 1,000 subscribers. The accompanying text states that the video was published on social media, suggesting Libya Al-Mostakbal reuploaded the video and is not the original source. It says that Werfalli is shown in the video, ordering the execution of two men on an unknown date at an unknown place.
\n
\n
The upload time of the video to Facebook is June 19, 2017, at 9:19 pm, which is consistent with the time mentioned in the ICC arrest warrant.
\n
It is unclear when this video was recorded.
\n
The exact location of this desert area has not been found yet, but you can help trying to find the exact location. Hints and possible locations can be posted on Check, where there is a specific entry for the geolocation of this video.
Mr Al-Werfalli, wearing a black cap, camouflage trousers and a black t-shirt bearing the logo of the Al-Saiqa Brigade, is seen in a video footage, together with two other persons wearing camouflage trousers, black t-shirts, and holding firearms. Mr Al-Werfalli is holding a white document from which he reads. He refers to the document as a “Decree decision” which “shall be executed today 17/07/2017”. The video also depicts 18 persons wearing orange jumpsuits and black hoods, with their hands tied behind their backs, kneeling bare feet on the ground, in four lines. After reading the document, Mr Al-Werfalli says “Attention! The first group!” and then “Ready! Aim! Fire!” As he does so, five men in camouflage uniform, hooded, and carrying firearms, walk from behind the group of kneeling persons to the front and position themselves behind the first line of five kneeling persons. They raise their firearms at them and shoot. The persons fall on the ground. One person in the second line also falls. The shooters then return behind the group of the other kneeling persons.\r\n\r\nThen Mr Al-Werfalli is recorded saying “The second group! Ready! Aim! Fire!” As he does so, five men walk again from behind the group and position themselves behind the second line of five kneeling persons. The person who has collapsed during the first shooting is pulled back to his knees by one of the men. The five men then raise their firearms and shoot the five kneeling persons, who fall on the ground. The shooters then return behind the group of the other kneeling persons.\r\n\r\nWhen Mr Al-Werfalli says “Attention! Attention, the third group!” and “Ready! Aim! Fire!” five men walk again from behind the group, position themselves behind the third line of five kneeling persons, raise their weapons and shoot them. The persons fall on the ground and the shooters return behind the group of the remaining kneeling persons.\r\n\r\nMr Al-Werfalli and the two other persons wearing camouflage trousers, black t-shirts, and holding firearms then walk behind the final line of three remaining kneeling persons. Mr Al-Werfalli is seen to still hold the white document from which he has read earlier. Mr Al-Werfalli and the other two men raise their firearms and shoot the three kneeling persons, who fall on the ground. He and one of the other men fire some additional shots towards the persons on the ground before walking away.\r\n\r\nIn the same video, in a new segment, Mr Al-Werfalli, wearing a black cap, camouflage trousers and a black t-shirt bearing the Al-Saiqa Brigade logo, is seen together with two men wearing orange jumpsuits and black hoods kneeling on the ground. One of the two kneeling men has his hands in front of him while the other has his hands on the back of his head. Mr Al-Werfalli is holding a white document folded in his hand. He is recorded saying “Ready! Aim! Fire!” As he does so, two hooded men in camouflage, carrying firearms, walk forward, raise their weapons and shoot at the two kneeling persons, who fall on the ground. The men fire some additional shots at the victims before walking away.\r\n\r\nThe video depicting this incident, involving in total 20 executed persons, was posted on social media on 23 July 2017. The data indicates that the video may have been uploaded in the Benghazi area, in Libya.
\n
\n
The seventh and last video is also still available online. A 4:46-long video was posted to Facebook by the same account that posted the Incident One and Incident Four videos. Segments of that video have been used by several other YouTube accounts from media organisations. It is unclear whether this profile is the original source of the video(s).
\n
The accompanying text with the video claims to show the execution of twenty “terrorists” sentenced to death because they were allegedly involved in kidnapping, torture, killing, bombing, and slaughtering.
\n
The upload time of the video to Facebook is July 23, 2017, at 5:14pm local time in Benghazi, which is consistent with the time mentioned in the arrest warrant.
\n
It likely that the execution took place on July 17, 2017. The accompanying text with the video claims it was filmed in Benghazi.
\n
\n
The exact location of this location has not been found yet, but you can help trying to find the exact location. Hints and possible locations can be posted on Check, where there is a specific entry for the geolocation of this video. A stitched panorama shown above may be of help. The buildings in the upper side of the panorama are a potential clue to the location.
\n
The Promise and Perils of Open Source Evidence
\n
The Promise: A Goldmine of Potential Evidence
\n
The widespread proliferation of new communication technologies, such as smartphones connected to the Internet, has increased the already unprecedented amount of information available about many of today’s conflicts.
\n
Whether it is an onlooker who recorded an airstrike on his or her mobile phone, or a militant group that filmed an offensive for propaganda purposes, the material is very often published on the Internet and, specifically, social media. It may be that possible human rights violations are recorded in that footage.
\n
Once corroborated and triangulated, this open source information can be a potential mine of evidence. Some jurisdictions have already used it to convict a suspect. A Swedish district court, for example, convicted a Syrian national of shooting a person with an assault rifle in 2012, partly basing the judgement on a video that was posted to Facebook. The man was sentenced to life in prison.
\n
The international community has limited ability to collect impartial information and freely assist civilians inside many of today’s conflict zones like Myanmar, Syria, and Iraq due to restriction of movement. With the decrease in international observers that are able to conduct field investigations, open source evidence is promising. Open source information enables anyone with an Internet connection to mine information from places that were once blind spots. International human rights courts and tribunals “have expressed hope that new information technologies might bring about better, cheaper, and safer prosecutions,” Keith Hiatt writes in The Yale Law Journal.
\n
Nevertheless, there are also methodological challenges of using open source information in legal cases, such as reliability, availability, and security, which will be discussed in the next section.
\n
The Perils: Reliability, Availability, Security
\n
Some of these challenges have been outlined by academics, including Hiatt in the aforementioned article, as well as by Emma Irving in an Opinio Juris article as a reaction to the ICC arrest warrant.
\n
Firstly, open source information is “notoriously susceptible to problems of verifiability, which naturally affects its reliability in any kind of criminal proceeding” (Irving). Questions that arise may be the establishment of the exact time, date or location of the evidence, or even potentially digitally altered footage.
\n
However, there is a wide variety of tools and methods which can be used to determine the exact location of a video or photograph, and sometimes even the approximate date and time. Verifiability is a challenge that can be overcome in many cases. It is unclear whether the ICC knows the exact location of the execution sites.
\n
With the rapid development of technology and artificial intelligence, a new challenge is on the horizon: Can we be sure a person seen in a video is in fact the same person the video purports to depict? Offenders have been incorrectly identified before, as in the aftermath of the Boston Marathon bombing or the assaults at the Charlottesville white supremacist rally. Even more worrying is the possibility to produce a photorealistic video, say, of former US President Barack Obama speaking sentences he never said.
\n
But neither of these issues are a challenge in the Werfalli case. Most, if not all, of the footage mentioned by the ICC appears to be produced – and shared – by individual members or supporters of the Saiqa Brigade.[1] The more salient question here, if this case goes to trial, is whether the defence team will contest the authenticity of the footage, according to Hiatt. In many international criminal cases, Hiatt says, “the issue is not whether the atrocity occurred, as a matter of actual fact. Instead, it is the context that matters.” For example, was there a fair trial before the execution, or not? That question is unlikely to hinge on whether an execution video was authentic. Indeed, the contested elements of a crime under the Rome Statute are more likely to concern chain of command, or the widespread, systematic nature of a criminal act.
\n
Secondly, the availability of open source information may be another challenge. While the arrest warrant for Werfalli was issued shortly after the alleged crimes, human rights tribunals are usually only established after the cessation of hostilities. “As a result, today’s investigations concern yesterday’s atrocities,” as the formation of these tribunals may take years (Hiatt).
\n
This can be a problem, as was highlighted by YouTube’s recent removal of many videos, playlists, accounts, and channels documenting conflict zones. As outlined in a New York Times article, YouTube “inadvertently removed thousands of videos that could be used to document atrocities in Syria, potentially jeopardizing future war crimes prosecutions.” The same has happened with videos from other areas where human rights violations are reportedly happening, such as Myanmar.
\n
For that very reason, it is important to preserve, for the long term, open source documentation relating to possible human rights violations and other crimes. Projects like the Syrian Archive are entirely focused on this, and Bellingcat hopes to broaden these efforts to other conflict areas in the future too. (Also, as you may have noticed, most of the hyperlinks in this article link to archived versions of the original page in case a source is taken offline).
\n
The Syrian Archive is entirely focused to preserve, for the long term, open source documentation relating to possible human rights violations and other crimes.
\n
A third challenge is security. Open source evidence may increase threats for eyewitnesses and those who gathered the information first-hand, or even bystanders visible in a photo or video. As Hiatt said,
\n
“By using these materials, investigators may draw unwanted attention to people depicted in them. The disclosure of a photograph to a powerful defendant or a hostile government might expose the identity of a witness, or otherwise endanger third parties.”
\n
This challenge does not seem of particular relevance to the Werfalli case, as all of the material appears to have been recorded and subsequently disseminated on social media by members or sympathisers of the Saiqa Brigade themselves.
\n
In case eyewitnesses are used, the use of open source evidence may actually decrease the risk for eyewitnesses as fewer witnesses may have to testify, and in case they do their claims can be corroborated with open source information.
\n
The fourth and last challenge is forward-looking, and was put forward by Irving: the weight of evidence. Will the approach to open source evidence change depending on the stage of proceedings?
\n
“The standard of proof for the issuance of an arrest warrant is ‘reasonable grounds to believe’ (Article 58(1) of the Rome Statute). One might assume that the approach to open source evidence would not be stricter when it came to initiating an investigation (‘reasonable basis to believe’, Article 53(1)(a)) but the situation may be different as regards the higher standards of ‘substantial grounds to believe’ (confirmation of charges, Article 61(5)) and ‘beyond reasonable doubt’ (conviction, Article 66(3)).”
\n
Will it be the case that the higher the burden, the lesser the chance that open source evidence can meet it? The decision on the arrest warrant was not the place to answer such questions, but they will need to be addressed in future.
\n
Discussion
\n
Human rights organisations like Amnesty International welcomed the arrest warrant, stating the ICC’s decision is “a significant step towards ending the rampant impunity for war crimes in Libya.” Amnesty International has previously reported on potential war crimes committed by the LNA.
\n
The US, the UK, and France welcome the ICC arrest warrant in a joint statement. This statement can be read in two possible ways, Alex Whiting writes for Just Security.
\n
The first is that “the U.S. (a non-State Party of the ICC) is joining two ICC State Parties [France and the UK] to support and encourage a positive and pro-active response to the ICC warrant [..meaning..] support for the Court consistent with U.S. law […] as long as it is in the national interest.”
\n
The second and more plausible read of the joint statement is, according to Whiting, that the US, France, and the UK, “are looking for a way to avoid the ICC. If the LNA did in fact pursue a proper investigation and prosecution of Al-Werfalli for the crimes for which he is charged, an unlikely proposition to be sure, it could render him ineligible to be prosecuted at all by the ICC.”
\n
The Werfalli case has also potential implications for the US and other states engaged in Libya, since General Haftar is a dual US-Libyan citizen, Whiting notes. For more information on this issue, Whiting’s article is recommended reading.
\n
On Libyan social media, criticism has been raised about the ICC’s arrest warrant. Questions are asked publicly about why “the heroic” Werfalli – who killed alleged Islamic State militants – would be prosecuted. Why are his enemies not prosecuted? Or, for instance, why not culprits of the forced displacement and killing of the Rohingya in Myanmar, or occupation forces in the West Bank?
\n
Crimes committed in these areas may have been recorded and, for example, have been uploaded to YouTube. If these videos are deleted by YouTube’s AI, the platform risks losing “the richest source of information about human rights violations in closed societies” Hiatt told the New York Times. The ICC faced similar criticism from many top African officials, who accused the court of solely targeting their continent.
\n
Either way, the decision of the ICC to rely on social media to support an arrest warrant is an important development, and may be only the beginning of a much larger increase of open source evidence by the Court. Open source information is a promise too important to ignore.
\n\n
\n[1]However, there have been cases in which armed groups were imitated by others, such as a video in which militants from the Ukrainian Azov Battalion threatened the Netherlands. That video turned out to be fake.
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nWarning: this article contains graphic material. Update: Incident One, Four and Five have been geolocated.\nIn Short:\nThe International Criminal Court issued its first ever arrest warrant solely based on social media evidence;\nThis warrant is a significant development and aligns the Court with the realities of today’s conflict zones;\nAll seven videos on which Libyan commander Werfalli is charged with murder as a war crime can be found online;\nWith the help of the verification platform\nCheck\n, Bellingcat is crowdsourcing the geolocation of the locations shown in these videos. Three out of seven videos have been geolocated so far.\nOpen source information is a potential goldmine of evidence, though there are potential challenges with using this information in a legal setting, which will be discussed in relation to the Werfalli case.\nIntroduction\nOn August 15, 2017, the International Criminal Court (ICC) issued its first ever arrest warrant solely based on evidence collected on social media.\nThe\narrest warrant\naccuses Mahmoud Mustafa Busayf Al-Werfalli, an alleged commander of the Al-Saiqa Brigade in Libya, of mass executions in or near Benghazi.\nThe use of social media evidence is a significant development, as it aligns the ICC with the realities of many of today’s conflict zones. It also allows anyone with an Internet connection to access the evidence themselves, as it is all open source information – the kind of information we always base our investigations on at Bellingcat. For that reason, this article examines the social media evidence used in the case in further depth.\nFirstly, a contextual background will be given to Werfalli and where the alleged incidents took place related to so-called “Operation Dignity” in and around the Libyan city of Benghazi.\nSecondly, the seven incidents based on seven separate videos will be examined in further detail. We have tried to extract as much information as possible out of the videos, and asking our readers to help geolocate the execution site. (Update: So far, already two videos have been geolocated.)\nThirdly, the social media evidence of this specific arrest warrant will be discussed in the light of the promise and peril of open source evidence. There is no doubt that open source information is a huge potential source for legal evidence, but what are the methodological challenges in assessing and eventually using these materials?\nContextual Background: Operation Dignity\nSince a bloody uprising started against the government of then-President Muammar Qaddafi in 2011, Libya has been in a state of unabated violence and conflict. After the relatively quick fall of Qaddafi, the country has been divided into rival political and armed groups. Opposing government and parliaments roughly divide Libya in an eastern and western side.\nLibya is roughly divided into an eastern and western section, as this LiveUAmap of approximate territorial control as of August 31, 2017, shows. The LNA-controlled areas are in red, the Libyan Dawn-controlled areas in blue, and the pink area represents the territory held by Tuareg militias. The black strip at Benghazi are jihadist groups. This is an oversimplified map of territorial control, as it does not show Amazigh (Tuareg) or Toubou controlled areas.\nThe Government of National Accord (GNA) is seated in the western part of Libya. As a rival of the eastern-based House of Representatives, the GNA has the armed support of several groups, most notably the Libya Dawn group – a coalition of pro-Islamist militias which attacked Tripoli International Airport in the summer of 2014 and has controlled many of Libya’s coastal cities since. The GNA are the internationally recognised authorities of Libya, and has received\noutspoken support\nfrom a number of Western countries, including the United States (US), the United Kingdom, and France.\nThe eastern part of Libya, and a small area west of Tripoli, are controlled by a coalition of armed groups named the Libyan National Army (LNA). This force of former army units, ex-revolutionary groups and tribal coalitions is under the command of General Khalifa Haftar, a dual citizen of Libya and the US. The eastern-based parliament is known as the House of Representatives.\nParts of Benghazi, Libya’s second-most populous city on the eastern Mediterranean coast, have been under control of a coalition of Islamist groups since the 2011 revolution. This coalition of Islamist groups became known as the Benghazi Revolutionary Shura Council (BRSC), which includes groups like Ansar Al-Sharia, the February 17 Revolutionary Martyrs’ Brigade, the Rafallah Al-Sahati militia, and the eastern Libya Shield Brigade.\nIn May 2014, General Haftar launched “Operation Dignity” against the BRSC in and around Benghazi. The operation continued until at least March 18, 2017. One of the groups taking part in the anti-BRSC offensive were the Libyan Special Forces, colloquially known as the Al-Saiqa Brigade.\nThe Saiqa Brigade was originally part of Qaddafi’s army, but defected in an early stage of the 2011 uprising. Being active especially in western Benghazi in recent years, the group reportedly consists of around 5,000 soldiers. Werfalli, the subject of the ICC arrest warrant, is said to be a Saiqa Brigade commander since at least 2015.\nDuring the periodic fighting in and around Benghazi, Werfalli is said to have been directly responsible for the execution of 33 persons who were either civilians or persons\nhors de combat\n. “There is no information in the evidence to show that they have been afforded a trial by a legitimate court, whether military or otherwise”, the arrest warrant states. There was thus an alleged\ndenial of fair trial\nfor the killed individuals – making them unlawful.\nWerfalli announced his resignation on May 15 in a video published on\nYouTube\n, though the reason why is unclear. There has been chatter that it may have to do with an attack on a police station in Benghazi by the Saiqa Brigade that left one officer dead.\nOn May 15, 2017, Werfalli announced his resignation. The reason as to why the Saiqa Brigae commander resigned unclear.\nFollowing the ICC arrest warrant, the LNA\narrested\nWerfalli, stating he was under investigation by a military prosecutor. The LNA said that it is ready “to cooperate with [the ICC] in informing you of the result and course of the judicial case”. The LNA statement did not mention whether it would hand over Werfalli to the ICC.\nHowever, a spokesperson for the Saiqa Brigades dismissed the ICC arrest warrant, saying that the ICC should rather arrest the opponents of the armed group,\nAl Jazeera\nreported.\nThe Footage: Seven Incidents in Seven Videos\nAl-Werfalli is charged with murder as a war crime under Article 8(2)(c)(i) of the\nRome Statute\n, the treaty that established the ICC, which is the “Violence to life and person, in particular murder of all kinds, mutilation, cruel treatment and torture”.\nThe charge is based on seven incidents shown in seven separate social media videos. All of these videos have been found online. Each incident and video will be presented and discussed below, following the order of the arrest warrant.\nIncident One\nLink to the video on Check\n[geolocated].\nThe ICC arrest warrant states the following:\nMr Al-Werfalli, wearing camouflage trousers and carrying a weapon, is seen in a video footage to stand near a hooded, unidentified person, who is moving around an open dirt area with his arms in the air. Mr Al-Werfalli is heard saying “Put your hands up! Put your hands up! Put your hands up!”. After that, he shoots the hooded person with his left hand a number of times and the person falls on the ground. Mr Al-Werfalli approaches the body, shoots again the body on the ground for a number of times and states “You have been misled by he who did you harm. You have been misled by Satan”. The video depicting this incident was posted on Facebook on 3 June 2016.\nThe first video is still available online on\nFacebook\n, with a duration of 2:16. It has been watched almost 200,000 times as of writing this article, and shared over 5,000 times. The logo and the name of the Saiqa Brigades (Arabic: القوات الخاصة الليبية) is visible throughout the video.\nA still of the video showing Werfalli, wearing camouflage trousers and carrying what appears to be a Russian-made PK machine gun, moments before he shoots an unidentified person.\nThe source is a Facebook profile under the name Mohamed Al-Ghali with Facebook identification number 100012315483959 and username hamoalgali01. It is unclear whether this profile is the original source of the video. [\nUpdate: Al-Ghali has reached out to Bellingcat to specify that he was\nnot\nthe original source of the video]\nThe accompanying text with the video claims to show the execution of a Syrian national who is a member of the so-called Islamic State (IS). In the comments section, the user uploaded a\nphoto\nseemingly of the executed individual at the same scene.\nThe upload time of the video to Facebook is June 4, 2016, at 1:12 am local time in Benghazi. This means there is an inconsistency with the arrest warrant’s claim that it was uploaded on June 3, though this may have to do with\nthe time Facebook displays\nas your local time zone based on your computer time. If you figure out what time settings your computer is in, there are plenty of useful online tools to\nconvert different time zones\n.\nIt is unclear when this video was recorded, though it must have been before June 4, 2017.\nThe exact location of this execution site is at 32.014221, 20.113838 (\nGoogle Maps\n,\nWikimapia\n), an orchard in between the Benghazi suburbs of Al-Hawari and Al-Qawarishah.\nA stitched panorama of still from the execution video gives a wider perspective of the area, which is useful for the geolocation process. Visual clues are the buildings on the left and right edge of the photos, the orchard, the two lines of trees next to what appears to be a narrow road, and an electricity pylon in the background.\nThe video was geolocated by Bellingcat’s “Daniel”, who first made a map of what was visible in the video.\nBellingcat’s Timmi Allen created the following image showing some of the elements of the video that match the satellite imagery of the area.\nIncident Two\nLink to the video on Check\n.\nThe ICC arrest warrant states the following:\nMr Al-Werfalli, wearing camouflage trousers and a black t-shirt with the logo of the Al-Saiqa Brigade, and carrying a weapon, is seen in a video footage shooting with his left hand three male figures in the head. The three persons are kneeling in front of a wall with their arms tied behind their backs. After the bodies of the three men fall, Mr Al-Werfalli shoots at the bodies a few more times. Two of the victims are alleged to have been revolutionaries in Benghazi. The video depicting this incident was posted on social media on 20 March 2017.\nThe second video is available via\nYouTube\nas a 18-second clip, uploaded by a user named “Hassan Mahmoud” (Arabic: حسن محمود). It has been watched over 10,000 times at the time of writing this article.\nUsing Amnesty International’s YouTube Data Viewer, the exact upload time to YouTube can established at March 21, 2017, at 4:09 am local time in Benghazi. The date is inconsistent with the time mentioned by the ICC warrant, though this may have to do with the display date of YouTube and the actual uploading time (hence the importance of Amnesty’s tool).\nIt is unclear when this video was recorded.\nThe exact location of this street has not been found yet, but you can help trying to find the exact location. Hints and possible locations can be posted on Check, where there is a specific entry for the geolocation of this video.\nIncident Three\nLink to the video on Check\n.\nThe ICC arrest warrant states the following:\nMr Al-Werfalli, wearing camouflage trousers and a black t-shirt with the logo of the Al-Saiqa Brigade, and carrying a weapon, is seen in a video footage in a room, with others being present as well. He is standing next to an unidentified man in a white t-shirt, who is kneeling bare feet, with his hands behind his head. Mr Al-Werfalli shoots the person with his left hand in the head and continues to do so after the man falls down. The men around Mr Al-Werfalli cheer in approval and another unidentified man from the group comes forward and shoots the victim two more times. The videos depicting this incident were posted on social media on 7 and 8 May 2017.\nThe third video is still available online on\nFacebook\n, and its duration is 1:17. It has been watched over 32,000 times and shared 242 times at the time of writing.\nThe source is a Facebook profile going by name Bin Mansour with Facebook identification number 100012980119097. It is unclear whether this profile is the original uploader of the video.\nThe accompanying text with the video claims to show the execution of “one of the Kharijites” by Werfalli.\nThe upload time of the video to Facebook is May 7, 2017, at 5:12 pm local time in Benghazi, which is consistent with the time mentioned in the ICC warrant.\nIt is unclear when the video was recorded.\nThe exact location of this house has not been found yet, and may be close to impossible without insider knowledge. You can help trying to find the exact location. Hints and possible locations can be posted on Check.\nIncident Four\nLink to the video on Check\n[geolocated].\nThe ICC arrest warrant states the following:\nMr Al-Werfalli is seen in a video footage in an outdoor setting together with two other men who carry weapons and whose heads are covered. The three men stand behind two other men, who are dressed in camouflage and are kneeling bare feet on the ground. The two men are claimed to be, the first, a member of the BRSC, and the second, possibly a member of Ansar al-Sharia. The two men can be seen earlier in the video footage being detained in a cage. Mr Al-Werfalli is seen walking away from the kneeling men, while the other two men with firearms walk towards them. Mr Al-Werfalli has his left hand in the air and sweeps it down towards the ground in a manner that suggests that he is ordering them to proceed with the execution. The two men shoot the kneeling persons who fall on the ground. The video depicting this incident was posted on social media on or about 22 May 2017.\nThe fourth video is also still available online on\nFacebook\n, with a duration of 2:47. It has been watched almost 37,000 times and shared 340 times as of writing this article. Only the last 9 seconds of the video show the execution, and the rest of the video shows interviews with the two individuals who are claimed to be members of the BRSC and Ansar Al-Sharia.\nThe source is the same Facebook profile as the one that posted the video showing Incident One. It is unclear whether this profile is the original source of the video.\nThe accompanying text with the video claims to show “the execution of the death penalty” (Arabic: تنفيذ حكم الإعدام) in Benghazi of a BRSC member and an Ansar Al-Sharia member.\nThe upload time of the video to Facebook is May 22, 2017, at 11:27 am local time in Benghazi. This date is consistent with the date mentioned by the ICC.\nIn the fourth video, Werfalli is seen in an outdoor setting with two other men wearing balaclavas. They stand behind two other men who are claimed to be a member of the BRSC and Ansar Al-Sharia. Werfalli orders the two man wearing balaclavas to proceed with the execution of the two other men.\nThe execution site has been\ngeolocated\nby Twitter-user @5urpher to a compound in Al-Hawari, a southern suburb of Benghazi. The coordinates are 32.046755, 20.116750 (\nGoogle Maps\n,\nWikimapia\n).\nA satellite image of Google Earth of the compound in southern Benghazi where the execution shown in the Incident Four video took place.\nIt is unclear when this video was recorded, though once it is geolocated it may be able to determine an approximate time of the day based on the clearly visible shadows.\nIncident Five\nhttps://www.youtube.com/watch?time_continue=1&v=kievfxloCcU\nLink to the video on Check\n[geolocated].\nThe ICC arrest warrant states the following:\nMr Al-Werfalli, wearing a full camouflage uniform, is seen in a video footage in an outdoor setting together with five men wearing hoods and camouflage trousers and holding firearms. Those five hooded, armed men stand behind four other hooded men, who are kneeling barefoot on the ground with their hands behind their backs. They point their firearms at each kneeling person’s head. Mr Al-Werfalli raises his left hand in the air and sweeps it down towards the ground in a manner that suggests that he is ordering them to proceed with the execution. The five men shoot the four kneeling persons, who fall on the ground. The video depicting this incident was posted on social media on 9 June 2017.\nThe fifth video can be found online on\nFacebook\n, uploaded by someone using the username “aalathry1” with Facebook identification number 1797850107. The execution takes place in the last 47 seconds of the video.\nThe video was uploaded on June 9, 2017 at 2:29 pm local time in Benghazi. This date is consistent with the ICC warrant.\nThe exact date when this execution took place is unknown.\nThe exact location of this execution site has been\ngeolocated\nby Twitter-user @5urpher to a military compound between Benghazi and the airport. The coordinates are\n32.122988, 20.164492\n. The location was found by\nusing\nthe lines of electricity of OpenStreetMap in Benghazi.\nThe location of Incident Five (yellow pin) in Google Earth, with an overlay of electricity pylons (black squares).\nIncident Six\nLink to the video on Check\n.\nThe ICC arrest warrant states the following:\nMr Al-Werfalli is seen in a video footage in an area of desert together with two other men who are holding firearms. They stand behind two other persons, who are kneeling on the ground with their hands behind their back. Mr Al-Werfalli is seen speaking into the camera and then raising his left hand in the air and sweeping it down towards the ground in a manner that suggests that he is ordering the two men to proceed with the execution. The men shoot the persons kneeling, who fall on the ground. The video depicting this incident was posted on social media on 19 June 2017.\nThe sixth video is available on both\nFacebook\nand\nYouTube\nwith a duration of 59 seconds. The video has been watched 80,000 times on Facebook and 18,000 times on YouTube.\nThe Facebook source is a page – not an account – with username “Benghazi2”.\nThe YouTube source is a channel named “Libya Al-Mostakbal”, which has over 1,000 subscribers. The accompanying text states that the video was published on social media, suggesting Libya Al-Mostakbal reuploaded the video and is not the original source. It says that Werfalli is shown in the video, ordering the execution of two men on an unknown date at an unknown place.\nThe upload time of the video to Facebook is June 19, 2017, at 9:19 pm, which is consistent with the time mentioned in the ICC arrest warrant.\nIt is unclear when this video was recorded.\nThe exact location of this desert area has not been found yet, but you can help trying to find the exact location. Hints and possible locations can be posted on Check, where there is a specific entry for the geolocation of this video.\nIncident Seven\nhttps://www.youtube.com/watch?v=oMfohLyw8d8\nLink to the video on Check\n.\nThe ICC arrest warrant states the following:\nMr Al-Werfalli, wearing a black cap, camouflage trousers and a black t-shirt bearing the logo of the Al-Saiqa Brigade, is seen in a video footage, together with two other persons wearing camouflage trousers, black t-shirts, and holding firearms. Mr Al-Werfalli is holding a white document from which he reads. He refers to the document as a “Decree decision” which “shall be executed today 17/07/2017”. The video also depicts 18 persons wearing orange jumpsuits and black hoods, with their hands tied behind their backs, kneeling bare feet on the ground, in four lines. After reading the document, Mr Al-Werfalli says “Attention! The first group!” and then “Ready! Aim! Fire!” As he does so, five men in camouflage uniform, hooded, and carrying firearms, walk from behind the group of kneeling persons to the front and position themselves behind the first line of five kneeling persons. They raise their firearms at them and shoot. The persons fall on the ground. One person in the second line also falls. The shooters then return behind the group of the other kneeling persons.\nThen Mr Al-Werfalli is recorded saying “The second group! Ready! Aim! Fire!” As he does so, five men walk again from behind the group and position themselves behind the second line of five kneeling persons. The person who has collapsed during the first shooting is pulled back to his knees by one of the men. The five men then raise their firearms and shoot the five kneeling persons, who fall on the ground. The shooters then return behind the group of the other kneeling persons.\nWhen Mr Al-Werfalli says “Attention! Attention, the third group!” and “Ready! Aim! Fire!” five men walk again from behind the group, position themselves behind the third line of five kneeling persons, raise their weapons and shoot them. The persons fall on the ground and the shooters return behind the group of the remaining kneeling persons.\nMr Al-Werfalli and the two other persons wearing camouflage trousers, black t-shirts, and holding firearms then walk behind the final line of three remaining kneeling persons. Mr Al-Werfalli is seen to still hold the white document from which he has read earlier. Mr Al-Werfalli and the other two men raise their firearms and shoot the three kneeling persons, who fall on the ground. He and one of the other men fire some additional shots towards the persons on the ground before walking away.\nIn the same video, in a new segment, Mr Al-Werfalli, wearing a black cap, camouflage trousers and a black t-shirt bearing the Al-Saiqa Brigade logo, is seen together with two men wearing orange jumpsuits and black hoods kneeling on the ground. One of the two kneeling men has his hands in front of him while the other has his hands on the back of his head. Mr Al-Werfalli is holding a white document folded in his hand. He is recorded saying “Ready! Aim! Fire!” As he does so, two hooded men in camouflage, carrying firearms, walk forward, raise their weapons and shoot at the two kneeling persons, who fall on the ground. The men fire some additional shots at the victims before walking away.\nThe video depicting this incident, involving in total 20 executed persons, was posted on social media on 23 July 2017. The data indicates that the video may have been uploaded in the Benghazi area, in Libya.\nThe seventh and last video is also still available online. A 4:46-long video was posted to\nFacebook\nby the same account that posted the Incident One and Incident Four videos. Segments of that video have been used by several other YouTube accounts from\nmedia organisations\n. It is unclear whether this profile is the original source of the video(s).\nThe accompanying text with the video claims to show the execution of twenty “terrorists” sentenced to death because they were allegedly involved in kidnapping, torture, killing, bombing, and slaughtering.\nThe upload time of the video to Facebook is July 23, 2017, at 5:14pm local time in Benghazi, which is consistent with the time mentioned in the arrest warrant.\nIt likely that the execution took place on July 17, 2017. The accompanying text with the video claims it was filmed in Benghazi.\nThe exact location of this location has not been found yet, but you can help trying to find the exact location. Hints and possible locations can be posted on Check, where there is a specific entry for the geolocation of this video. A stitched panorama shown above may be of help. The buildings in the upper side of the panorama are a potential clue to the location.\nThe Promise and Perils of Open Source Evidence\nThe Promise: A Goldmine of Potential Evidence\nThe widespread proliferation of new communication technologies, such as smartphones connected to the Internet, has increased the already unprecedented amount of information available about many of today’s conflicts.\nWhether it is an onlooker who recorded an airstrike on his or her mobile phone, or a militant group that filmed an offensive for propaganda purposes, the material is very often published on the Internet and, specifically, social media. It may be that possible human rights violations are recorded in that footage.\nOnce corroborated and triangulated, this open source information can be a potential mine of evidence. Some jurisdictions have already used it to convict a suspect. A Swedish district court, for example,\nconvicted\na Syrian national of shooting a person with an assault rifle in 2012, partly basing the judgement on a video that was posted to Facebook. The man was sentenced to life in prison.\nThe international community has limited ability to collect impartial information and freely assist civilians inside many of today’s conflict zones like Myanmar, Syria, and Iraq due to restriction of movement. With the decrease in international observers that are able to conduct field investigations, open source evidence is promising. Open source information enables anyone with an Internet connection to mine information from places that were once blind spots. International human rights courts and tribunals “have expressed hope that new information technologies might bring about better, cheaper, and safer prosecutions,” Keith Hiatt writes in\nThe Yale Law Journal\n.\nNevertheless, there are also methodological challenges of using open source information in legal cases, such as reliability, availability, and security, which will be discussed in the next section.\nThe Perils: Reliability, Availability, Security\nSome of these challenges have been outlined by academics, including Hiatt in the aforementioned article, as well as by Emma Irving in an\nOpinio Juris\narticle as a reaction to the ICC arrest warrant.\nFirstly, open source information is “notoriously susceptible to problems of verifiability, which naturally affects its reliability in any kind of criminal proceeding” (Irving). Questions that arise may be the establishment of the exact time, date or location of the evidence, or even potentially digitally altered footage.\nHowever, there is a wide variety of\ntools and methods\nwhich can be used to determine the exact location of a video or photograph, and sometimes even the approximate date and time. Verifiability is a challenge that can be overcome in many cases. It is unclear whether the ICC knows the exact location of the execution sites.\nWith the rapid development of technology and artificial intelligence, a new challenge is on the horizon: Can we be sure a person seen in a video is in fact the same person the video purports to depict? Offenders have been incorrectly identified before, as in the aftermath of the Boston Marathon bombing or the assaults at the Charlottesville white supremacist rally. Even more worrying is the possibility to produce a photorealistic video, say, of former US President Barack Obama\nspeaking sentences he never said\n.\nBut neither of these issues are a challenge in the Werfalli case. Most, if not all, of the footage mentioned by the ICC appears to be produced – and shared – by individual members or supporters of the Saiqa Brigade.\n[1]\nThe more salient question here, if this case goes to trial, is whether the defence team will contest the authenticity of the footage, according to Hiatt. In many international criminal cases, Hiatt says, “the issue is not whether the atrocity occurred, as a matter of actual fact. Instead, it is the context that matters.” For example, was there a fair trial before the execution, or not? That question is unlikely to hinge on whether an execution video was authentic. Indeed, the contested elements of a crime under the Rome Statute are more likely to concern chain of command, or the widespread, systematic nature of a criminal act.\nSecondly, the availability of open source information may be another challenge. While the arrest warrant for Werfalli was issued shortly after the alleged crimes, human rights tribunals are usually only established after the cessation of hostilities. “As a result, today’s investigations concern yesterday’s atrocities,” as the formation of these tribunals may take years (Hiatt).\nThis can be a problem, as was highlighted by YouTube’s recent removal of many videos, playlists, accounts, and channels documenting conflict zones. As outlined in a\nNew York Times\narticle, YouTube “inadvertently removed thousands of videos that could be used to document atrocities in Syria, potentially jeopardizing future war crimes prosecutions.” The same has happened with videos from other areas where human rights violations are reportedly happening, such as\nMyanmar\n.\nFor that very reason, it is important to preserve, for the long term, open source documentation relating to possible human rights violations and other crimes. Projects like the\nSyrian Archive\nare entirely focused on this, and Bellingcat hopes to broaden these efforts to other conflict areas in the future too. (Also, as you may have noticed, most of the hyperlinks in this article link to archived versions of the original page in case a source is taken offline).\nThe Syrian Archive is entirely focused to preserve, for the long term, open source documentation relating to possible human rights violations and other crimes.\nA third challenge is security. Open source evidence may increase threats for eyewitnesses and those who gathered the information first-hand, or even bystanders visible in a photo or video. As Hiatt said,\n“By using these materials, investigators may draw unwanted attention to people depicted in them. The disclosure of a photograph to a powerful defendant or a hostile government might expose the identity of a witness, or otherwise endanger third parties.”\nThis challenge does not seem of particular relevance to the Werfalli case, as all of the material appears to have been recorded and subsequently disseminated on social media by members or sympathisers of the Saiqa Brigade themselves.\nIn case eyewitnesses are used, the use of open source evidence may actually decrease the risk for eyewitnesses as fewer witnesses may have to testify, and in case they do their claims can be corroborated with open source information.\nThe fourth and last challenge is forward-looking, and was put forward by Irving: the weight of evidence. Will the approach to open source evidence change depending on the stage of proceedings?\n“The standard of proof for the issuance of an arrest warrant is ‘reasonable grounds to believe’ (Article 58(1) of the Rome Statute). One might assume that the approach to open source evidence would not be stricter when it came to initiating an investigation (‘reasonable basis to believe’, Article 53(1)(a)) but the situation may be different as regards the higher standards of ‘substantial grounds to believe’ (confirmation of charges, Article 61(5)) and ‘beyond reasonable doubt’ (conviction, Article 66(3)).”\nWill it be the case that the higher the burden, the lesser the chance that open source evidence can meet it? The decision on the arrest warrant was not the place to answer such questions, but they will need to be addressed in future.\nDiscussion\nHuman rights organisations like Amnesty International welcomed the arrest warrant,\nstating\nthe ICC’s decision is “a significant step towards ending the rampant impunity for war crimes in Libya.” Amnesty International has previously\nreported\non potential war crimes committed by the LNA.\nThe US, the UK, and France welcome the ICC arrest warrant in a joint statement. This statement can be read in two possible ways, Alex Whiting writes for\nJust Security\n.\nThe first is that “the U.S. (a non-State Party of the ICC) is joining two ICC State Parties [France and the UK] to support and encourage a positive and pro-active response to the ICC warrant [..meaning..] support for the Court consistent with U.S. law […] as long as it is in the national interest.”\nThe second and more plausible read of the joint statement is, according to Whiting, that the US, France, and the UK, “are looking for a way to\navoid\nthe ICC. If the LNA did in fact pursue a proper investigation and prosecution of Al-Werfalli for the crimes for which he is charged, an unlikely proposition to be sure, it could render him ineligible to be prosecuted at all by the ICC.”\nThe Werfalli case has also potential implications for the US and other states engaged in Libya, since General Haftar is a dual US-Libyan citizen, Whiting notes. For more information on this issue, Whiting’s article is recommended reading.\nOn Libyan social media, criticism has been raised about the ICC’s arrest warrant. Questions are asked\npublicly\nabout why “\nthe heroic\n” Werfalli – who killed alleged Islamic State militants – would be prosecuted. Why are his enemies not prosecuted? Or, for instance, why not culprits of the forced displacement and killing of the\nRohingya in Myanmar\n, or occupation forces in the\nWest Bank\n?\nCrimes committed in these areas may have been recorded and, for example, have been uploaded to YouTube. If these videos are deleted by YouTube’s AI, the platform risks losing “the richest source of information about human rights violations in closed societies” Hiatt told the\nNew York Times\n. The ICC faced similar criticism from many top African officials, who\naccused\nthe court of solely targeting their continent.\nEither way, the decision of the ICC to rely on social media to support an arrest warrant is an important development, and may be only the beginning of a much larger increase of open source evidence by the Court. Open source information is a promise too important to ignore.\n[1]\nHowever, there have been cases in which armed groups were imitated by others, such as a video in which militants from the Ukrainian Azov Battalion threatened the Netherlands. That video turned out to be\nfake\n.\nSpecial thanks to\nKeith Hiatt\nand\nThomas Feneux\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/territorial-contorl-1200x826.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/Werfalli-resignation-1200x675.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/Incident-One-Still-1-1200x675.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/Incident-One-3-1200x576.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/Incident-One-2.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/Incident-One-4-1200x293.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/DI826MwWsAErDJ7.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/Incident1_15-03-2017_matching.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/Incident-Four-1200x640.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/DI448YPW4AAVh7o.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/Panorama-Video6-1200x396.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/electricity_poles_visible.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/Incident-Six-1200x675.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/Incident-Seven-1200x341.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/09/syrian-archive-1200x447.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2017-10-03_how-an-execution-site-was-geolocated.json b/bellingcat_geolocation_json/2017-10-03_how-an-execution-site-was-geolocated.json
new file mode 100644
index 0000000000000000000000000000000000000000..83b76ea84a6781084c2eaf9b0be0af6ee385555c
--- /dev/null
+++ b/bellingcat_geolocation_json/2017-10-03_how-an-execution-site-was-geolocated.json
@@ -0,0 +1,97 @@
+{
+ "slug": "how-an-execution-site-was-geolocated",
+ "title": "How a Werfalli Execution Site Was Geolocated",
+ "url": "https://www.bellingcat.com/news/middle-east/2017/10/03/how-an-execution-site-was-geolocated/",
+ "date": "2017-10-03T16:30:33+00:00",
+ "modified": "2018-02-10T00:27:20+00:00",
+ "author": "Bellingcat Investigation Team",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "This article was collaboratively researched and written by the Bellingcat Investigation Team. Introduction In August 2017, the International Criminal Court (ICC) issued its first ever arrest warrant solely based on social media evidence. The arrest warrant accuses Mahmoud Mustafa Busayf Al-Werfalli (hereafter referred to as ‘Werfalli’), an alleged commander of the Al-Saiqa Brigade of the […]",
+ "content_html": "
\n
This article was collaboratively researched and written by the Bellingcat Investigation Team.
\n\n
Introduction
\n
In August 2017, the International Criminal Court (ICC) issued its first ever arrest warrant solely based on social media evidence. The arrest warrant accuses Mahmoud Mustafa Busayf Al-Werfalli (hereafter referred to as ‘Werfalli’), an alleged commander of the Al-Saiqa Brigade of the Libyan National Army (LNA), of mass executions in or near Benghazi.
\n
Bellingcat has highlighted the importance of this development, which aligns the ICC with the realities of many of today’s conflicts, in an earlier article. The article marked the start of a crowdsourcing campaign to geolocate the seven locations of the executions on which the arrest warrant is based. Four execution sites have been geolocated so far, and one of them – Incident Seven – is worth discussing in further depth as it is an excellent example of team work and using a wide variety of open source information such as satellite imagery, geotagged photos, conflict context, geolocations from other sources and news articles.
\n
The Evidence: A Facebook Video
\n
The video which prompted the ICC arrest warrant was uploaded on July 23, 2017, by a Facebook user named ‘Mohammed Al-Gali’. The video has been deleted since, though an archived version exists.
\n
The video shows the execution of twenty individuals who are, according to the description, “terrorists found guilty of kidnapping, torture, killing, bombing, and slaughtering” and members of the so-called Islamic State (IS). Werfalli is visible in the video and gives the order to execute the group of twenty individuals. According to the arrest warrant, Werfalli reads from a white document in the video of Incident Seven, and refers to a ‘decree decision’, which ‘shall be executed today 17/07/2017’.
\n
The first part of the video shows the execution of eighteen people and the second part the execution of another two people. The long shadows in the video indicate it was filmed in the early morning (from north to south) or late afternoon (from south to north).
\n
The Geolocation: Bushes, Buildings, and Blood Stains
\n
At first sight the video seems very difficult to geolocate, because of the apparent lack of identifiable landmarks. Aside from the twenty executed people, their executioners and a white car, mainly a desert landscape with two sand tracks, a few sand hills and some vegetation is visible. However, a stitched panorama of the video, published in Bellingcat’s previous article, shows buildings and a fence on a distance in the background.
\n
The buildings are visible for less than a second, at 3m25s. A fence consisting of poles is visible in the second part of the video, from 4m19s onward. Besides the vegetation pattern, these are the only visual clues to go on for the geolocation.
\n
Image 1. A panorama of the execution site of Incident Seven. The panorama is composed out of stills of the Facebook video.
\n
Image 2. A still from the Incident Seven video at 3m25s, showing the structures in the background of the execution site.
\n
Image 3. A fence is visible in the background of a still of the video at 4m25s.
\n
The Bellingcat team was able to geolocate the video in Ghanfuda, which is situated in the western part of Benghazi.
\n
The video was filmed from north to south, very likely on July 17, 2017, approximately at 6:37 am. The coordinates of the execution area are 32.023144, 20.029181.
Image 5. Screenschot of SunCalc at the execution location at 17 July 2017, 6:37 am.
\n
Image 6. An overlay of a still of the Incident Seven video on satellite imagery with a 3D reconstruction of the buildings in that area.
\n
On satellite imagery, vegetation patches and a hill are visible. These features directly correspond with the vegetation patches and hill which are visible in the Incident Seven video. Furthermore, the two sand tracks shown in the video are clearly visible on the satellite imagery.
Traces of the execution are also distinctly visible on July 17, 2017, satellite imagery of TerraServer. The satellite image’s resolution is 0.30m, the highest resolution available for consumers. Fifteen black spots are visible, which are very likely blood stains of the executed people. These spots are at the exact same location as where the twenty individuals were shot. Satellite imagery of the 11th and 14th of July 2017 clearly shows these spots were not visible yet on those dates.
Image 9. Comparison of the location of the bodies in the video and the spots in the satellite imagery makes clear the bodies in the video and the spots in the satellite imagery are on exactly the same location.
\n
How Was the Location Found?
\n
As mentioned in the previous article about the Werfalli executions, the buildings in the left upper side of the panorama gave a strong clue about the location of the video. The buildings visible in the video seem to be gray concrete buildings with “arches” without doors.
\n
Twitter user @ROELart mentioned that the video possibly was filmed in a southwest area of Benghazi because the color of the sand is more grey than in other areas of Benghazi, where colors of the sand are more orange or yellow.
\n
In that area we noticed many concrete buildings that were not finished yet, such as an area with new buildings that is named ‘Bingazi Meammar’, located to the southwest of Benghazi, as described on Wikimapia.
\n
Another area with ‘Chinese Buildings 24’ is mentioned on Wikimapia as an area captured by the Libyan National Army (LNA) from the Benghazi Revolutionaries’ Shura Council (BRSC) in January 2017. These buildings are a part of a much bigger project of new real estate, developed by China, for 25.000 housing units in the Ghanfuda area. Construction of the ‘Chinese Buildings’ started in 2008 or 2009, but stopped in 2011 because of the civil war in Libya and thousands of Chinese construction workers left Libya. The same year there were plans to continue the construction, but that never happened and the construction never was completed. In January 2017 terrorists were hiding inside ‘Buildings 12’, likely an area with 12 buildings in Ghanfuda, Benghazi.
\n
One photograph of the ‘Chinese Buildings’ strongly resembles to the buildings visible in the execution video, especially because of a structure where the front and side of the building are alternated. Satellite imagery of ‘Chinese Buildings’ confirmed the strong similarity. However, since the area consist of hundreds of these buildings, geolocating the right spot seemed a very difficult task.
\n
Image 10. Buildings in the execution video (top) compared to the ‘Chinese Buildings’ (bottom).
\n
Image 11. Similarities between the ‘Chinese Buildings’ on satellite imagery (top) and the buildings visible on a distance in the execution video (bottom).
\n
Mohammed Al-Gali uploaded several videos where similar unfinished concrete buildings are visible. On 2 February 2017 he uploaded a video showing a panorama view around, filmed from the roof of one of the buildings. This video could be geolocated to 32° 1’31.29″N, 20° 1’48.84″E (32.025358, 20.030233) in a north western corner of the area with ‘Chinese Buildings’ in Ghanfuda. The description of the video likely refers to the ’12 buildings’ where terrorists were hiding and a battle against these terrorists.
\n
Image 12. Screenshot of the 2 February 2017 video, showing an area with unfinished concrete buildings and several cranes. The description mentions ’12 buildings’ and a battle with terrorists, who were hiding there.
\n
Two other videos, filmed on the 12th (but uploaded at the 13th) and 14th of June 2017, show a group of people, including Werfalli and what appear to be soldiers of his unit, digging up a large number of munition and weapons, buried by the terrorists, with a bulldozer. These videos were geolocated at 32° 1’40.13″N, 20° 1’43.89″E (32.027814, 20.028858) and 32° 1’37.65″N, 20° 1’42.59″E (32.027125, 20.028497), just a few blocks of buildings away from the 2 February video location. Both videos mention that the ammunition and weapons were buried at sites inside ‘Chinese Buildings’.
\n
Image 13. Screenshots from the 12 and 14 June 2017 videos, showing unfinished concrete buildings, according to the description weapons and munition were buried inside ‘Chinese Buildings’.
\n
Image 14. Screenshots from the 12 and 14 June 2017 videos, showing unfinished concrete buildings, according to the description weapons and munition were buried inside ‘Chinese Buildings’.
\n
The 12th June 2017 video shows similar white cars as visible in the execution video, although a definitive match between these cars could not be made.
\n
Because the 2nd of February and the 12th and 14th June 2017 videos were filmed in the same area, Bellingcat expected the execution location to be in the same area as well. It was eventually a fence visible in the 12 June 2017 video that led to the execution location.
\n
Image 15. A fence to the north of the buildings is visible in the 12 June 2017 video (marked in red rectangle).
\n
On the northern side of the area with buildings and continuing to the western side of this area, a long fence consisting of poles leads from north to south. Following that fence to the south reveals an open area with several dirt tracks crossing each other. One of these crossings and the vegetation around turned out to be a perfect match with the landscape in the execution video.
\n
Image 16. Google Earth satellite imagery showing the locations of the 2 February, 12 and 14 June 2017 videos. The 17 July 2017 execution video was filmed in the same area.
\n
Conclusion
\n
The geolocation of the execution video of Incident Seven of the Werfalli executions is a very strong match. The similarities between the vegetation, sand tracks, buildings, fence and spots on the exact location where the twenty individuals were executed, allows the conclusion that the exact location of the executions has been found. Changes to satellite imagery strongly indicates Mahmoud Al-Werfalli commanded the execution of twenty individuals on this location at some point between the 15th and 17th of July, 2017 during the early morning. Werfalli himself states in the video the execution took place on July 17th, which is consistent in the changes in the satellite imagery.
\n
Update: In a previous version of this article, Mohammed Al-Gali was described as a potential person who filmed one or more videos described in this post along with the filmed individuals. Al-Gali has reached out to Bellingcat and refuted this, saying that he had nothing to do with the filming of the video, and that he reshared it from an earlier source. We have updated the article to remove any mention of Al-Gali as a potential original source of the video.
On October 19, 2017, the Joint Investigation Team (JIT), the Dutch-led criminal investigation into the downing of Malaysian Airlines Flight 17 (MH17) over Ukraine, published a new photograph Buk 332, the Russian Buk missile launcher that downed MH17 on July 17, 2014. Buk 332, previously identified by Bellingcat as ‘Buk 3×2,’ is an anti-aircraft missile launcher belonging to Russia’s Kursk-based 53rd Anti-Aircraft Missile Brigade. The image shows the right side of the Buk, loaded on the red low loader. In publishing the photograph, the JIT wrote that the photograph was “probably taken on July 17, 2014 in the town of Makeevka, Ukraine,” but we can now conclusively say that the photograph was actually taken at 78 Prospekt Ilycha, at the intersection with Shakhtostroiteley, in Donetsk.
\n
\n
The Hunt
\n
Using Check, a crowdsourced verification platform, we asked our readers to contribute tips regarding the location of this photograph so that we could collectively geolocate this key piece of evidence concerning the downing of MH17.
One of the strongest candidates was 78 Prospekt Ilycha in Donetsk, due to previous witness accounts identifying a Buk missile launcher with its Russian/separatist convoy at this site.
“Bad news. Around 9am, a hauler was going along the Makeevka highway from Makeevka in the direction of Donetsk. On the platform was a BukM1-M2? This AAMS proceeded to the intersection with Shakhtostroiteley Boulevard. The system was accompanied by a convoy that was composed of 1 gray Rav4 SUV, a camouflaged UAZ, and a dark blue Hyundai van with tinted windows. As of 9:15am, the vehicle was located at the intersection of Shakhtostroiteley and Ilycha. The militants got out of their cars, blocking 2 of the far left lanes. Obviously, they were waiting for logistical guidance.”
\n
Numerous features in the photograph were also visible in Google Street View and Yandex Panorama imagery of the same location, including the two trees in the foreground, a gate beneath the Buk’s low-loader, a rock next to the nearest tree, and notches in the curb blocks.
\n
\n
Our first guess is in Donetsk, at the intersection of Sakhtostroiteley and Ilych (near a beauty salon). Weigh in at the Check page. pic.twitter.com/1TlxxgEAtT
The walkway along the salon, where cameraman stood, is much lower than the curb, which explains low position of camera on the \"BUK\" picture. pic.twitter.com/xTb37YesAF
Taking these tips together, the following comparisons show how some key details match between the two photographs, making the site worthy of further investigation to confirm the geolocation.
\n
\n
Left: A part of a gate visible in an open space through the red low loader, just under the Buk missile launcher. \nRight: the same type of gate visible on October 2011 Google Street View of Prospekt Illycha in Donetsk, Ukraine.
\n
\n
Left: The new Buk 332 photograph with a thick tree, a slanting branch of another tree, a little rock and notches in the curb, all marked in red. \nRight: the same landmarks in Google Street View marked in red.
\n\n
Conclusive Evidence
\n
Today, at least one local in Donetsk snapped two photographs from this same location, attempting to confirm or refute a geolocation to Prospekt Ilycha in central Donetsk. These photographs were shared online by Christo Grozev and Rudy Bouma.
(1) Gisteren vrijgegeven foto Buk lijkt op foto's die ik vandaag liet maken niet in Makeevka (zoals #JIT vermoedt) maar hier Donetsk. #MH17pic.twitter.com/oJGyGqkvd6
With this geolocation, we now have photographic evidence of a scene that was reported on by numerous locals: Buk 332, originating from Russia’s 53rd Anti-Aircraft Missile Brigade, arrived in Donetsk from Makiivka around 9am on July 17, 2014, parked at the intersection of Prospekt Ilycha and Shakhtostroiteley, and then eventually turned towards the east-bound Prospekt Ilycha towards the Motel roundabout in Donetsk. The movement is shown in the map below, with the photographer of the original Buk photo facing south from building number 78.
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nOn October 19, 2017, the Joint Investigation Team (JIT), the Dutch-led criminal investigation into the downing of Malaysian Airlines Flight 17 (MH17) over Ukraine,\npublished a new photograph Buk 332\n, the Russian Buk missile launcher that downed MH17 on July 17, 2014. Buk 332, previously identified by Bellingcat as ‘Buk 3×2,’ is an anti-aircraft missile launcher belonging to Russia’s Kursk-based 53rd Anti-Aircraft Missile Brigade. The image shows the right side of the Buk, loaded on the red low loader. In publishing the photograph, the JIT wrote that the photograph was “probably taken on July 17, 2014 in the town of Makeevka, Ukraine,” but we can now conclusively say that the photograph was actually taken at 78 Prospekt Ilycha, at the intersection with Shakhtostroiteley, in Donetsk.\nThe Hunt\nUsing Check, a crowdsourced verification platform,\nwe asked our readers\nto contribute tips regarding the location of this photograph so that we could collectively geolocate this key piece of evidence concerning the downing of MH17.\nGeolocating this\n#MH17\nphotograph seems impossible, but we already have a lead. Help us to crowdsource this task:\nhttps://t.co/OuZcd7fZbA\npic.twitter.com/2P90SHMhxw\n— bellingcat (@bellingcat)\nOctober 19, 2017\nOne of the strongest candidates was 78 Prospekt Ilycha in Donetsk, due to previous witness accounts identifying a Buk missile launcher with its Russian/separatist convoy at this site.\nSource\n(\nArchive\n)\n“Bad news. Around 9am, a hauler was going along the\nMakeevka highway\nfrom Makeevka in the direction of Donetsk. On the platform was a BukM1-M2? This\nAAMS\nproceeded to the\nintersection with Shakhtostroiteley Boulevard\n. The system was accompanied by a convoy that was composed of 1 gray Rav4 SUV, a camouflaged UAZ, and a dark blue Hyundai van with tinted windows.\nAs of 9:15am, the vehicle was located at the intersection of Shakhtostroiteley and Ilycha. The militants got out of their cars, blocking 2 of the far left lanes.\nObviously, they were waiting for logistical guidance.”\nNumerous features in the photograph were also visible in Google Street View and Yandex Panorama imagery of the same location, including the two trees in the foreground, a gate beneath the Buk’s low-loader, a rock next to the nearest tree, and notches in the curb blocks.\nOur first guess is in Donetsk, at the intersection of Sakhtostroiteley and Ilych (near a beauty salon). Weigh in at the Check page.\npic.twitter.com/1TlxxgEAtT\n— bellingcat (@bellingcat)\nOctober 19, 2017\nThe walkway along the salon, where cameraman stood, is much lower than the curb, which explains low position of camera on the \"BUK\" picture.\npic.twitter.com/xTb37YesAF\n— Zitty (@zitty055)\nOctober 20, 2017\nOne of the branches behind the hauler looks the same as a tree on Google Maps.\npic.twitter.com/1pKh0N3eJq\n— Wout Mager (@Eworm_)\nOctober 19, 2017\nLooking the other side becomes even clearer; this is at the Illicha Ave in Donetsk.\n#mh17\npic.twitter.com/TkYpW0O665\n— Roelard Smit (@ROELart)\nOctober 20, 2017\n99% a match\npic.twitter.com/mbwHxG5ohE\n— Lorenzo Romani (@lorenzoromani)\nOctober 19, 2017\nTaking these tips together, the following comparisons show how some key details match between the two photographs, making the site worthy of further investigation to confirm the geolocation.\nLeft: A part of a gate visible in an open space through the red low loader, just under the Buk missile launcher.\nRight: the same type of gate visible on October 2011 Google Street View of Prospekt Illycha in Donetsk, Ukraine.\nLeft: The new Buk 332 photograph with a thick tree, a slanting branch of another tree, a little rock and notches in the curb, all marked in red.\nRight: the same landmarks in Google Street View marked in red.\nConclusive Evidence\nToday, at least one local in Donetsk snapped two photographs from this same location, attempting to confirm or refute a geolocation to Prospekt Ilycha in central Donetsk. These photographs were shared online by Christo Grozev and Rudy Bouma.\nPhoto from same location TODAY proves\n@bellingcat\n& global volunteers were on target\npic.twitter.com/TyENGC7qaI\n— Christo Grozev (@christogrozev)\nOctober 20, 2017\nAnd here is a more comprehensive angle from today\npic.twitter.com/KiQL3aAras\n— Christo Grozev (@christogrozev)\nOctober 20, 2017\n(1) Pictures I had made today in DNR are placing new picture of\n#Buk\nin Donetsk, not in Makeevka like\n@JITMH17\nis suspecting.\n#MH17\npic.twitter.com/yNOTfTyxcM\n— rudy bouma (@rudybouma)\nOctober 20, 2017\n(1) Gisteren vrijgegeven foto Buk lijkt op foto's die ik vandaag liet maken niet in Makeevka (zoals\n#JIT\nvermoedt) maar hier Donetsk.\n#MH17\npic.twitter.com/oJGyGqkvd6\n— rudy bouma (@rudybouma)\nOctober 20, 2017\nConclusion\nWith this geolocation, we now have photographic evidence of a scene that was reported on by numerous locals: Buk 332, originating from Russia’s 53rd Anti-Aircraft Missile Brigade, arrived in Donetsk from Makiivka around 9am on July 17, 2014, parked at the intersection of Prospekt Ilycha and Shakhtostroiteley, and then eventually turned towards the east-bound Prospekt Ilycha towards the Motel roundabout in Donetsk. The movement is shown in the map below, with the photographer of the original Buk photo facing south from building number 78.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/10/9OF67tBB15o.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/10/donetsketoukraina-1.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/10/gate_geolocation.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/10/notches.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/10/buk_comparison.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2017/10/map-kurazh.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git "a/bellingcat_geolocation_json/2018-09-04_identify-burnt-villages-satellite-imagery\342\200\212-case-studies-california-nigeria-myanmar.json" "b/bellingcat_geolocation_json/2018-09-04_identify-burnt-villages-satellite-imagery\342\200\212-case-studies-california-nigeria-myanmar.json"
new file mode 100644
index 0000000000000000000000000000000000000000..d55aaeb2871354e58f8b46c831890a122b8c2d39
--- /dev/null
+++ "b/bellingcat_geolocation_json/2018-09-04_identify-burnt-villages-satellite-imagery\342\200\212-case-studies-california-nigeria-myanmar.json"
@@ -0,0 +1,107 @@
+{
+ "slug": "identify-burnt-villages-satellite-imagery -case-studies-california-nigeria-myanmar",
+ "title": "How to Identify Burnt Villages by Satellite Imagery — Case-Studies from California, Nigeria and Myanmar",
+ "url": "https://www.bellingcat.com/resources/how-tos/2018/09/04/identify-burnt-villages-satellite-imagery -case-studies-california-nigeria-myanmar/",
+ "date": "2018-09-04T05:16:49+00:00",
+ "modified": "2018-09-04T11:53:57+00:00",
+ "author": "Benjamin Strick",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)Evidence of human rights abuses often remains in some of the world’s most regional and unreachable parts. It is in these areas, where a view from satellite imagery above can be the sole method of visual evidence of bombings, attacks, and the destruction of communities. As satellite imagery becomes more available and technology […]",
+ "content_html": "
Evidence of human rights abuses often remains in some of the world’s most regional and unreachable parts.
\n
It is in these areas, where a view from satellite imagery above can be the sole method of visual evidence of bombings, attacks, and the destruction of communities.
\n
As satellite imagery becomes more available and technology permits for more access, we are seeing overhead mapping play more of an integral role in media and human rights bodies to identify a chronological story of an area.
\n
For some areas, this may be through infrastructure expansions, military operations, or what new equipment has just landed on the tarmac. For other parts, satellite imagery can reveal signs of chemical attacks or villages that have been burnt to the ground.
\n
While this imagery does help, it does not replace evidence obtained from the ground — but it goes a long way to help us know what happens in these conflict areas.
\n
Creating a Fire Damage Reference Scale with Google Earth
\n
Satellite images courtesy of Google Earth / DigitalGlobe.
\n
As a reference point for what fire damage looks like on satellite imagery, we can use California’s devastating 2017 fires as a litmus scale.
\n
The above images are available on Google Earth and are from Coffey Park, a suburban area in Santa Rosa, California, that saw widespread destruction of houses, buildings, and bordering reserves.
\n
In the chronological satellite timeline above, you can see specific signs that a fire has taken place. These are signatures left on the earth when damage has been caused by fire, namely:
\n
\n
Blackening of the earth where any grass or shrub once stood
\n
Levelling of infrastructure exposing its foundations
\n
Clear white patches formed by beds of ash
\n
\n
The most significant of these indicators to appear on imagery are the white patches of ash left behind, as you will see, these are significant markers when scanning more remote areas for signs of attacks against people and their homes.
\n
There are also other ways to confirm fire damage in satellite imagery, for this we can use Sentinel Hub to determine vegetation levels and other useful options.
\n
Using Sentinel Hub to Confirm Fire Damage
\n
Sentinel Hub’s playground feature has a number of interesting rendering options on its satellite imagery.
\n
A number of these can support our findings of fire damage. In the example above of the same urban area in Coffey Park, Santa Rosa, we can view the “Color Infrared (vegetation)” band on the satellite imagery.
\n
The zoom level is limited in Sentinel Hub, however using this filter to identify vegetation reveals a significant difference between pre-burn and post-burn. For widespread bushfires such as California’s fire, much of the red vegetation was burnt and destroyed, but there is a definite change in the level of bright red to a much darker image, indicating a significant loss of vegetation.
\n
This change in vegetation levels is illustrated in the images below. At the top is the concentration of vegetation in our area in Coffey Park from July, 2017, while on the bottom is the vegetation from October, 2017, after the fire had swept through.
\n
\n
Using these open source tools, we can find, identify, and analyse areas that have been caused by means other than natural causes.
\n
Identifying Villages Destroyed by Fire
\n
Context also plays an important role in satellite imagery, as these overhead views will only go so far as to tell us that a building, hut, or community, has been destroyed.
\n
They will not tell us how it has been destroyed, or why, but using evidence from satellite imagery provides us with a firm timeline from which we can base our search filters on.
\n
This is where open source information, through the use of reports from non-governmental organisations (NGOs), human rights bodies, news agencies, and social media, in conjunction with satellite imagery, will create a much better picture of what has occurred.
\n
In Musari Village, east of Maiduguri in Nigeria’s Borno State, we can see evidence of damage to a number of the village’s huts.
\n
\n
The area surrounding the village is dry with a lack of vegetation, therefore we do not see the blackening of the earth as widespread as in the Californian fires examples. However, it is evident that the small hut infrastructure has burned, leaving ash markings.
\n
In a closer view, we can see more of those signature markings.
\n
\n
Correlating this with a Google search using the given dates on the satellite imagery indicate that this is an attack by Boko Haram militants.
\n
Dalori (below), also near Maiduguri in Nigeria, shows evidence of more permanent structures that have been destroyed by fire.
\n
\n
Using Sentinel Hub’s vegetation feature, we can further highlight the changes we see on Google Earth.
\n
Sentinel Hub imagery is also more frequently updated than Google Earth satellite imagery, so we are able to narrow down dates using these features.
\n
\n
To substantiate the damage, a comparison between two dates can be made to see what has changed in this village.
\n
As you can see below, the area is the same, but on a chronological scale, they are completely different. Roofs are missing, walls destroyed, significant trees burnt and there is a blackening of the earth around the structures, indicative of scorching.
\n
\n
Running a chronologically filtered Google search between these dates provides us with reports of scores of children and adults that were burned to death in this village by more than 80 militants.
\n
Using the Burned Village Litmus on Myanmar
\n
As a standard approach, reporters and human rights researchers can use these tools to identify clear abuses, in conjunction with contextual information.
\n
For instance, in Myanmar’s Rakhine State, where mass village destruction has been reported, the signs of village destruction are very evident.
\n
\n
Using Sentinel Hub, it is very easy to determine where fire has caused a rapid decrease in the vegetation count. You can see this in the following satellite imagery with the view at the top from July, 2017, and on the bottom, from October, 2017.
\n
\n
As we can see in Sentinel Hub, vegetation can reveal where fires have been, whether it is in a Nigerian desert environment such as in the previous examples, or in a more vegetated environment such as Myanmar.
\n
Taking a closer look at Google Earth imagery provides a more clear image as to the destruction of the overview of villages in the Rakhine State.
\n
\n
\n
In a closer look, we can see the same aspects that both the natural Californian fires had, as well as the deliberately burnt villages in Nigeria.
\n
\n
These signature traits of white spots of ash, charred ground and a substantial decrease in vegetation (both visually and via band imaging) are present in all markings of burnt villages.
\n
All in all, satellite imagery allows anyone with an Internet connection to identify potential evidence of new human rights abuses, and such open-source information — from Google Earth or Sentinel Hub — and may just as well assist media and human rights organisations to verify such claims.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nEvidence of human rights abuses often remains in some of the world’s most regional and unreachable parts.\nIt is in these areas, where a view from satellite imagery above can be the sole method of visual evidence of bombings, attacks, and the destruction of communities.\nAs satellite imagery becomes more available and technology permits for more access, we are seeing overhead mapping play more of an integral role in media and human rights bodies to identify a chronological story of an area.\nFor some areas, this may be through infrastructure expansions, military operations, or what new equipment has just landed on the tarmac. For other parts, satellite imagery can reveal signs of chemical attacks or villages that have been burnt to the ground.\nWhile this imagery does help, it does not replace evidence obtained from the ground — but it goes a long way to help us know what happens in these conflict areas.\nCreating a Fire Damage Reference Scale with Google Earth\nSatellite images courtesy of Google Earth / DigitalGlobe.\nAs a reference point for what fire damage looks like on satellite imagery, we can use California’s devastating 2017 fires as a litmus scale.\nThe above images are available on Google Earth and are from\nCoffey Park\n, a suburban area in Santa Rosa, California, that saw widespread destruction of houses, buildings, and bordering reserves.\nIn the chronological satellite timeline above, you can see specific signs that a fire has taken place. These are signatures left on the earth when damage has been caused by fire, namely:\nBlackening of the earth where any grass or shrub once stood\nLevelling of infrastructure exposing its foundations\nClear white patches formed by beds of ash\nThe most significant of these indicators to appear on imagery are the white patches of ash left behind, as you will see, these are significant markers when scanning more remote areas for signs of attacks against people and their homes.\nThere are also other ways to confirm fire damage in satellite imagery, for this we can use\nSentinel Hub\nto determine vegetation levels and other useful options.\nUsing Sentinel Hub to Confirm Fire Damage\nSentinel Hub’s playground feature has a number of interesting rendering options on its satellite imagery.\nA number of these can support our findings of fire damage. In the example above of the same urban area in Coffey Park, Santa Rosa, we can view the “Color Infrared (vegetation)” band on the satellite imagery.\nThe zoom level is limited in Sentinel Hub, however using this filter to identify vegetation reveals a significant difference between pre-burn and post-burn. For widespread bushfires such as California’s fire, much of the red vegetation was burnt and destroyed, but there is a definite change in the level of bright red to a much darker image, indicating a significant loss of vegetation.\nThis change in vegetation levels is illustrated in the images below. At the top is the concentration of vegetation in our area in Coffey Park from July, 2017, while on the bottom is the vegetation from October, 2017, after the fire had swept through.\nUsing these open source tools, we can find, identify, and analyse areas that have been caused by means other than natural causes.\nIdentifying Villages Destroyed by Fire\nContext also plays an important role in satellite imagery, as these overhead views will only go so far as to tell us that a building, hut, or community, has been destroyed.\nThey will not tell us how it has been destroyed, or why, but using evidence from satellite imagery provides us with a firm timeline from which we can base our search filters on.\nThis is where open source information, through the use of reports from non-governmental organisations (NGOs), human rights bodies, news agencies, and social media, in conjunction with satellite imagery, will create a much better picture of what has occurred.\nIn\nMusari Village\n, east of Maiduguri in Nigeria’s Borno State, we can see evidence of damage to a number of the village’s huts.\nThe area surrounding the village is dry with a lack of vegetation, therefore we do not see the blackening of the earth as widespread as in the Californian fires examples. However, it is evident that the small hut infrastructure has burned, leaving ash markings.\nIn a closer view, we can see more of those signature markings.\nCorrelating this with a\nGoogle search\nusing the given dates on the satellite imagery indicate that this is an attack by Boko Haram militants.\nDalori\n(below), also near Maiduguri in Nigeria, shows evidence of more permanent structures that have been destroyed by fire.\nUsing Sentinel Hub’s vegetation feature, we can further highlight the changes we see on Google Earth.\nSentinel Hub imagery is also more frequently updated than Google Earth satellite imagery, so we are able to narrow down dates using these features.\nTo substantiate the damage, a comparison between two dates can be made to see what has changed in this village.\nAs you can see below, the area is the same, but on a chronological scale, they are completely different. Roofs are missing, walls destroyed, significant trees burnt and there is a blackening of the earth around the structures, indicative of scorching.\nRunning a chronologically filtered Google search between these dates provides us with\nreports\nof scores of children and adults that were burned to death in this village by more than 80 militants.\nUsing the Burned Village Litmus on Myanmar\nAs a standard approach, reporters and human rights researchers can use these tools to identify clear abuses, in conjunction with contextual information.\nFor instance, in\nMyanmar’s Rakhine State\n, where mass village destruction has been reported, the signs of village destruction are very evident.\nUsing Sentinel Hub, it is very easy to determine where fire has caused a rapid decrease in the vegetation count. You can see this in the following satellite imagery with the view at the top from July, 2017, and on the bottom, from October, 2017.\nAs we can see in Sentinel Hub, vegetation can reveal where fires have been, whether it is in a Nigerian desert environment such as in the previous examples, or in a more vegetated environment such as Myanmar.\nTaking a closer look at Google Earth imagery provides a more clear image as to the destruction of the overview of villages in the Rakhine State.\nIn a closer look, we can see the same aspects that both the natural Californian fires had, as well as the deliberately burnt villages in Nigeria.\nThese signature traits of white spots of ash, charred ground and a substantial decrease in vegetation (both visually and via band imaging) are present in all markings of burnt villages.\nAll in all, satellite imagery allows anyone with an Internet connection to identify potential evidence of new human rights abuses, and such open-source information — from Google Earth or Sentinel Hub — and may just as well assist media and human rights organisations to verify such claims.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS1.jpeg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS2-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS3-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS4.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/vs5.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS6.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS7.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS8.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS9.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS10.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS11.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS12.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS13.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS14.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS15.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS16.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS17.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/09/VS18.png",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2018-11-08_europols-asian-city-child-abuse-photographs-geolocated.json b/bellingcat_geolocation_json/2018-11-08_europols-asian-city-child-abuse-photographs-geolocated.json
new file mode 100644
index 0000000000000000000000000000000000000000..fc682800312e32372fb0c948e1b3fd48637c92ec
--- /dev/null
+++ b/bellingcat_geolocation_json/2018-11-08_europols-asian-city-child-abuse-photographs-geolocated.json
@@ -0,0 +1,77 @@
+{
+ "slug": "europols-asian-city-child-abuse-photographs-geolocated",
+ "title": "Europol’s Asian City Child Abuse Photographs Geolocated",
+ "url": "https://www.bellingcat.com/resources/case-studies/2018/11/08/europols-asian-city-child-abuse-photographs-geolocated/",
+ "date": "2018-11-08T09:21:33+00:00",
+ "modified": "2018-11-08T09:55:32+00:00",
+ "author": "Daniel Romein",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)Since 2017 the European Union Agency of Law Enforcement Cooperation, better known as Europol, has been crowdsourcing parts of or heavily censored photographs, related to child abuse crimes in their “Stop Child Abuse – Trace an Object” campaign. Within two weeks after the start of their campaign Europol received 10.000 contributions from the […]",
+ "content_html": "
Since 2017 the European Union Agency of Law Enforcement Cooperation, better known as Europol, has been crowdsourcing parts of or heavily censored photographs, related to child abuse crimes in their “Stop Child Abuse – Trace an Object” campaign.
\n
Within two weeks after the start of their campaign Europol received 10.000 contributions from the public. Bellingcat, among many others, shared this campaign on the crowdsourcing platform Check. Many followers of this campaign have helped by identifying objects and/or geolocating photographs. The location of a photograph in a hotel room was proved to be taken in a hotel on Mauritius within 48 hours after Europol shared the photograph on Twitter.
\n
Europol published a press release on the success of their campaign earlier this year and mentioned that out of 70 objects, 25 of them were identified to one country or else to a reasonable number of countries of production and on 1 June 2018, one year after the start of the campaign, a press release describes that Europol received 21.000 tips and more objects and countries of production were identified. Another press release from October 2018 makes clear that eight children have been identified and one offender has been arrested because of citizen efforts.
\n
Europol shared new images on their website and via Twitter on 15 October 2018. These were mostly of objects that need to be identified, but there were also a few photographs that were taken outside and are possible to geolocate because of recognizable landmarks. Two of these photographs, taken from a roof of a building, show concrete buildings, and were presumably taken in an Asian city. According to Europol, a child was sexually abused in this city. Europol’s investigators need the location of the photograph to be able to trace the abuser and save a victim. The photographs are heavily censored, because of the sensitivity of the material.
\n
Europol’s tweet from 15 October 2018 (left) and the two photographs taken from a roof in an Asian city (right).
\n
Geolocation of the Photographs
\n
At first glance, the photograph doesn’t seem to contain many recognizable landmarks. No text is displayed on the buildings and the concrete architecture of the buildings only indicates that the photographs have been taken in an Asian city. Many Twitter users responded to the tweet and mentioned a former Soviet Republic, Malaysia, Philippines or Indonesia as possible locations where the photographs were taken, yet the majority of the responses indicated Southern China as the most likely location.
\n
Twitter user “Bo” contacted Bellingcat and mentioned the architecture especially shows similarity to the city of Shenzhen in Southern China. Bellingcat responded to Europol’s tweet with this information, included a photograph of similar architecture and an overlay image of the two photographs, mentioning a blue road sign and a structure similar to a satellite receiver on top of a building visible in the photographs.
\n
Sixteen days later, on 31 October 2018 Twitter user Olli Enne from Finland geolocated the exact location of the photographs in the Bao’an district of Shenzhen in Southern China. According to Olli, the photographs have been taken from the roof of a building with coordinates 22.722917, 114.053194 and, he explained, several buildings and a hill in the photographs match the buildings visible in satellite imagery. Also, a view line across a building with a blue roof to a building with arch shaped windows in the distance lines up with the view line in satellite imagery. In later tweets, Olli explained that he searched in Shenzhen and other major cities in China for several hours, looking for little green hills and road shapes, and that he draw a map how the area would look on satellite imagery.
\n
Tweet from 31 October 2018 from Olli Enne (left) and the image that explains the geolocation of the photographs (right).
\n
The geolocation of the photographs could not be immediately verified by Bellingcat, as it wasn’t easy to match the buildings of the photographs to the buildings visible in satellite imagery, mainly because of the distance from the location where the photographs were presumably taken to these buildings.
\n
Yet due to street view on Baidu maps, a Chinese web mapping system, we were able to verify that Olli’s geolocation is a perfect match. A white building visible on the left side in one of the photographs shows strong similarity to a building visible in a 2014 street view on Baidu, just as a white building with several balconies visible on the right side in one photograph shows a match to a building visible in an older street view as well. Also the buildings with red and orange tiles on the roof from the photographs, as well as a blue road sign, are visible street view. A hotel further away, the building with arch shaped windows and several buildings at the end of the road near a hill with trees, one of them demolished in 2017, are all visible in recent or older street views as well. Street view shows a white building with three antennas on the roof, similar to a white building with what looks like three antennas on a roof in the photographs.
\n
Overlay image of Europol’s photographs with enhanced contrast. The distance between the buildings on the left and on the right side, as well as a blue road sign, indicate there’s a bigger road in between these buildings. A smaller street along the white building with balconies crosses this road. Several buildings and also the blue road sign from the photographs are visible in recent or older Baidu maps street views.
\n
Also, the building from where the photographs were taken can be geolocated because of the red tiles on the eaves. Street view from 2014 shows red tiles and even a pole on the roof, which is similar to a pole visible in one of the photographs.
\n
The location from where the photographs were taken is the roof of a building visible in Baidu street view (red rectangle). Red tiles on the eaves and also a pole on the roof at the same location as in one of the photographs are clearly visible.
\n
A Google Earth 3D view of the building the photographs were taken from in the same direction as in those photographs shows the same mountainscape. Particularly the shape of the mountain on the left side of the photographs is very similar to the shape of the mountains in the 3D view. A smaller mountain with a relatively high peak is more difficult to spot, but following a view line in the photographs from the location where they were taken in the direction of that mountain shows the same buildings in the 3D view that are visible in the photographs in that view line. Also the partly visible small green hill at the end of the road is clearly visible in Google’s 3D view.
\n
Google Earth 3D view shows the same mountainscape and the nearby small green hill as visible in the photographs (marked in red rectangles). The green lines are the view lines that show the area that is visible in the photographs, while the yellow line shows the view from the building where the photographs have been taken to the smaller mountain with high peak.
\n
Estimating the Year the Photographs Were Taken
\n
Because the buildings in the photographs show more similarity to the same buildings in older street view captures, it is clear that the photographs were not taken recently. A 2017 street view shows one of the buildings was even demolished and the buildings near to the location where the photographs were taken have brown netting, i.e. protective constructions for air conditioning. A 2014 street view capture shows GSM antennas on the white building on the left that are not visible in Europol’s photographs. Recent street view shows many GSM antennas and satellite receivers in the area, while the Europol photographs only show one satellite receiver and three antennas.
\n
Satellite imagery tells us the photographs had to have been taken after September 2002, as in the photographs one building clearly has a blue painted roof, which is not visible in satellite imagery from September 2002, but is visible in satellite imagery from February 2008. Also, two buildings visible slightly in the distance on the right side in one of the two photographs are not visible in 2002 satellite imagery, but are visible in 2008 satellite imagery. A factory and other tall buildings, visible in the photographs behind the building with arch shaped windows were not constructed before February 2003. Satellite imagery from November 2013 shows the construction of a high rise building that should be visible in one of the photographs if they were taken in or after 2013. However, a smaller building is visible in the photograph, but the high rise building that is standing behind it, as visible in 2016 street view, is not visible in the photograph. It is very difficult, perhaps even impossible, to say exactly in which year the photographs were taken. The roughest estimation is between 2003 and 2013, so it’s most likely they were taken around 2008.
\n
Left: two buildings visible on the right side in one of the two photographs that have been constructed after September 2002. Right: A high rise building constructed in 2013 that is not visible in Europol’s photographs, but the smaller building in front is visible.
\n
Satellite imagery of the area visible in the photographs from 2002 compared to imagery of 2008: a blue painted roof (blue square) and two buildings (green rectangle) from the photographs are visible in 2008, but not in the 2002 satellite imagery.
\n
The Initial Research
\n
The next part describes the initial research of Bellingcat and others, that eventually helped Olli Enne in finding the exact location. This initial research specifies many details, which mainly are added to show our method of working, but also to show which exact steps were needed to come to our results.
\n
Bellingcat’s research started when Twitter user “Bo” shared a video filmed in Shenzhen, China, showing very similar architecture as in Europol’s photographs at the 4:17 time mark. Important note: The people in the video have no relation at all to Europol’s campaign, the video is just meant as an illustration of what the architecture in Shenzhen looks like.
\n
“Bo,” who is very involved in Europol’s campaign, has cooperated with Bellingcat for this cause and is a member of research collective Serendip, which played an important role in solving one of the biggest child abuse cases in the Netherlands.
\n
A screenshot from a video filmed in Shenzhen, China, showing very similar architecture as in Europol’s photographs. The building on the front from the video has similar windows as a building on the left side in one of the photographs. Also, the smaller buildings in the area with red and orange tiles bear similarities to buildings visible in the photographs.
\n
The video doesn’t give much clues as to the several locations where it was filmed, certainly not for people who are not able to read Chinese. One part obviously is filmed in Shekou, a modern harbour area with many high rise buildings and clearly not in the same location as the older buildings.
\n
A Cantonese restaurant, visible shortly before the part featuring the older buildings and filmed likely in the same area, is geolocatable for a Chinese speaker. However, text recognition with a OCR-tool didn’t work, because the characters are in a particular style and also not filmed in a straight angle. An earlier part in the video, on the other hand, shows a street with shops. At the wall on the side of one of the shops a small blue sign with a street name is visible, but the same street name and house number is also more clearly visible on the store sign itself. The colon after two characters indicates the name of the street starts after that colon.
\n
An aforementioned free online OCR-tool, NewOCR, can be used to recognize text in many languages. However, the tool is quite sensitive in a way that the text has to be straight and no other parts of the image can be included, the smallest deviation results in the tool not recognizing the text. After a few (partly) failed attempts, the tool recognized the text as 南岭东路49号, which (literally) translates as “Nanling East Road 49 number.” Google maps doesn’t display the correct location, but it seems to be the correct area. However, Baidu maps shows the correct location with a search for 南岭东路49 (without word “number”). The stores have changed between the street view date and the date the video was filmed, but it’s obviously the same location.
\n
Screenshots from the video showing a Cantonese restaurant (left) and shops on Nanling East Road nr. 49 (right). The name of the street including house number is marked with a blue rectangle and this part of the image was used to read the text with an OCR-tool.
\n
A short part of the video shows a drone shot from the area and suggests the Cantonese restaurant is one of the buildings visible in the drone shot. The area of the drone shot is found easily near to the location of the shops on Nanling East Road. Particularly, a building with a blue and white roof is very recognizable.
\n
Screenshot from the video showing a drone shot (left) and the same location on Google Earth (right).
\n
From here, the exact location of the buildings visible on 4:17 in the video is also not hard to find, as it is just a bit further south from the drone shot area. The high rise buildings visible in the video, but also the area in the video at 4:20 (showing a green edge and a small building with a blue roof next to a higher building) is visible in satellite imagery.
\n
A closer look at the part of video that shows the Cantonese restaurant at 3:43 shows the same green edge in the distance and the same buildings visible at 4:17. Bellingcat contacted a person living in Shenzhen about the location of the restaurant and the spelling in Chinese characters. This individual told us that the name of the restaurant is spelled 金沙湾茶餐厅, which translates as Golden Sands Bay Tea Restaurant. The restaurant is displayed on Google maps in the correct street, but the map doesn’t match the satellite imagery exactly. The map, in fact, moves the restaurant 500 meters to the east of the real location on satellite imagery. Baidu maps shows the restaurant almost at the right location and displays a few reviews, as well as an older photograph of the restaurant. The address of the restaurant, 深圳市龙岗区樟树布新村一巷10, which translates very odd, is in fact the first alley on nr. 10 of that neighbourhood (Yushubu new village) in Longgang district, Shenzhen.
\n
The exact location marked by a red dot where the video is filmed from at the 4:17 mark: the buildings with similar architecture as in Europol’s photographs and the high rise buildings are visible to the east from the red dot. The green edge and small building with a blue roof are to the south-southwest of the red dot. The Cantonese restaurant is marked in an orange square.
\n
Unfortunately, no match was found between this eastern suburb of Shenzhen and Europol’s photographs, although the mountainscape in eastern direction seems to be a match. After finding many more areas in Shenzhen with similar architecture and mountain scapes, no precise match to the photographs was found. We realized that the chance to find the exact location in a city of 2000 square kilometers (compare it to London, with 1500 square kilometers) is like trying to find a needle in a haystack. At that time it was not even clear the location is in Shenzhen, as other major cities in the area like Dongguan, Guangzhou and likely many more cities in China have similar architecture.
\n
Yet the Bellingcat team found another clue that the photographs possibly were taken in a northern or eastern suburb of Shenzhen: a video filmed from a tram in Shenzhen shows a building with three antennas on the roof, very similar to what looks like three antennas on top of a building in one of the photographs. We also noticed many satellite receivers on the roofs of many buildings in the same area. By following the direction of the tram in the video simultaneously to the street view in Baidu we were able to find the exact location of the building with three antennas.
\n
In this area, we also could not find an exact match with Europol’s photographs, but we noticed again the similarity between the architecture of this area and the photographs. A few days later, when the exact location was found by Twitter user Olli Enne, we realized how close we were to the exact spot: 1.2 kilometer as the crow flies or a 20 minute, 1.5 kilometer walk. The tram in the video stops at Xinlan, which is even less than a kilometer from the exact photograph’s location.
\n
A screenshot of a video filmed from the Qinghu-Xinlan tram in Shenzhen shows three antennas on a roof of a building with a similar shape as a structure visible in Europol’s photographs and a satellite dish is also visible. A street view of the same area shows many more satellite dishes.
\n
Conclusion
\n
This research shows that in a complex investigation tiny details can be very important. The architecture, the tiles with red and orange colors, the blue roof and the mountainous area narrowed down the research to Southern China. Details in a video lead us to an eastern suburb of Shenzhen with similar architecture. Another video shows us GSM antennas and a satellite dish, also visible in the photographs, and lead us to a northern suburb of Shenzhen. Tiny details as buildings that are not visible in the photographs, but are visible in satellite imagery, and GSM antennas missing in the photographs, but which are visible in street view, show the photographs are at least five years old, likely even older.
\n
It takes hours of research, and a complex investigation as this one clearly can not be solved by one person or a small team; a crowdsourced wisdom is required to solve cases like this. When even with crowdsourcing a location or an object can’t be found, it can feel very frustrating. Yet, when a photograph is geolocated or an object is recognized, you realize that there is a chance that the case will be solved and someone will get arrested for a serious crime — and it was you who contributed to that.
\n
Research by Daniel Romein and Timmi Allen, editing by Natalia Antonova.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nSince 2017 the European Union Agency of Law Enforcement Cooperation, better known as Europol, has been crowdsourcing parts of or heavily censored photographs, related to child abuse crimes in their “\nStop Child Abuse – Trace an Object\n” campaign.\nWithin two weeks after the start of their campaign Europol received 10.000 contributions from the public. Bellingcat, among many others, shared this\ncampaign\non the crowdsourcing platform\nCheck\n. Many followers of this campaign have helped by identifying objects and/or geolocating photographs. The location of a photograph in a hotel room was proved to be taken in a\nhotel on Mauritius\nwithin 48 hours after Europol shared the photograph on Twitter.\nEuropol published a\npress release\non the success of their campaign earlier this year and mentioned that out of 70 objects, 25 of them were identified to one country or else to a reasonable number of countries of production and on\n1 June 2018\n, one year after the start of the campaign, a press release describes that Europol received 21.000 tips and more objects and countries of production were identified. Another press release from\nOctober 2018\nmakes clear that eight children have been identified and one offender has been arrested because of citizen efforts.\nEuropol shared new images on their website and via Twitter on 15 October 2018. These were mostly of objects that need to be identified, but there were also a few photographs that were taken outside and are possible to geolocate because of recognizable landmarks. Two of these photographs, taken from a roof of a building, show concrete buildings, and were presumably taken in an Asian city. According to Europol, a child was sexually abused in this city. Europol’s investigators need the location of the photograph to be able to trace the abuser and save a victim. The photographs are heavily censored, because of the sensitivity of the material.\nEuropol’s tweet from 15 October 2018 (left) and the two photographs taken from a roof in an Asian city (right).\nGeolocation of the Photographs\nAt first glance, the photograph doesn’t seem to contain many recognizable landmarks. No text is displayed on the buildings and the concrete architecture of the buildings only indicates that the photographs have been taken in an Asian city. Many Twitter users responded to the tweet and mentioned a former Soviet Republic, Malaysia, Philippines or Indonesia as possible locations where the photographs were taken, yet the majority of the responses indicated Southern China as the most likely location.\nTwitter user “Bo” contacted Bellingcat and mentioned the architecture especially shows similarity to the city of Shenzhen in Southern China. Bellingcat responded to Europol’s tweet with this information, included a photograph of similar architecture and an overlay image of the two photographs, mentioning a blue road sign and a structure similar to a satellite receiver on top of a building visible in the photographs.\nSixteen days later, on\n31 October 2018\nTwitter user Olli Enne from Finland geolocated the exact location of the photographs in the Bao’an district of Shenzhen in Southern China. According to Olli, the photographs have been taken from the roof of a building with coordinates\n22.722917, 114.053194\nand, he explained, several buildings and a hill in the photographs match the buildings visible in satellite imagery. Also, a view line across a building with a blue roof to a building with arch shaped windows in the distance lines up with the view line in satellite imagery. In later tweets, Olli explained that he searched in Shenzhen and other major cities in China for several hours, looking for little green hills and road shapes, and that he draw a map how the area would look on satellite imagery.\nTweet from 31 October 2018 from Olli Enne (left) and the image that explains the geolocation of the photographs (right).\nThe geolocation of the photographs could not be immediately verified by Bellingcat, as it wasn’t easy to match the buildings of the photographs to the buildings visible in satellite imagery, mainly because of the distance from the location where the photographs were presumably taken to these buildings.\nYet due to street view on\nBaidu maps\n, a Chinese web mapping system, we were able to verify that Olli’s geolocation is a perfect match. A white building visible on the left side in one of the photographs shows strong similarity to a building visible in a\n2014 street view\non Baidu, just as a white building with several balconies visible on the right side in one photograph shows a match to a building visible in an\nolder street view\nas well. Also the buildings with\nred\nand\norange\ntiles on the roof from the photographs, as well as a blue road sign, are visible street view.\nA hotel\nfurther away, the building with\narch shaped windows\nand\nseveral buildings\nat the end of the road near a hill with trees, one of them\ndemolished in 2017\n, are all visible in recent or older street views as well. Street view shows a white building with\nthree antennas on the roof\n, similar to a white building with what looks like three antennas on a roof in the photographs.\nOverlay image of Europol’s photographs with enhanced contrast. The distance between the buildings on the left and on the right side, as well as a blue road sign, indicate there’s a bigger road in between these buildings. A smaller street along the white building with balconies crosses this road. Several buildings and also the blue road sign from the photographs are visible in recent or older Baidu maps street views.\nAlso, the building from where the photographs were taken can be geolocated because of the red tiles on the eaves.\nStreet view from 2014\nshows red tiles and even a pole on the roof, which is similar to a pole visible in one of the photographs.\nThe location from where the photographs were taken is the roof of a building visible in Baidu street view (red rectangle). Red tiles on the eaves and also a pole on the roof at the same location as in one of the photographs are clearly visible.\nA Google Earth 3D view of the building the photographs were taken from in the same direction as in those photographs shows the same mountainscape. Particularly the shape of the mountain on the left side of the photographs is very similar to the shape of the mountains in the 3D view. A smaller mountain with a relatively high peak is more difficult to spot, but following a view line in the photographs from the location where they were taken in the direction of that mountain shows the same buildings in the 3D view that are visible in the photographs in that view line. Also the partly visible small green hill at the end of the road is clearly visible in Google’s 3D view.\nGoogle Earth 3D view shows the same mountainscape and the nearby small green hill as visible in the photographs (marked in red rectangles). The green lines are the view lines that show the area that is visible in the photographs, while the yellow line shows the view from the building where the photographs have been taken to the smaller mountain with high peak.\nEstimating the Year the Photographs Were Taken\nBecause the buildings in the photographs show more similarity to the same buildings in older street view captures, it is clear that the photographs were not taken recently. A 2017 street view shows one of the buildings was even demolished and the buildings near to the location where the photographs were taken have brown netting, i.e. protective constructions for air conditioning. A 2014 street view capture shows GSM antennas on the white building on the left that are not visible in Europol’s photographs. Recent street view shows many GSM antennas and satellite receivers in the area, while the Europol photographs only show one satellite receiver and three antennas.\nSatellite imagery tells us the photographs had to have been taken after September 2002, as in the photographs one building clearly has a blue painted roof, which is not visible in satellite imagery from September 2002, but is visible in satellite imagery from February 2008. Also,\ntwo buildings\nvisible slightly in the distance on the right side in one of the two photographs are not visible in 2002 satellite imagery, but are visible in 2008 satellite imagery.\nA factory and other tall buildings\n, visible in the photographs behind the building with arch shaped windows were not constructed before February 2003. Satellite imagery from November 2013 shows the construction of a high rise building that should be visible in one of the photographs if they were taken in or after 2013. However, a smaller building is visible in the photograph, but the high rise building that is standing behind it, as visible in\n2016 street view\n, is not visible in the photograph. It is very difficult, perhaps even impossible, to say exactly in which year the photographs were taken. The roughest estimation is between 2003 and 2013, so it’s most likely they were taken around 2008.\nLeft: two buildings visible on the right side in one of the two photographs that have been constructed after September 2002. Right: A high rise building constructed in 2013 that is not visible in Europol’s photographs, but the smaller building in front is visible.\nSatellite imagery of the area visible in the photographs from 2002 compared to imagery of 2008: a blue painted roof (blue square) and two buildings (green rectangle) from the photographs are visible in 2008, but not in the 2002 satellite imagery.\nThe Initial Research\nThe next part describes the initial research of Bellingcat and others, that eventually helped Olli Enne in finding the exact location. This initial research specifies many details, which mainly are added to show our method of working, but also to show which exact steps were needed to come to our results.\nBellingcat’s research started when\nTwitter user “Bo\n” shared a\nvideo\nfilmed in Shenzhen, China, showing very similar architecture as in Europol’s photographs at the 4:17 time mark.\nImportant note\n: The people in the video have no relation at all to Europol’s campaign, the video is just meant as an illustration of what the architecture in Shenzhen looks like.\n“Bo,” who is very involved in Europol’s campaign, has cooperated with Bellingcat for this cause and is a member of research collective\nSerendip\n, which played an important role in solving one of the biggest child abuse cases in the Netherlands.\nA screenshot from a video filmed in Shenzhen, China, showing very similar architecture as in Europol’s photographs. The building on the front from the video has similar windows as a building on the left side in one of the photographs. Also, the smaller buildings in the area with red and orange tiles bear similarities to buildings visible in the photographs.\nThe video doesn’t give much clues as to the several locations where it was filmed, certainly not for people who are not able to read Chinese. One part obviously is filmed in\nShekou\n, a modern harbour area with many high rise buildings and clearly not in the same location as the older buildings.\nA Cantonese restaurant, visible shortly before the part featuring the older buildings and filmed likely in the same area, is geolocatable for a Chinese speaker. However, text recognition with a OCR-tool didn’t work, because the characters are in a particular style and also not filmed in a straight angle. An earlier part in the video, on the other hand, shows a street with shops. At the wall on the side of one of the shops a small blue sign with a street name is visible, but the same street name and house number is also more clearly visible on the store sign itself. The colon after two characters indicates the name of the street starts after that colon.\nAn aforementioned free online OCR-tool,\nNewOCR\n, can be used to recognize text in many languages. However, the tool is quite sensitive in a way that the text has to be straight and no other parts of the image can be included, the smallest deviation results in the tool not recognizing the text. After a few (partly) failed attempts, the tool recognized the text as 南岭东路49号, which (literally) translates as “Nanling East Road 49 number.”\nGoogle maps\ndoesn’t display the correct location, but it seems to be the correct area. However, Baidu maps shows the\ncorrect location\nwith a search for 南岭东路49 (without word “number”). The stores have changed between the street view date and the date the video was filmed, but it’s obviously the same location.\nScreenshots from the video showing a Cantonese restaurant (left) and shops on Nanling East Road nr. 49 (right). The name of the street including house number is marked with a blue rectangle and this part of the image was used to read the text with an OCR-tool.\nA short part of the video shows a drone shot from the area and suggests the Cantonese restaurant is one of the buildings visible in the drone shot.\nThe area of the drone shot\nis found easily near to the location of the shops on Nanling East Road. Particularly, a building with a blue and white roof is very recognizable.\nScreenshot from the video showing a drone shot (left) and the same location on Google Earth (right).\nFrom here, the exact location of the buildings visible on 4:17 in the video is also not hard to find, as it is just a bit further south from the drone shot area. The high rise buildings visible in the video, but also the area in the video at 4:20 (showing a green edge and a small building with a blue roof next to a higher building) is visible in satellite imagery.\nA closer look at the part of video that shows the Cantonese restaurant at 3:43 shows the same green edge in the distance and the same buildings visible at 4:17. Bellingcat contacted a person living in Shenzhen about the location of the restaurant and the spelling in Chinese characters. This individual told us that the name of the restaurant is spelled 金沙湾茶餐厅, which translates as Golden Sands Bay Tea Restaurant. The restaurant is displayed on\nGoogle maps\nin the correct street, but the map doesn’t match the satellite imagery exactly. The map, in fact, moves the restaurant 500 meters to the east of the real location on satellite imagery.\nBaidu maps\nshows the restaurant almost at the right location and displays a few reviews, as well as an older photograph of the restaurant. The address of the restaurant, 深圳市龙岗区樟树布新村一巷10, which translates very odd, is in fact the\nfirst alley\non nr. 10 of that neighbourhood (Yushubu new village) in Longgang district, Shenzhen.\nThe exact location marked by a red dot where the video is filmed from at the 4:17 mark: the buildings with similar architecture as in Europol’s photographs and the high rise buildings are visible to the east from the red dot. The green edge and small building with a blue roof are to the south-southwest of the red dot. The Cantonese restaurant is marked in an orange square.\nUnfortunately, no match was found between this eastern suburb of Shenzhen and Europol’s photographs, although the mountainscape in eastern direction seems to be a match. After finding many more areas in Shenzhen with similar architecture and mountain scapes, no precise match to the photographs was found. We realized that the chance to find the exact location in a city of 2000 square kilometers (compare it to London, with 1500 square kilometers) is like trying to find a needle in a haystack. At that time it was not even clear the location is in Shenzhen, as other major cities in the area like Dongguan, Guangzhou and likely many more cities in China have similar architecture.\nYet the Bellingcat team found another clue that the photographs possibly were taken in a northern or eastern suburb of Shenzhen:\na video filmed from a tram\nin Shenzhen shows a building with three antennas on the roof, very similar to what looks like three antennas on top of a building in one of the photographs. We also noticed many satellite receivers on the roofs of many buildings in the same area. By following the direction of the tram in the video simultaneously to the\nstreet view in Baidu\nwe were able to find the\nexact location\nof the building with three antennas.\nIn this area, we also could not find an exact match with Europol’s photographs, but we noticed again the similarity between the architecture of this area and the photographs. A few days later, when the exact location was found by Twitter user Olli Enne, we realized how close we were to the exact spot: 1.2 kilometer as the crow flies or a 20 minute, 1.5 kilometer walk. The tram in the video stops at Xinlan, which is even less than a kilometer from the exact photograph’s location.\nA screenshot of a video filmed from the Qinghu-Xinlan tram in Shenzhen shows three antennas on a roof of a building with a similar shape as a structure visible in Europol’s photographs and a satellite dish is also visible. A street view of the same area shows many more satellite dishes.\nConclusion\nThis research shows that in a complex investigation tiny details can be very important. The architecture, the tiles with red and orange colors, the blue roof and the mountainous area narrowed down the research to Southern China. Details in a video lead us to an eastern suburb of Shenzhen with similar architecture. Another video shows us GSM antennas and a satellite dish, also visible in the photographs, and lead us to a northern suburb of Shenzhen. Tiny details as buildings that are not visible in the photographs, but are visible in satellite imagery, and GSM antennas missing in the photographs, but which are visible in street view, show the photographs are at least five years old, likely even older.\nIt takes hours of research, and a complex investigation as this one clearly can not be solved by one person or a small team; a crowdsourced wisdom is required to solve cases like this. When even with crowdsourcing a location or an object can’t be found, it can feel very frustrating. Yet, when a photograph is geolocated or an object is recognized, you realize that there is a chance that the case will be solved and someone will get arrested for a serious crime — and it was you who contributed to that.\nResearch by Daniel Romein and Timmi Allen, editing by Natalia Antonova.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/Europol_tweet_and_image-1200x670.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/Tweet_and_image_Olli_Enne.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_overview_lines.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_photo_location_2014.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_photo_view_Google-1200x718.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/buildings_right_far_away_and_huge_building_missing.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/Google_view_2002_2008-1200x523.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_first_location-1200x834.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_restaurant_shops-1200x449.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_map_comparison-1200x339.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_roll_area.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/satellite-receiver_and-dish-00002-1200x675.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2018-11-19_using-time-lapse-satellite-imagery-detect-infrastructure-changes-case-studies-via-myanmar-nigeria-south-china-sea.json b/bellingcat_geolocation_json/2018-11-19_using-time-lapse-satellite-imagery-detect-infrastructure-changes-case-studies-via-myanmar-nigeria-south-china-sea.json
new file mode 100644
index 0000000000000000000000000000000000000000..0c7024f327a9f251d3ee1f3c969fa59d6af2bcea
--- /dev/null
+++ b/bellingcat_geolocation_json/2018-11-19_using-time-lapse-satellite-imagery-detect-infrastructure-changes-case-studies-via-myanmar-nigeria-south-china-sea.json
@@ -0,0 +1,92 @@
+{
+ "slug": "using-time-lapse-satellite-imagery-detect-infrastructure-changes-case-studies-via-myanmar-nigeria-south-china-sea",
+ "title": "Using Time-Lapse Satellite Imagery To Detect Infrastructure Changes: Case-Studies via Myanmar, Nigeria and the South China Sea",
+ "url": "https://www.bellingcat.com/resources/how-tos/2018/11/19/using-time-lapse-satellite-imagery-detect-infrastructure-changes-case-studies-via-myanmar-nigeria-south-china-sea/",
+ "date": "2018-11-19T18:52:35+00:00",
+ "modified": "2019-09-02T13:17:18+00:00",
+ "author": "Benjamin Strick",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)Identifying changes in infrastructure is an important task for those of us surveilling areas for investigative purposes. Satellite imagery can show evidence of everything from the destruction of buildings, the erection of temporary structures, as well as newly constructed permanent facilities. This kind of information is important for analysis of events in conflict […]",
+ "content_html": "
Identifying changes in infrastructure is an important task for those of us surveilling areas for investigative purposes.
\n
Satellite imagery can show evidence of everything from the destruction of buildings, the erection of temporary structures, as well as newly constructed permanent facilities.
\n
This kind of information is important for analysis of events in conflict areas, as these areas often cannot be reached by unbiased observers. Satellite imagery can therefore tell a truer/more complete account of what occurred in a specific conflict area.
\n
By being able to view changes taking place in a specific area, we can utilize free open-source tools which create what is effectively a time-lapse of the Earth’s surface.
\n
Using imagery from the Copernicus Sentinel or NASA Landsat programs, which vary in their daily updates, a timeline of chronological satellite imagery can be compiled, contextualized, and compared to observe and analyze changes taking place below.
\n
Those changes, which can be either disastrous or merely eventful, are helpful in telling the facts of what is happening in a particular area. The best part is that tools that can be used to view those changes can be used by anyone.
\n
Using Satellite Time-Lapse Imagery to Identify Infrastructure Changes in Myanmar
\n
To identify clear signs of the razing of villages and development of new infrastructure, researchers are able to build a time-lapse of satellite imagery using Sentinel Hub’s freely availableEO browser.
\n
In Myanmar’sRakhine State, for example, there have been reports of a mass genocide of the Rohingya. Below is aview of an area through Sentinel Hub that experienced significant change between 2017 and 2018 via Copernicus’ Sentinel-2 satellite imagery.
\n
\n
After creating a free account with Sentinel Hub, the function for time-lapse creation becomes available.
\n
\n
To achieve an effective result for your satellite time-lapse, and to observe change during a set period, select your two dates in the window as seen in the image below.
\n
Knowing that village razing occurred in the Rakhine State between August to December 2017, the selection should capture that particular timeframe.
\n
\n
Be mindful of cloud coverage when selecting the date, as satellite imagery with high cloud coverage may hinder visuals. An advisable setting for cloud coverage is either 20% or less.
\n
\n
Having selected the frames to include in the satellite time-lapse, a presentation will be created to play the imagery in a chronological manner.
\n
Speed adjustments at the bottom can be selected for viewing more frames per second.
\n
The final result for this area shows that a village (previously identified as having beenburned to the ground) was cleared along with surrounding vegetation and a large camp facility was constructed.
\n
\n
Via this manner of presenting chronological consecutive satellite mapping, we can demonstrate that there is evidence of both the removal of structures and the implementation of new, large-scale facilities.
\n
False colour bands can be utilised to identify other areas of Rakhine State that underwent significant change.
\n
\n
As an overview, sections with significantly lighter colours can be seen to emerge in early 2018. The previous time-lapse, for example, shows the development of the large facility in the north.
\n
\n
A closer view of clearing (below) in the south identifies new infrastructure already in place. The development in the centre of the time-lapse shows what used to be village huts (which wereburnt) and is now government buildings. Here is the location onGoogle Maps.
\n
\n
Adding Сustom Script Banding to Time-Lapse Imagery
\n
With satellite imagery, custom rendering can also be used to manipulate, highlight and affect the way a satellite image is viewed.
\n
For instance, should you require the identification of areas that are experiencing large-scale fires and devastation, custom scripting can be used to highlight active fire locations. This works with the satellite time-lapse of Sentinel Hub, or can be used with GIS packaging such asQGIS.
\n
A selection of custom scripts for varying differentiations can be found onGitHub, and onespecific script highlights active fires for Sentinel-2 imagery.
\n
The custom rendering panel can be viewed under “Visualization” (1), and “Custom” (2).
\n
\n
Applying the script has the visual effect of highlighting the intended values.
\n
\n
Using this script, in conjunction with the time-lapse function, shows the before, during, and the after-effects of the fire.
\n
\n
Using this type of fire identification works in effect to identify specific dates for when destructive activity may have occurred, which further assists with research.
\n
As an example, in the time-lapse below, here is how the Nigerian villageNgudoram was engulfed by destructive fire.
\n
\n
False Colour Banding in the South China Sea
\n
Repeating the same process in varying satellite banding can produce effective results for monitoring different areas where new infrastructure is identified.
\n
China’s island military expansion in the South China Sea can be evidenced through the manner of time-lapse imagery. Looking atMischief Reef, which was once a deserted island, and fusing the time-lapse with false colour banding, the island’s new infrastructure can be viewed.
\n
\n
Further scrutiny of satellite imagery, such as using moisture index banding, can be applied. This identifies the level of moisture through the satellite banding.
\n
\n
In this time-lapse above, of Mischief Reef, warmer temperature colours indicate less moisture. The specific areas identified are the tarmac, and one building in the middle of the island — likely to be warm.
\n
Open-Source Remote Sensing Tool Available for Public Use
\n
The ability to view a satellite imagery time-lapse and identify changes taking place on the Earth’s surface is free to use by anyone with a computer and an internet connection.
\n
Using the methods identified in this guide allows anyone to monitor unreachable areas with an analytical approach and identify infrastructure destruction/destructive activities, changes and developments in infrastructure, or anything else, for that matter, on the planet.
\n
Bellingcat’s research for this publication was supported by PAXforPeace.
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nIdentifying changes in infrastructure is an important task for those of us surveilling areas for investigative purposes.\nSatellite imagery can show evidence of everything from the destruction of buildings, the erection of temporary structures, as well as newly constructed permanent facilities.\nThis kind of information is important for analysis of events in conflict areas, as these areas often cannot be reached by unbiased observers. Satellite imagery can therefore tell a truer/more complete account of what occurred in a specific conflict area.\nBy being able to view changes taking place in a specific area, we can utilize free open-source tools which create what is effectively a time-lapse of the Earth’s surface.\nUsing imagery from the Copernicus Sentinel or NASA Landsat programs, which vary in their daily updates, a timeline of chronological satellite imagery can be compiled, contextualized, and compared to observe and analyze changes taking place below.\nThose changes, which can be either disastrous or merely eventful, are helpful in telling the facts of what is happening in a particular area. The best part is that tools that can be used to view those changes can be used by anyone.\nUsing Satellite Time-Lapse Imagery to Identify Infrastructure Changes in Myanmar\nTo identify clear signs of the razing of villages and development of new infrastructure, researchers are able to build a time-lapse of satellite imagery using Sentinel Hub’s freely available\nEO browser\n.\nIn Myanmar’s\nRakhine State\n, for example, there have been reports of a mass genocide of the Rohingya. Below is a\nview of an area\nthrough Sentinel Hub that experienced significant change between 2017 and 2018 via Copernicus’ Sentinel-2 satellite imagery.\nAfter creating a free account with Sentinel Hub, the function for time-lapse creation becomes available.\nTo achieve an effective result for your satellite time-lapse, and to observe change during a set period, select your two dates in the window as seen in the image below.\nKnowing that village razing occurred in the Rakhine State between August to December 2017, the selection should capture that particular timeframe.\nBe mindful of cloud coverage when selecting the date, as satellite imagery with high cloud coverage may hinder visuals. An advisable setting for cloud coverage is either 20% or less.\nHaving selected the frames to include in the satellite time-lapse, a presentation will be created to play the imagery in a chronological manner.\nSpeed adjustments at the bottom can be selected for viewing more frames per second.\nThe final result for this area shows that a village (previously identified as having been\nburned\nto the ground) was cleared along with surrounding vegetation and a large camp facility was constructed.\nVia this manner of presenting chronological consecutive satellite mapping, we can demonstrate that there is evidence of both the removal of structures and the implementation of new, large-scale facilities.\nFalse colour bands\ncan be utilised to identify other areas of Rakhine State that underwent significant change.\nAs an overview, sections with significantly lighter colours can be seen to emerge in early 2018. The previous time-lapse, for example, shows the development of the large facility in the north.\nA closer view of clearing (below) in the south identifies new infrastructure already in place. The development in the centre of the time-lapse shows what used to be village huts (which were\nburnt\n) and is now government buildings. Here is the location on\nGoogle Maps\n.\nAdding Сustom Script Banding to Time-Lapse Imagery\nWith satellite imagery, custom rendering can also be used to manipulate, highlight and affect the way a satellite image is viewed.\nFor instance, should you require the identification of areas that are experiencing large-scale fires and devastation, custom scripting can be used to highlight active fire locations. This works with the satellite time-lapse of Sentinel Hub, or can be used with GIS packaging such as\nQGIS\n.\nA selection of custom scripts for varying differentiations can be found on\nGitHub\n, and o\nne\nspecific script\nhighlights active fires for Sentinel-2 imagery.\nThe custom rendering panel can be viewed under “Visualization” (1), and “Custom” (2).\nApplying the script has the visual effect of highlighting the intended values.\nUsing this script, in conjunction with the time-lapse function, shows the before, during, and the after-effects of the fire.\nUsing this type of fire identification works in effect to identify specific dates for when destructive activity may have occurred, which further assists with research.\nAs an example, in the time-lapse below, here is how the Nigerian village\nNgudoram\nwas engulfed by destructive fire.\nFalse Colour Banding in the South China Sea\nRepeating the same process in varying satellite banding can produce effective results for monitoring different areas where new infrastructure is identified.\nChina’s island military expansion in the South China Sea can be evidenced through the manner of time-lapse imagery. Looking at\nMischief Reef\n, which was once a deserted island, and fusing the time-lapse with false colour banding, the island’s new infrastructure can be viewed.\nFurther scrutiny of satellite imagery, such as using moisture index banding, can be applied. This identifies the level of moisture through the satellite banding.\nIn this time-lapse above, of Mischief Reef, warmer temperature colours indicate less moisture. The specific areas identified are the tarmac, and one building in the middle of the island — likely to be warm.\nOpen-Source Remote Sensing Tool Available for Public Use\nThe ability to view a satellite imagery time-lapse and identify changes taking place on the Earth’s surface is free to use by anyone with a computer and an internet connection.\nUsing the methods identified in this guide allows anyone to monitor unreachable areas with an analytical approach and identify infrastructure destruction/destructive activities, changes and developments in infrastructure, or anything else, for that matter, on the planet.\nBellingcat’s research\nfor\nthis publication was supported by\nPAX\nfor\nPeace.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/3.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/4.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/5.gif",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/6.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/7.gif",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/8.gif",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/9.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/10.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/11.gif",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/12.gif",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/13.gif",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/14.gif",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/01/pax-logo-web-fc-900-300x141.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2018-11-21_geolocation-infrastructure-destruction-cameroon-case-study-kumbo-kumfutu.json b/bellingcat_geolocation_json/2018-11-21_geolocation-infrastructure-destruction-cameroon-case-study-kumbo-kumfutu.json
new file mode 100644
index 0000000000000000000000000000000000000000..cc271d989e09049947e4859ff1d4843578a18b1f
--- /dev/null
+++ b/bellingcat_geolocation_json/2018-11-21_geolocation-infrastructure-destruction-cameroon-case-study-kumbo-kumfutu.json
@@ -0,0 +1,137 @@
+{
+ "slug": "geolocation-infrastructure-destruction-cameroon-case-study-kumbo-kumfutu",
+ "title": "Geolocation of Infrastructure Destruction in Cameroon: A Case Study of Kumbo and Kumfutu",
+ "url": "https://www.bellingcat.com/resources/case-studies/2018/11/21/geolocation-infrastructure-destruction-cameroon-case-study-kumbo-kumfutu/",
+ "date": "2018-11-21T23:41:12+00:00",
+ "modified": "2019-09-02T13:19:11+00:00",
+ "author": "Benjamin Strick",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)Geolocation is crucial in open source investigations. In countries in conflict and/or countries under dictatorship, verifying evidence on the ground can be difficult. Video and photographic materials make it possible to have a working knowledge of events occurring out of reach — and it is always important to geolocate such materials. Cameroon, for […]",
+ "content_html": "
Geolocation is crucial in open source investigations. In countries in conflict and/or countries under dictatorship, verifying evidence on the ground can be difficult. Video and photographic materials make it possible to have a working knowledge of events occurring out of reach — and it is always important to geolocate such materials.
\n
Cameroon, for example, has a history of government-led attacks on villages. Fighting between military and separatists has seen many Cameroonians displaced or killed. The BBC showed in an open source investigation evidence of villages being burned to the ground by government forces earlier this year.
\n
Footage from various Cameroonian hotspots, namely Cameroon’s anglophone region, show killings, further destruction of villages, and acts of violence towards civilians caught in the middle of fighting.
\n
While we can’t right the wrongs committed against people on the ground, open source investigators are able to analyse the video content to assist reporting and verification of such violent, destructive events.
\n
In November of this year, two videos of attacks on Cameroonian villages surfaced on the internet. They showed clear signs of deliberate arson, as there was no indication of surrounding vegetation burning — just individual buildings ablaze. One of the videos, according to its narrator, showed compounds where people were trapped and burned inside.
\n
Both videos were narrated by the people doing the actual filming, as a means of explaining what the video authors were seeing. The case study below explains how we geolocated this grim footage.
\n
Video of Burning Buildings in Kumbo, Cameroon
\n
The first video was filmed in Kumbo, Cameroon, at this location.
\n
The video, as seen below, was provided by investigative journalist Emmanuel Freudenthal, who was been documenting ongoing violence in Cameroon.
\n
\n
Great example of how geolocation and crowdsourcing works to assist investigations. @EmmanuelFreuden needed a video geolocated of buildings that had been burnt – hints were it was Kumbo in Cameroon. The following is how it was found. #Geolocation#OSINT Thread 1/? pic.twitter.com/uAHamh4fNe
Freudenthal approached both Bellingcat and independent open source analysts for assistance in verifying the video. The one clue he immediately had as to the location was that it was filmed recently in the city of Kumbo, Cameroon.
\n
For many, knowing that the video was filmed in this specific area would be enough. But for a thorough verification, and to help with investigating what occurred, a geolocation of the video is a priority task.
\n
The key markers that stood out in the video, those that would reflect on satellite imagery, are as follows.
\n
This long red roof next to a road:
\n
\n
These shops and an unfinished building on the other side of the road from the red roof:
\n
\n
After an initial scan through Google Earth, it was clear that with a long red building on one side of the road and shops on the other side, this could be the exact location, as found by open source analyst Sector035.
\n
\n
To further verify the location as an exact match, we checked out other clues in the video. For instance, the name of one of the stores seen in the video is “GPS Bookshop.”
\n
\n
Performing a simple Google Search for “GPS Bookshop” in Kumbo does not yield definitive results, but it does appear that the business is listed on a business registry site.
\n
The registry does not provide much detail, and the only other attached information is “Squars.” In viewing other listings, we conclude that Squars must refer to a geographic location.
\n
\n
With open source analysis, an important rule of thumb is to “test” information just to see what can come up. In the case of “Squars” in Kumbo, punching it into Google provides the autocorrected search version for Kumbo Squares — a specific area in Kumbo.
\n
From those results, some YouTube videos are listed at the top. The thumbnail of one of those yields an interesting result — the long red building from the original video.
\n
\n
In this video, we can see a type of dance being performed in front of the long red building. Here is a comparison between the building we saw in the video of the torched ruins, and the one we have found via a Google search.
\n
\n
It is apparent that the two buildings are very similar in design, with oval archways, a unique entrance facade, red roof, and black and white stones.
\n
\n
In the second YouTube video from a simple Google search, we can see the same costumed dancer in the same area, but this time from an alternative view — where we can see the other side of the street.
\n
The businesses in the backdrop were filmed with this view of the long red building:
\n
\n
This means that in the YouTube video, the view from the businesses was parallel to the long red building here:
\n
\n
In the video of the burned down businesses, we were shown the long red building from this viewpoint:
\n
\n
Which means the businesses that were destroyed were under this roof:
\n
\n
To further establish this location, we have these features from that exact viewpoint:
\n
\n
We can verify their presence on Google Earth from here:
\n
\n
Since I posted these findings on Twitter, a number of Cameroonians have come forward with video and photo evidence of damage to nearby buildings. It is clear that the torching of these stores was not an isolated instance.
\n
Geolocating a Video from Kumfutu, a Village Burned Down in Cameroon
\n
Another video that surfaced recently was that of the village of Kumfutu, also in Cameroon. Kumfutu is 61km from Kumbo, where the previous video was filmed.
\n
While the previous video did not include evidence of human casualties, the video from Kumfutu came with disturbing photos of charred bodies and injured villagers.
\n
The long version of the second video was a documentation of many buildings that had been destroyed. The narrator talks about what was done to each building. The narrator also says that one of the compounds had human remains in it. At the end of the second video, we see a family whose members had been left homeless.
\n
Here is the video, which was uploaded to a public Cameroonian social group on Facebook (click on embedded tweet to view):
Along with the post came two images of unique building designs:
\n
\n
\n
Both have noticeable archway entrances, distinct roofing, and were allegedly located near each other.
\n
In the video, these buildings can be matched to the following scenes:
\n
\n
We can confirm that they are next to each other from the video, as the video author walks up a small hill to film the second house, and proceeds to again show the roof of the first here, in this scene:
\n
\n
Because of the unique design of the first house, and the fact that it has a shed on the left and a long shelter at the rear, it was easy to find a match on Google satellite imagery. The red-roofed house next door is also visible on the map.
\n
\n
Open source analyst Jura Lugaverified the other buildings that were destroyed in the video.
Having taken into consideration the locations of the buildings, and with the differences between the buildings (pre and post-destruction) on available satellite imagery from Digital Globe, open source analyst Nathan Ruser demonstrated that these villages had been destroyed in Kumfutu between October 26 and November 11, 2018 (as seen below):
\n
\n
In Kumfutu, the houses highlighted in red were destroyed or damaged in fire between Oct 26th and Nov 11th. pic.twitter.com/DHXZCPznfj
The red indicates buildings which were either completely, or partially, destroyed in the near-complete destruction of Kumfutu.
\n
Context to Cameroonian Village Burnings
\n
Previous open source investigations have shown that the Cameroon military has systematically burned villages to the ground in Cameroon’s anglophone region, as you can see in the video Cameroon Burning.
\n
Further proof of atrocities committed by the military can be seen in another open source investigation on Cameroonian extra-judicial killings, which also involved many of the analysts and investigators in this case study.
\n
Bellingcat’s research for this publication was supported by PAXforPeace.
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nGeolocation is crucial in open source investigations. In countries in conflict and/or countries under dictatorship, verifying evidence on the ground can be difficult. Video and photographic materials make it possible to have a working knowledge of events occurring out of reach — and it is always important to geolocate such materials.\nCameroon, for example, has a history of government-led attacks on villages. Fighting between military and separatists has seen many Cameroonians displaced or killed. The\nBBC\nshowed in an open source investigation evidence of villages being burned to the ground by government forces earlier this year.\nFootage from various Cameroonian hotspots, namely Cameroon’s anglophone region, show killings, further destruction of villages, and acts of violence towards civilians caught in the middle of fighting.\nWhile we can’t right the wrongs committed against people on the ground, open source investigators are able to analyse the video content to assist reporting and verification of such violent, destructive events.\nIn November of this year, two videos of attacks on Cameroonian villages surfaced on the internet. They showed clear signs of deliberate arson, as there was no indication of surrounding vegetation burning — just individual buildings ablaze. One of the videos, according to its narrator, showed compounds where people were trapped and burned inside.\nBoth videos were narrated by the people doing the actual filming, as a means of explaining what the video authors were seeing. The case study below explains how we geolocated this grim footage.\nVideo of Burning Buildings in Kumbo, Cameroon\nThe first video was filmed in Kumbo, Cameroon, at\nthis location\n.\nThe video, as seen below, was provided by investigative journalist\nEmmanuel Freudenthal\n, who was been documenting ongoing violence in Cameroon.\nGreat example of how geolocation and crowdsourcing works to assist investigations.\n@EmmanuelFreuden\nneeded a video geolocated of buildings that had been burnt – hints were it was Kumbo in Cameroon. The following is how it was found.\n#Geolocation\n#OSINT\nThread 1/?\npic.twitter.com/uAHamh4fNe\n— Benjamin Strick (@BenDoBrown)\nNovember 16, 2018\nFreudenthal approached both Bellingcat and independent open source analysts for assistance in verifying the video. The one clue he immediately had as to the location was that it was filmed recently in the city of Kumbo, Cameroon.\nFor many, knowing that the video was filmed in this specific area would be enough. But for a thorough verification, and to help with investigating what occurred, a geolocation of the video is a priority task.\nThe key markers that stood out in the video, those that would reflect on satellite imagery, are as follows.\nThis long red roof next to a road:\nThese shops and an unfinished building on the other side of the road from the red roof:\nAfter an initial scan through Google Earth, it was clear that with a long red building on one side of the road and shops on the other side, this could be\nthe exact location\n, as found by open source analyst\nSector035\n.\nTo further verify the location as an exact match, we checked out other clues in the video. For instance, the name of one of the stores seen in the video is “GPS Bookshop.”\nPerforming a simple\nGoogle Search\nfor “GPS Bookshop” in Kumbo does not yield definitive results, but it does appear that the business is listed on a\nbusiness registry site\n.\nThe registry does not provide much detail, and the only other attached information is “Squars.” In viewing other listings, we conclude that Squars must refer to a geographic location.\nWith open source analysis, an important rule of thumb is to “test” information just to see what can come up. In the case of “Squars” in Kumbo, punching it into Google provides the\nautocorrected search version\nfor Kumbo Squares — a specific area in Kumbo.\nFrom those results, some YouTube videos are listed at the top. The thumbnail of one of those yields an interesting result — the long red building from the original video.\nIn this\nvideo\n, we can see a type of dance being performed in front of the long red building. Here is a comparison between the building we saw in the video of the torched ruins, and the one we have found via a Google search.\nIt is apparent that the two buildings are very similar in design, with oval archways, a unique entrance facade, red roof, and black and white stones.\nIn the second\nYouTube video\nfrom a simple Google search, we can see the same costumed dancer in the same area, but this time from an alternative view — where we can see the other side of the street.\nThe businesses in the backdrop were filmed with this view of the long red building:\nThis means that in the YouTube video, the view from the businesses was parallel to the long red building here:\nIn the video of the burned down businesses, we were shown the long red building from this viewpoint:\nWhich means the businesses that were destroyed were under this roof:\nTo further establish this location, we have these features from that exact viewpoint:\nWe can verify their presence on\nGoogle Earth\nfrom here:\nSince I posted these findings on\nTwitter\n, a number of Cameroonians have come forward with video and photo evidence of damage to nearby buildings. It is clear that the torching of these stores was not an isolated instance.\nGeolocating a Video from Kumfutu, a Village Burned Down in Cameroon\nAnother video that surfaced recently was that of the village of Kumfutu, also in Cameroon. Kumfutu is 61km from Kumbo, where the previous video was filmed.\nWhile the previous video did not include evidence of human casualties, the video from Kumfutu came with disturbing photos of charred bodies and injured villagers.\nThe long version of the second video was a documentation of many buildings that had been destroyed. The narrator talks about what was done to each building. The narrator also says that one of the compounds had human remains in it. At the end of the second video, we see a family whose members had been left homeless.\nHere is the video, which was uploaded to a public Cameroonian social group on\nFacebook\n(click on embedded tweet to view):\n11/ Thanks to open source investigator\n@y_vdw\n(who worked on the\n#BBCAfricaEye\nCameroon investigation) we have found out about another devastating event of a village burning in\n#Cameroon\n– this time in Kumfutu (\nhttps://t.co/HUSuvra4M5\n) Evidence is in the video ?\npic.twitter.com/M1LeeAXwAa\n— Benjamin Strick (@BenDoBrown)\nNovember 19, 2018\nAlong with the post came two images of unique building designs:\nBoth have noticeable archway entrances, distinct roofing, and were allegedly located near each other.\nIn the video, these buildings can be matched to the following scenes:\nWe can confirm that they are next to each other from the video, as the video author walks up a small hill to film the second house, and proceeds to again show the roof of the first here, in this scene:\nBecause of the unique design of the first house, and the fact that it has a shed on the left and a long shelter at the rear, it was easy to find a match on\nGoogle\nsatellite imagery. The red-roofed house next door is also visible on the map.\nOpen source analyst\nJura Luga\nverified\nthe other buildings that were destroyed in the video.\nThese destroyed buildings were shown here:\nThey can be geolocated to here on\nGoogle Maps\n:\nJura also verified buildings that were destroyed in Kumfutu\nhere\nand\nhere\n.\nIn the video, the narrator said people were trapped in the burned buildings shown here:\nThese buildings are located here on\nGoogle Maps\n:\nHaving taken into consideration the locations of the buildings, and with the differences between the buildings (pre and post-destruction) on available satellite imagery from Digital Globe, open source analyst\nNathan Ruser\ndemonstrated that these\nvillages had been destroyed in Kumfutu\nbetween October 26 and November 11, 2018 (as seen below):\nIn Kumfutu, the houses highlighted in red were destroyed or damaged in fire between Oct 26th and Nov 11th.\npic.twitter.com/DHXZCPznfj\n— Nathan Ruser (@Nrg8000)\nNovember 20, 2018\nThe red indicates buildings which were either completely, or partially, destroyed in the near-complete destruction of Kumfutu.\nContext to Cameroonian Village Burnings\nPrevious open source investigations have shown that the Cameroon military has systematically burned villages to the ground in Cameroon’s anglophone region, as you can see in the video\nCameroon Burning\n.\nFurther proof of atrocities committed by the military can be seen in another\nopen source investigation\non Cameroonian extra-judicial killings, which also involved many of the analysts and investigators in this case study.\nBellingcat’s research\nfor\nthis publication was supported by\nPAX\nfor\nPeace.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/1-7-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/2-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/3-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/4-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/5-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/6-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/8-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/9-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/10-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/11-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/12-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/13-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/14-CAMEROON.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/15-CAMEROON.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/16-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/17-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/18-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/19-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/20-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/21-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/22-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/23-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/11/24-CAMEROON.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2018/01/pax-logo-web-fc-900-300x141.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2019-02-07_wheres-babis-geolocating-the-alleged-father-son-kidnapping-mystery.json b/bellingcat_geolocation_json/2019-02-07_wheres-babis-geolocating-the-alleged-father-son-kidnapping-mystery.json
new file mode 100644
index 0000000000000000000000000000000000000000..e056c711a8acbf51d0525daf0e077ed870a91cb0
--- /dev/null
+++ b/bellingcat_geolocation_json/2019-02-07_wheres-babis-geolocating-the-alleged-father-son-kidnapping-mystery.json
@@ -0,0 +1,247 @@
+{
+ "slug": "wheres-babis-geolocating-the-alleged-father-son-kidnapping-mystery",
+ "title": "Where’s Babiš? Geolocating the Alleged Father-Son Kidnapping Mystery",
+ "url": "https://www.bellingcat.com/news/2019/02/07/wheres-babis-geolocating-the-alleged-father-son-kidnapping-mystery/",
+ "date": "2019-02-07T11:00:54+00:00",
+ "modified": "2019-02-09T20:59:40+00:00",
+ "author": "Christiaan Triebert",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)This is a joint publication with Czech medium Houpací Osel. In November 2018, Andrej Babiš, the Prime Minister (PM) of the Czech Republic, came under heavy pressure as thousands of protestors took the streets and demanded his resignation. Days earlier, on November 13, 2018, the Czech media outlet Seznam Zprávy interviewed Andrej Babiš […]",
+ "content_html": "
This is a joint publication with Czech medium Houpací Osel.
\n
\n
In November 2018, Andrej Babiš, the Prime Minister (PM) of the Czech Republic, came under heavy pressure as thousands of protestors took the streets and demanded his resignation.
\n
Days earlier, on November 13, 2018, the Czech media outlet Seznam Zprávy interviewed Andrej Babiš Jr., the eldest son of the Czech PM, who claimed that he had been kidnapped by an associate of his father and held outside of the country to prevent him giving evidence in a corruption case.
\n
Mr. Babiš and two of his children — including Andrej Jr. — are accused of illegally obtaining two million euros of European Union small-business subsidy for a luxury resort and farm known as the “Stork’s Nest” (Čapí Hnízdo in Czech), located around 40 kilometers south of Prague. It is suspected that the Czech PM concealed his ownership of the 4-star hotel.
\n
\n
In response to the extraordinary claims of his son, the Czech PM said that Babiš Jr. had been diagnosed with schizophrenia in 2015. He accused the Seznam Zprávy reporters of exploiting his son’s mental illness when they tracked him down to Switzerland and interviewed him. “No one kidnapped my son. He travelled from the Czech Republic voluntarily,” Mr. Babiš wrote on Facebook.
\n
\n
To support the claim that his son was not held against his will, the Czech PM published several photos of his son, allegedly vacationing in Russian-occupied Crimea.
\n
\n
While Babiš’s coalition government survived a no-confidence vote in November 2018, it is still unclear whether the claims of Mr. Babiš Jr. are factual, and what the full context of his alleged presence in Russian-annexed Crimea is. On top of that, there have been a variety of wild theories about the origins of the images circulating on Czech social media. Was it Photoshop? Was it manipulated in an other way?
\n
For this reason, Houpací Osel contacted Bellingcat to see if the photos presented as evidence by Babiš can shed a bit more light on the claims. To answer these questions, we decided to crowdsource the geolocation (“Where?”) and chronolocation (“When?”) of these photos via Twitter.
\n
Before we turn to the results of the crowdsourcing campaign, let’s have a look at the specific allegations and statements made in what is by far the most talked-about event in the Czech Republic in 2018 — and perhaps 2019, too.
\n
Allegations and Statements: What We Know
\n
The Son: Andrej Babiš Junior
\n
Andrej Babiš Jr. as seen in the Seznam Zprávy report.
\n
Andrej Babiš Jr. (born in 1982, currently 35 years old) was tracked down to his current home in Switzerland by Seznam Zprávy, something the Czech police didn’t manage to do in two years, the reporters wrote. It’s worth mentioning that Czech Radio (Český Rozhlas) later reported that, according to the public prosecutor Jaroslav Šaroch, the Czech police investigative team was told the location of Andrej Babiš Jr. by the defence when the prosecution’s case was launched.
\n
Babiš Jr. told the reporters that he was given the choice of “taking an extended holiday” in Crimea, or being locked up in a mental institution. He chose the former, he says, and claimed to have been held in Russian-annexed Crimea (i.e. Russian-occupied Ukraine), specifically in rented apartments in Sevastopol and Yalta, as well as in Kaliningrad (Russia), Moscow (Russia), and Kryvyi Rih (Ukraine), thereby preventing him from testifying in the “Stork’s Nest” affair.
\n
In an email to Czech reporters, Mr Babiš said that he was held against his will “for a long time” in Crimea, but was later able to travel to Slovakia and “eventually escaped to Switzerland by bus.” He did not mention specific dates.
Dubbed the “Czech Donald Trump”, Andrej Babiš is the country’s second-richest person (number 456 on Forbes), who became PM in December 2017. The 64-year-old billionaire has repeatedly denied any wrongdoing and dismissed the claims of his son, which he attributed to “pressure by journalists on law enforcement authorities”.
\n
He said his son was living with his mother in Switzerland, and would require permanent supervision due to his mental illness. It is reported that Babiš Jr. has dual Czech and Swiss citizenship.
\n
The Caretaker: Petr Protopopov
\n
\n
The associate who allegedly kidnapped Babiš Jr. was a man working as a driver at his father’s company, Agrofert, Babiš Jr. told the reporters. He identified the man as “Petr Protopopov,” a citizen of the Russian Federation, and identified his wife “Dita Protopopova”, a psychiatrist who treated him and wrote a medical report that said he cannot testify in the Stork’s Nest case. Babiš Jr. said he trusted both of these individuals, and that this trust was abused by them.
\n
Protopopov was also interviewed by Seznam reporters, as well as Novinky.cz. To the latter, he said he had travelled to Russia and twice to Russian-annexed Crimea with Babiš Jr. in 2017.
\n
In September 2017, Protopopov said spent a few days in Moscow and then continued on to Crimea. He said that the choice for these locations was the language barrier, stating that he can only speak Czech and Russian and was afraid that medical assistance would be needed for Babiš Jr. Protopopov’s wife arranged for Babiš Jr to go to a clinic in Russia as well as Ukraine, he told Novinky.cz, while stating that he does not considers Crimea to be annexed territory: “Whenever we travelled, my wife ensured that even abroad [Russia and Ukraine] he was going to psychiatry.” The newspaper published one picture of Mr. Protopopov.
\n
In November 2017, Protopopov and Babiš Jr.visited the city of Yalta on Crimea, where the former introduced the latter to the woman later identified by Radio Free Europe as Jelizaveta M. from Kryvyi Rih, Ukraine. “My good friend recommended her to me and she knew that [Babiš Jr] is “different”. So I left [Babiš Jr.] there and returned to the Czech Republic,” Protopopov said.
\n
During the interview, Protopopov said that they took plenty of photos of their trips to Russia and Ukraine — photos that Babiš Sr. would later upload to his Facebook, only to delete them again. These photos will be analysed further down in this piece.
Dita Protopopova is the psychiatrist that took over the care of Babiš Jr. in 2015. We should reiterate that she was the author of the medical report which said he cannot testify in the Stork’s Nest case. (After the Seznam Zprávy‘s report, Cyril Höschl, director of the National Institute of Mental Health said that the institute verified the medical report made by Prototopova and confirmed its findings).
\n
A year later, in 2016, she became an advisor to the Minister of Finance — Andrej Babiš. In 2018, she led the ticket from his ANO party but announced she would step down as a deputy in a district of the Czech capital, Prague 8, after the Seznam report was issued, Radio Free Europe reported. She cited “the pressure imposed in the last few hours” on her and her family.
\n
The Czech Police: Nothing To See Here?
\n
The Czech police said on November 13, 2018, that they had investigated Babiš Jr.’s allegation and concluded that there was no suspicion of a crime.
\n
The Visual Evidence
\n
The PM’s Photos
\n
Following the uproar caused by the Seznam report, Mr. Babiš published three of photos his son on his Facebook page as evidence that his son was not kidnapped but visiting those places voluntarily. Three photos show Babiš Jr, and one photo shows Mr. Protopopov.
\n
The three photos that were serving as evidence were removed from the Czech PM’s Facebook page within days. For that reason, Jan Zabka of Czech student outlet Houpaciosel.cz and Jaroslav Valůch, a trainer with the journalism education institution Transitions, reached out to Bellingcat to ask for help with determining where (geolocation) and when (chronolocation) these pictures were taken.
\n
The Protopopov Photo
\n
A single photo of Mr. Protopopov was published by Novinky.cz. It is not clear whether this photo is related to the case, but we added it to the crowdsourcing effort nonetheless.
\n
The Girlfriend’s Photo
\n
Another eightphotos of relevance were posted to the VKontakte (VK) profile of Babiš Jr.’s girlfriend. While the photos have been deleted from the social medium since Czech media reported about it, Czech medium NEOVLIVNI.CZ archived the photos, which will also be analysed.
\n
Geo- and Chronolocating the Images
\n
Facebook Photo 1
\n
\n
Pictured: Andrej Babiš Jr. and Jelizaveta \nGeolocation: 44.612319, 33.489343 (Sevastopol, Russian-annexed Crimea) \nChronolocation: around September (before 15 November 2018)
\n
In the first Facebook photo, we see Babiš Jr. hugging a lady — who has been identified by Radio Free Europeas Jelizaveta from Ukraine’s Kryvyi Rih — on a rocky beach. A skyline is faintly visible in the background.
\n
The picture was geolocated to the Chersonesus preserve in Sevastopol, Crimea, by Haley Willis, an open source researcher at UC Berkeley’s Human Rights Center. Based on the perspective in the photo with some buildings faintly visible in the background, she narrowed it down a few possible locations along the coastline of Sevastopol. Yandex Panorama — the Russian equivalent of Google Streetview — was used to find reference photos that confirmed that it was the same location. In Willis’ own words, it was due to, “an unglamorous amount of searching through panoramic and reference imagery.”
\n
\n
\n
Further reference imagery found via geo-based searches on VKontakte appear to show the same rock formation in the background as well — you can read all about how to conduct geo-based searches on VK and other social media here.
\n
Geolocation (the “where” question) leads to chronolocation (the “when” question), and here’s how: the shadow Andrej Babis Jr. is casting is like a sundial. Based on the exact location, we now that the shadow points almost directly to the east.
\n
Using sun movement services like SunCalc.org, we can see that the shadow cannot point to that direction during January, February, March, as well as October, November and December.
\n
We are thus left with a time period between April and September. Twitter-user @Sector035 was able to narrow down the month-frame even more, by estimating the length of Mr. Babiš Jr. and the length of his shadow.
\n
\n
About chronolocation of Photo 1: There aren't a lot of months at all! The angle of the shadow points roughly towards 17:30 in the afternoon. From April till mid August the shadow would not be long enough to be cast. I could get closer, but I bet on mid September ? pic.twitter.com/Ddoa2cDzUs
Given the direction of the shadow, the photo was taken in the evening hours, likely before sunset.
\n
Maximum temperatures in Sevastopol are relatively cold, around 15 degrees, Celsius in April, and are quite warmer in September — around 22 degrees Celsius. Most people in the photo are wearing T-shirts and shorts. However, this information is not enough to conclusively determine whether the photo was taken near the end of the established period — they could have just not minded the cold, of course.
\n
Unfortunately, no visual clues have been discovered so far to determine which approximate year the photo was taken.
\n
Facebook Photo 2
\n
\n
Pictured: Andrej Babiš Jr. \nGeolocation: 44.613203, 33.518702 (Sevastopol, Russian-annexed Crimea) \nChronolocation: after 4 September 2017 and before 15 November 2018 (the upload date)
\n
The second Facebook photo shows Andrej Babis Jr. sitting on a terrace, apparently smoking shisha, and enjoying a drink in a large lemonade glass.
\n
There are several clues in the photo which allow for geolocation. The most obvious (but relatively useless) clues are the writings on the net and the chair: кафе (Russian for “cafe”) and Балтика (“Baltika”), a leading Russian beer brewery. Such table sets are often featured in cafes and restaurants, making it harder to pinpoint where it may be: the Russian-speaking and Baltika-drinking world is large, and it is the second largest brewing company in Europe. However, it still narrows down the search area to countries where Baltika is consumed.
\n
Some intrepid researchers went on a search to find similar table combinations. For example, Twitter-user Matt McCarthy found a place called “Piccolini” with the exact same chair and table combination, including the warning sign against excessive drinking on the table (thanks @stysin) — in Sevastopol!
\n
\n
However, further examination showed that it was not possible to find a match with the windows of the Piccolino cafe. The method was interesting though, as Matt McCarthy explained: he read that 4square is often used in Ukraine like how Yelp is used in the United States. He searched for “food” in “crimea” and found user submitted photos. He then continued looking for places with outdoor seating and stumbled upon the photo above. The glasses and straws were a hint too.
\n
\n
But there is a less obvious clue in the photo: The reflections in the windows, as mentioned by Twitter-user Egon de Bruin. We see, for example, distinctive buildings and long stripes (three white, one yellow) painted on the asphalt.
\n
Meanwhile, Bellingcat Investigation Team member Daniel Romein found a cafe in Sevastopol with a matching font type for the word “кафе” (Russian for “cafe). Further inspection of the reflection confirms the location: Namely the stripes on the ground and the “Artbukhta” ferry terminal building give it away, as found by Jason Paladino, a journalist with the Investigative Reporting Project at UC Berkeley.
\n
\n
The visual clues in the window reflections lead us to the exact place: Lounge-cafe “Лица.” It can be confirmed with reference photos of the location, such as this one from TripAdvisor, which show the same pillar and table, and thus the exact position where Babiš Jr. was sitting. But when exactly? The combination of tables and chairs holds the answer.
\n
\n
Based on geotagged reference photos from the lounge-cafe, we can see that the table and chair combo changed somewhere between 2016 and 2018. The combination of Baltika chairs and the tables in Babiš Jr.’s photo can be spotted on photos (mostly Instagram and Google Photos) uploaded in the period of June and July 2017. It seems lounge-cafe “Лица” had Warsteiner in 2016 (last photo → “Eine Königin unter den Bieren”).
\n\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n \n
\n
Now that we have a geolocation, we can focus on determining the time-frame in which the photo was taken. Reference photos on Yandex Panorama and 360cities show that the place next to our location of interest was called “Кафе бар АртБухта” (Cafe Bar Artbukhta) a few years ago, but there has been a renovation and a renaming to “Confectionary Hedgehogs” (Confectionary Cafe) — based on Yandex Panorama imagery somewhere in or before 2016.
\n
Satellite imagery is even more helpful in narrowing down the date.
\n
In the window reflection, four lines are visible on the asphalt. Google Earth historical satellite imagery shows that these appeared between 9 August 2017 and 9 September 2017 respectively.
\n
\n
\n
Daria Sazanovich, an artist and illustrator from Minsk, started searches in Russian to narrow down that timeframe by asking when those lines appeared.
\n
Reference photos on a local website Primechaniya show that the striped were not there on 18 August 2017. A local media report from 4 September 2017 — also dug up by Ms. Sazanovich — shows that the stripes were not visible either. The lounge-cafe is clearly visible in the background.
\n
\n
A local blogger writes on Livejournal that the stripes appeared in early September, showing a picture dated 8 September 2017 featuring the stripes, Ms. Sazanovich found.
\n
\n
Based on these reference photos, it is thus possible to say that the photo of Mr. Babis Jr. must have been taken after 4 September 2017, as the stripes were not there yet.
\n
Facebook Photo 3
\n
\n
Pictured: Andrej Babiš Jr. \nGeolocation: 44.614328, 33.521112 (Sevastopol, Russian-annexed Crimea) \nChronolocation: after Spring 2016 and before April 2018
\n
The third photo shows Mr. Babiš Jr. pictured with what appear to be two chinchillas on his shoulders. There are several visual clues.
\n
Behind him, for example, is an advertisement for a new music school (Russian: Севастопольская Музыкальная школа № 1 Имени Николая Андреевича, Римского-корсакова) in Russian-annexed Sevastopol is visible. The advertisement includes an address. Daria Sazanovich found that the opening ceremony for the school was held on June 17, 2018. It’d be unlikely that the advertisement for a new school to be constructed to be still hanging there after the opening, but we cannot be entirely sure. Either way, it is a start.
\n
Now that we know that the photo was likely taken in or around Sevastopol (who would place such an advertisement somewhere else?), we can narrow down the location. On top of that, Twitter user @firstkaffee noticed that an advertisement for ice in Crimea was also visible.
\n
Twitter-user Rytis Alekna suggests that having chinchillas on your shoulders may be a paid tourist attraction near the beach in Sevastopol. Googling and Yandexing in Cyrillic for “chinchillas” and “sevastopol” indeed give results, for example in 2013 article in ForPost, a local newspaper in Sevastopol.
\n
\n
Eventually, Twitter-user @zaska12345 was able to geolocate it to the Kornilova Embankment in Sevastopol, Russian-annexed Crimea. Yandex Panorama has good reference visuals of the place, showing that even the pole and the tree match.
\n
\n
Now that the location is known, time determination can start. Visual clues for the chronolocation are the mural paint (first it was grey, on the photo it is white), the advertisement (the school officially opened on June 17, 2017), and the chinchillas (present in some years going back to at least 2010 and also present in 2013, but may have become illegal over time like in Yalta according to ForPost). These clues have been further examined by @zaska12345, but it has not been possible to further narrow down the time frame.
\n
Taking the mural paint as reference, Twitter user @john_marquee was able to find a reference image from October 2017 showing that the paint on the ledge was there, as well as the advertisement for the music school. The Yandex image does not show the wall painted in white, and is dated early Spring 2016, suggesting the Babiš photo was taken after that.
\n
\n
Another Twitter user, @MCantow, found a clue for the last possible time. The second tree in the background (here in Spring 2016) was cut between October 2017 (tree visible on Google Earth) and April 2018 (tree not visible anymore on DigitalGlobe satellite imagery).
\n
Novinky.cz Photo
\n
\n
Pictured: Petr Protopopov \nGeolocation: 55.709491, 37.542220 (Moscow, Russia) \nChronolocation: unknown
\n
This photo of Petr Protopov was published alongside an interview with him in Czech media outlet Novinky.cz. Though there is no direct relation to the case, we thought it may be useful to geolocate it nonetheless.
\n
The photo can be geolocated to the Sparrow Hills (Vorobyovy Gory) in Moscow, Russia. It is a popular spot for locals and tourists alike, hence many people recognised the view and the famous Luzhniki Stadium in the background.
Pictured: Andrej Babiš Jr. and Jelizaveta \nGeolocation: 46.203528, 6.148187 \nChronolocation: before 5 March 2018
\n
The first photo was uploaded on 4 March 2018 to VK by Jelizaveta, the girlfriend of Babiš Jr. It shows the couple sitting in a restaurant with decorated ornaments, wooden seats, and a plate of oysters.
\n
On Twitter, Mark Reid was quick to notice that it may be a restaurant in Geneva: Café du Centre. Using online reference photos, he and Twitter-users @Sector035 and @cz_smidra were not only able to match the similar plates used for the oysters, but also the lights, ornaments, ceiling, and wooden chairs.
\n
\n
VKontakte Photo 2
\n
\n
Pictured: Andrej Babiš Jr., Jelizaveta, Santa Claus, and unidentified minor \nGeolocation: 47.903308, 33.400988 (Kryvyi Rih, Ukraine) \nChronolocation: before 2 January 2018
\n
The second VK photo was uploaded on January 2, 2018, five days before Babiš Jr. wrote Czech police that he was kidnapped by an associate of his father.
\n
An eagle-eyed Twitter-user “Lara van Gaalen” as well as Bellingcat-reader “Sergio.Beer” spotted that the shape of the window shape and style, as well the objects outside of the windows strongly resemble the location shown in VKontakte Photo 7, which is discussed below. The spacing between the vertical loudspeakers match as well, and so does the back side of the canvas banner.
\n
\n
And picture 2 might be at the same location as picture 7. Trough the window of photo 2 you can see two rectangular shapes (looks like speakers to me). These are also visible in picture 7, above the banner. pic.twitter.com/i2wwJlhQye
Twitter-user ”Abraxas Spa“ and Bellingcat-reader “Sergio.Beer” both noticed that Babiš Jr. and Jelizaveta are apparently also wearing similar clothes, and the lady is wearing the same bag. This suggest that it may have been the same time period as VKontakte Photo 7, which was taken between 3 December 2017 and 14 January 2018.
\n
VKontakte Photo 3
\n
\n
Pictured: Andrej Babiš Jr., Jelizaveta, Santa Claus and Santa’s helper \nGeolocation: 47.905535, 33.396696 (Multiplex Union, Kryvyi Rih, Ukraine) \nChronolocation: unknown (likely between 24 December and 16 January, the dates which encompasses both Christmas in Ukraine and Russia)
\n
The third VK photo shows the couple posing in front of a Christmas tree and a dressed-up Santa Claus (Ded Moroz, or Father Frost, as he is known in many post-Soviet countries) and his helper, the Snow Maiden (Snegurochka), inside what appears to be a building.
\n
With no distinctive visual clues such as a written text, interior photos are usually harder to geolocate. Nevertheless, Twitter-user @Sector035 was able to find a photo on VK that appears to be the exact same spot. However, no location was mentioned.. So how to figure out where it was taken?
\n
\n
Haley Willis also found a reference photo that was indexed as being from Kryvyi Rih in Ukraine. That was @Sector035’s initial hunch as well. Both started looking for shopping malls in the Ukrainian city, looking for red booths with tables as visible on the top left corner of the three photos. Eventually, Ms. Willis found reference photos on Foursquare, and managed to find a matching place: the Multiplex Union in Kryvyi Rih.
\n
\n
Found it! Looks like the Multiplex Union (вулиця Героїв АТО, 30Г, ТРЦ Union, Kryvyi Rih). Saw this first photo going through Foursquare and recognized the banisters. Then found second photo on Google Photos. pic.twitter.com/or4aMkxFxX
Pictured: Andrej Babiš Jr. and Jelizaveta \nGeolocation: 44.498875, 34.167997 (Saturn IMAX, Yalta, Russian-annexed Crimea) \nChronolocation: between late October and early November 2017
\n
This photo was published on VK by Jelizaveta in late October 2017.
\n
A Bellingcat-reader going by the name “Alex” managed to geolocate the photo. Here’s how.
\n
To the right of Mr. Babiš Jr. are two monitors, one vertical and one horizontal. By cropping, transforming, and reverse-image-searching the image, Alex identified the image on the vertical monitor as the film poster for “Black Panther”, an American superhero film that was released in January 2018.
\n
\n
Alex started comparing Black Panther posters in Russian and Ukrainian, and noticed that he white tet under the logo in the bottom of the poster varied slightly in Russian and Ukrainian. The Russian “в кино с 15 февраля” is slightly longer than the Ukrainian “скоро в кіно”. While there may be more variations, it led Alex to have a look around cinemas in Russian-annexed Crimea.
\n
\n
He thought it must be a a “modern” and “trendy” Crimean cinema, given that it looks that way. The air hockey table on the right side may hold a clue as well. Since there are not that many new or renovated cinemas in Crimea, Alex narrowed down to Yalta, as the air hockey tables in cinemas in Sevastopol did not have matching floors or air hockey tables. Eventually, the Saturn IMAX (Russian: Сатурн IMAX) in Yalta was a match, as can be seen below on tworeference photos on Google Maps.
\n
\n
Based on Alex’s geolocation, it is now possible to narrow down the timeframe by looking for geotagged photos of the Saturn IMAX in Yalta, and see when the Halloween corner appeared in the building. Instagram was specifically useful, as many users tagged their photos to the Saturn IMAX. The first photo of the setting appeared on 27 October 2017 and the last one in 1 November 2017. As Halloween is celebrated on the 30th of October, these dates do make sense to have a Halloween decor in the cinema.
\n
The first photo tagged to the Saturn IMAX showing the Halloween decor appeared on Instagram on 28 October 2017.
\n
VKontakte Photo 5 and 6
\n
\n
Pictured: Andrej Babiš Jr. and Jelizaveta \nGeolocation: 46.205829, 6.157553 (Geneva, Switzerland) \nChronolocation: between October 2017 and July 2018
\n
The fifth and sixth photos were uploaded to Jelizaveta’s VK profile in 2018. It shows the couple standing in front of a fountain.
\n
It was publicly known that Babiš Jr. was with Jelizaveta in Geneva, and for that reason it was easier to geolocate these two photos. Houpací Osel’s Jan Zabka simply wrote “fountain Swiss” into Google Images, and the first result was the “Jet d’Eau” in Geneva, Switzerland. Using reference photos, it is then possible to geolocate their exact positions to 46.205829, 6.157553 and 46.207357, 6.156086, respectively.
\n
With regards to chronolocation, the stickers next to couple hold a clue (very similar to the Papadopulos case). When was the sticker there, and when was it not? Using reference photos on a variety of online platforms, it is possible to determine that the photo must have been taken between October 2017 and July 2018.
\n
VKontakte Photo 7
\n
\n
Pictured: Andrej Babiš Jr. and Jelizaveta \nGeolocation: 47.903662, 33.401034 (Kryvyi Rih, Ukraine) \nChronolocation: between 3 December 2017 and 14 January 2018
\n
The seventh VK photo of interest shows the couple in front of a large building.
\n
The banner on the background is the first visual clue, as it allows to do a text-based search on ”бульвар чуд(…) Італійський (…)” on Google which results in a reference photo of the same poster.
\n
In that way, it is possible to figure out the whole name of the show, and the place where it took place: the Kryvorizkyi Circus in Kryvyi Rih, Ukraine. The couple is tanding specifically at 47.903662, 33.401034. It was geolocated by Houpací Osel’s Jan Zabka.
\n
The banner in the background allows chronolocation: the show “Бульвар чудес” (Boulevard of Miracles) only ran from December 16, 2017 to January 14, 2018, the banner reveals. Reference photos show there was a different banner until December 3, 2017.
\n
VKontakte Photo 8
\n
\n
Pictured: Andrej Babiš Jr. and Jelizaveta \nGeolocation: 44.494549, 34.117881 (Yalta, Russian-annexed Crimea) \nChronolocation: unknown
\n
The eighth and last VK photo of interest shows the couple standing in front of a caged giraffe. An orange, real-size statue of a giraffe is also visible in the background.
\n
Some Czech media mentioned that this photo was taken in a zoo in Yalta, Russian-annexed Crimea. Houpací Osel’s Jan Zabka managed to confirm that this is the “Skazka” Zoo (Зоопарк “Сказка”) in Yalta, Crimea, by geolocating the photo to 44.494549, 34.117881.
\n
\n
Reference photos found via Google and TripAdvisor allow to match the details such as the fence, the wall, and the orange giraffe, to confirm the geolocation.
\n
\n
We have not been able to determine the date and time the photo was taken.
\n
Analysis and Conclusion
\n
When the Czech Prime Minister published the three initial photos as a proof to the public that Babiš Jr. did not suffer while being in Crimea, a large discussion about the authenticity of the photos started among Czech citizens. For that very reason, it is important to geo- and chrono-locate the photos. Only then can we establish some facts.
\n
The crowdsourcing campaign resulted in successfully geolocating 12 out of 12 photos, and getting a better sense of time by roughly chronolocating half of them.
\n\n
The results show that Andrej Babis Jr. indeed visited the Ukrainian city of Kryvyi Rih, as well as Sevastopol and Yalta in Russian-annexed Crimea. With regards to Crimea, it is likely that this was indeed in the latter half of 2017. This backs up the claims related to the three initial photos published by the Czech Prime Minister: the “where” and “when” are correct.
\n
The geolocation does not say anything about the status of Babis Jr. when he was there. It does, at least, authenticate Babis Jr. presence in Russian-annexed Crimea in late 2017. This is an important fact in a highly politicised and sensitive case in which there has been a lot of speculation.
\n
It has been three months the first interview with Babiš Jr. was published. There were mass protests against the government, but it survived a no-confidence vote. Babiš Sr. is still the Prime Minister of the Czech Republic, and the turmoil surrounding the alleged kidnapping case has wound down. In fact, a recent survey suggests that the PM’s popularity actually increased as many voters accepted his narrative of “unethical journalists’ hunt for sensation” who harmed his “ill son” and “family as a whole.”
\n
Correction 7 February 2019: An earlier version mentioned that the location of VKontakte Photo 3 had not been found yet. However, it was already geolocated by Haley Willis on Twitter. \n
\n
Update 9 February 2019: A Google MyMaps of the geolocated photos has been added to the concluding section of the article. Furthermore, Bellingcat-readers and Twitter-users have found the location of VKontakte Photo 2, as well as a closer chronolocation of Facebook Photo 1. The article has been updated accordingly. An earlier version of the article also mentioned that there was reference photo of a tree in 2014. This should have been 2018. \n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nThis is a joint publication with Czech medium\nHoupací Osel\n.\nIn November 2018, Andrej Babiš, the Prime Minister (PM) of the Czech Republic, came under heavy pressure as thousands of protestors took the streets and demanded his resignation.\nDays earlier, on November 13, 2018, the Czech media outlet\nSeznam Zprávy\ninterviewed\nAndrej Babiš Jr., the eldest son of the Czech PM, who claimed that he had been kidnapped by an associate of his father and held outside of the country to prevent him giving evidence in a corruption case.\nMr. Babiš and two of his children — including Andrej Jr. — are accused of illegally obtaining two million euros of European Union small-business subsidy for a luxury resort and farm known as the “Stork’s Nest” (\nČapí Hnízdo\nin Czech),\nlocated\naround 40 kilometers south of Prague. It is suspected that the Czech PM concealed his ownership of the 4-star hotel.\nIn response to the extraordinary claims of his son, the Czech PM said that Babiš Jr. had been diagnosed with schizophrenia in 2015. He accused the\nSeznam Zprávy\nreporters of exploiting his son’s mental illness when they tracked him down to Switzerland and interviewed him. “No one kidnapped my son. He travelled from the Czech Republic voluntarily,” Mr. Babiš wrote on\nFacebook\n.\nTo support the claim that his son was not held against his will, the Czech PM published several photos of his son, allegedly vacationing in Russian-occupied Crimea.\nWhile Babiš’s coalition government survived a no-confidence vote in November 2018, it is still unclear whether the claims of Mr. Babiš Jr. are factual, and what the full context of his alleged presence in Russian-annexed Crimea is. On top of that, there have been a variety of wild theories about the origins of the images circulating on Czech social media. Was it Photoshop? Was it manipulated in an other way?\nFor this reason,\nHoupací Osel\ncontacted\nBellingcat\nto see if the photos presented as evidence by Babiš can shed a bit more light on the claims. To answer these questions, we decided to crowdsource the geolocation (“Where?”) and chronolocation (“When?”) of these photos via Twitter.\nBefore we turn to the results of the crowdsourcing campaign, let’s have a look at the specific allegations and statements made in what is by far the most talked-about event in the Czech Republic in 2018 — and perhaps 2019, too.\nAllegations and Statements: What We Know\nThe Son: Andrej Babiš Junior\nAndrej Babiš Jr. as seen in the Seznam Zprávy report.\nAndrej Babiš Jr. (born in 1982, currently 35 years old) was tracked down to his current home in Switzerland by\nSeznam Zprávy\n, something the Czech police didn’t manage to do in two years, the reporters wrote. It’s worth mentioning that Czech Radio (\nČeský Rozhlas\n) later\nreported\nthat, according to the public prosecutor Jaroslav Šaroch, the Czech police investigative team was told the location of Andrej Babiš Jr. by the defence when the prosecution’s case was launched.\nBabiš Jr. told the reporters that he was given the choice of “taking an extended holiday” in Crimea, or being locked up in a mental institution. He chose the former, he says, and claimed to have been held in Russian-annexed Crimea (i.e. Russian-occupied Ukraine), specifically in rented apartments in Sevastopol and Yalta, as well as in Kaliningrad (Russia), Moscow (Russia), and Kryvyi Rih (Ukraine), thereby preventing him from testifying in the “Stork’s Nest” affair.\nIn an email to Czech reporters, Mr Babiš said that he was held against his will “for a long time” in Crimea, but was later able to travel to Slovakia and “eventually escaped to Switzerland by bus.” He did not mention specific dates.\nThe Father: Andrej Babiš Senior\nImage courtesy of “vjkombajn” via\nPixabay\n.\nDubbed the “\nCzech Donald Trump\n”, Andrej Babiš is the country’s second-richest person (\nnumber 456\non\nForbes\n), who became PM in December 2017. The 64-year-old billionaire has repeatedly denied any wrongdoing and dismissed the claims of his son, which he attributed to “pressure by journalists on law enforcement authorities”.\nHe said his son was living with his mother in Switzerland, and would require permanent supervision due to his mental illness. It is reported that Babiš Jr. has dual Czech and Swiss citizenship.\nThe Caretaker: Petr Protopopov\nThe associate who allegedly kidnapped Babiš Jr. was a man working as a driver at his father’s company, Agrofert, Babiš Jr. told the reporters. He identified the man as “Petr Protopopov,” a citizen of the Russian Federation, and identified his wife “Dita Protopopova”, a psychiatrist who treated him and wrote a medical report that said he cannot testify in the Stork’s Nest case. Babiš Jr. said he trusted both of these individuals, and that this trust was abused by them.\nProtopopov was also interviewed by\nSeznam\nreporters, as well as\nNovinky.cz\n. To the latter, he said he had travelled to Russia and twice to Russian-annexed Crimea with Babiš Jr. in 2017.\nIn September 2017, Protopopov said spent a few days in Moscow and then continued on to Crimea. He said that the choice for these locations was the language barrier, stating that he can only speak Czech and Russian and was afraid that medical assistance would be needed for Babiš Jr. Protopopov’s wife arranged for Babiš Jr to go to a clinic in Russia as well as Ukraine, he told\nNovinky.cz\n, while stating that he does not considers Crimea to be annexed territory: “Whenever we travelled, my wife ensured that even abroad [Russia and Ukraine] he was going to psychiatry.” The newspaper published one picture of Mr. Protopopov.\nIn November 2017, Protopopov and Babiš Jr.\nvisited the city of Yalta on Crimea, where the former introduced the latter to the woman later identified by\nRadio Free Europe\nas Jelizaveta M. from Kryvyi Rih, Ukraine. “My good friend recommended her to me and she knew that [Babiš Jr] is “different”. So I left [Babiš Jr.] there and returned to the Czech Republic,” Protopopov said.\nDuring the interview, Protopopov said that they took plenty of photos of their trips to Russia and Ukraine — photos that Babiš Sr. would later upload to his Facebook, only to delete them again. These photos will be analysed further down in this piece.\nThe Psychiatrist: Dita Protopopova\nImage via\nBlesk.cz\n.\nDita Protopopova is the psychiatrist that took over the care of Babiš Jr. in 2015. We should reiterate that she was the author of the medical report which said he cannot testify in the Stork’s Nest case. (After the\nSeznam Zprávy\n‘s report, Cyril Höschl, director of the National Institute of Mental Health said that the institute verified the medical report made by Prototopova and\nconfirmed\nits findings).\nA year later, in 2016,\nshe became an advisor\nto the Minister of Finance — Andrej Babiš. In 2018, she led the ticket from his ANO party but announced she would step down as a deputy in a district of the Czech capital, Prague 8, after the\nSeznam\nreport was issued,\nRadio Free Europe\nreported\n. She cited “the pressure imposed in the last few hours” on her and her family.\nThe Czech Police: Nothing To See Here?\nThe Czech police said on November 13, 2018, that they had investigated Babiš Jr.’s allegation and concluded that there was no suspicion of a crime.\nThe Visual Evidence\nThe PM’s Photos\nFollowing the uproar caused by the\nSeznam\nreport, Mr. Babiš published three of photos his son on his Facebook page as evidence that his son was not kidnapped but visiting those places voluntarily. Three photos show Babiš Jr, and one photo shows Mr. Protopopov.\nThe three photos that were serving as evidence were removed from the Czech PM’s Facebook page within days. For that reason, Jan Zabka of Czech student outlet\nHoupaciosel.cz\nand Jaroslav Valůch, a trainer with the journalism education institution\nTransitions\n, reached out to\nBellingcat\nto ask for help with determining where (geolocation) and when (chronolocation) these pictures were taken.\nThe Protopopov Photo\nA single photo of Mr. Protopopov was published by\nNovinky.cz\n. It is not clear whether this photo is related to the case, but we added it to the crowdsourcing effort nonetheless.\nThe Girlfriend’s Photo\nAnother eight\nphotos of relevance were posted to the VKontakte (VK) profile of Babiš Jr.’s girlfriend. While the photos have been deleted from the social medium since Czech media reported about it, Czech medium NEOVLIVNI.CZ archived the photos, which will also be analysed.\nGeo- and Chronolocating the Images\nFacebook Photo 1\nPictured:\nAndrej Babiš Jr. and Jelizaveta\nGeolocation:\n44.612319, 33.489343 (Sevastopol, Russian-annexed Crimea)\nChronolocation:\naround September (before 15 November 2018)\nIn the first Facebook photo, we see Babiš Jr. hugging a lady — who has been\nidentified by\nRadio Free Europe\nas Jelizaveta from Ukraine’s Kryvyi Rih — on a rocky beach. A skyline is faintly visible in the background.\nThe picture\nwas geolocated\nto the Chersonesus preserve in Sevastopol, Crimea, by Haley Willis, an open source researcher at UC Berkeley’s Human Rights Center. Based on the perspective in the photo with some buildings faintly visible in the background, she narrowed it down a few possible locations along the coastline of Sevastopol. Yandex Panorama — the Russian equivalent of Google Streetview — was used to find reference photos that confirmed that it was the same location. In Willis’ own words, it was due to, “an unglamorous amount of searching through panoramic and reference imagery.”\nFurther reference imagery found via geo-based searches on VKontakte appear to show the same rock formation in the background as well — you can read all about how to conduct geo-based searches on VK and other social media\nhere\n.\nGeolocation (the “where” question) leads to chronolocation (the “when” question), and here’s how:\nthe shadow Andrej Babis Jr. is casting is like a sundial. Based on the exact location, we now that the shadow points almost directly to the east.\nUsing sun movement services like\nSunCalc.org\n, we can see that the shadow cannot point to that direction during\nJanuary\n,\nFebruary\n,\nMarch\n, as well as October, November and December.\nWe are thus left with a time period between April and September. Twitter-user @Sector035 was able to narrow down the month-frame even more, by estimating the length of Mr. Babiš Jr. and the length of his shadow.\nAbout chronolocation of Photo 1: There aren't a lot of months at all! The angle of the shadow points roughly towards 17:30 in the afternoon. From April till mid August the shadow would not be long enough to be cast. I could get closer, but I bet on mid September ?\npic.twitter.com/Ddoa2cDzUs\n— Sector035 (@Sector035)\nFebruary 7, 2019\nGiven the direction of the shadow, the photo was taken in the evening hours, likely before sunset.\nMaximum temperatures in Sevastopol are relatively cold,\naround 15 degrees\n, Celsius in April, and are quite warmer in September — around 22 degrees Celsius. Most people in the photo are wearing T-shirts and shorts. However, this information is not enough to conclusively determine whether the photo was taken near the end of the established period — they could have just not minded the cold, of course.\nUnfortunately, no visual clues have been discovered so far to determine which approximate year the photo was taken.\nFacebook Photo 2\nPictured:\nAndrej Babiš Jr.\nGeolocation:\n44.613203, 33.518702 (Sevastopol, Russian-annexed Crimea)\nChronolocation:\nafter 4 September 2017 and before 15 November 2018 (the upload date)\nThe second Facebook photo shows Andrej Babis Jr. sitting on a terrace, apparently smoking shisha, and enjoying a drink in a large lemonade glass.\nThere are several clues in the photo which allow for geolocation. The most obvious (but relatively useless) clues are the writings on the net and the chair: кафе (Russian for “cafe”) and Балтика (“Baltika”), a leading Russian beer brewery. Such table sets are often featured in cafes and restaurants, making it harder to pinpoint where it may be: the Russian-speaking and Baltika-drinking world is large, and it is the second largest brewing company in Europe. However, it still narrows down the search area to countries where Baltika is consumed.\nSome intrepid researchers went on a search to find similar table combinations. For example, Twitter-user Matt McCarthy\nfound a place\ncalled “Piccolini” with the exact same chair and table combination, including the warning sign against excessive drinking on the table (\nthanks @stysin\n) — in Sevastopol!\nHowever, further examination showed that it was not possible to find a match with the windows of the Piccolino cafe. The method was interesting though, as Matt McCarthy\nexplained\n: he read that 4square is often used in Ukraine like how Yelp is used in the United States. He searched for “food” in “crimea” and found user submitted photos. He then continued looking for places with outdoor seating and stumbled upon the photo above. The glasses and straws were a hint too.\nBut there is a less obvious clue in the photo: The reflections in the windows, as\nmentioned\nby Twitter-user Egon de Bruin. We see, for example, distinctive buildings and long stripes (three white, one yellow) painted on the asphalt.\nMeanwhile, Bellingcat Investigation Team member Daniel Romein found a cafe in Sevastopol with a matching font type for the word “кафе” (Russian for “cafe). Further inspection of the reflection confirms the location: Namely the stripes on the ground and the “Artbukhta” ferry terminal building give it away, as\nfound\nby Jason Paladino, a journalist with the Investigative Reporting Project at UC Berkeley.\nThe visual clues in the window reflections lead us to the exact place: Lounge-cafe “Лица.” It can be confirmed with reference photos of the location, such as this one from\nTripAdvisor\n, which show the same pillar and table, and thus the exact position where Babiš Jr. was sitting. But when exactly? The combination of tables and chairs holds the answer.\nBased on geotagged reference photos from the lounge-cafe, we can see that the table and chair combo changed somewhere between 2016 and 2018. The combination of Baltika chairs and the tables in Babiš Jr.’s photo can be spotted on photos (mostly Instagram and Google Photos) uploaded in the period of June and July 2017. It seems lounge-cafe “Лица” had Warsteiner in 2016 (last photo → “Eine Königin unter den Bieren”).\nNow that we have a geolocation, we can focus on determining the time-frame in which the photo was taken. Reference photos on\nYandex Panorama and 360cities\nshow that the place next to our location of interest was called “Кафе бар АртБухта” (Cafe Bar Artbukhta) a few years ago, but there has been a renovation and a renaming to “Confectionary Hedgehogs” (Confectionary Cafe) — based on Yandex Panorama imagery somewhere in or before 2016.\nSatellite imagery is even more helpful in narrowing down the date.\nIn the window reflection, four lines are visible on the asphalt. Google Earth historical satellite imagery shows that these appeared between 9 August 2017 and 9 September 2017 respectively.\nDaria Sazanovich, an artist and illustrator from Minsk, started searches in Russian to narrow down that timeframe by asking when those lines appeared.\nReference photos on a local website\nPrimechaniya\nshow that the striped were not there on 18 August 2017. A local media report from 4 September 2017 — also dug up by Ms. Sazanovich — shows that the stripes were not visible either. The lounge-cafe is clearly visible in the background.\nA local blogger\nwrites on\nLivejournal\nthat the stripes appeared in early September, showing a picture dated 8 September 2017 featuring the stripes, Ms. Sazanovich found.\nBased on these reference photos, it is thus possible to say that the photo of Mr. Babis Jr. must have been taken after 4 September 2017, as the stripes were not there yet.\nFacebook Photo 3\nPictured:\nAndrej Babiš Jr.\nGeolocation:\n44.614328, 33.521112 (Sevastopol, Russian-annexed Crimea)\nChronolocation:\nafter Spring 2016 and before April 2018\nThe third photo shows Mr. Babiš Jr. pictured with what appear to be two chinchillas on his shoulders. There are several visual clues.\nBehind him, for example, is an advertisement for a new music school (Russian: Севастопольская Музыкальная школа № 1 Имени Николая Андреевича, Римского-корсакова) in Russian-annexed Sevastopol is visible. The advertisement includes an address. Daria Sazanovich found that\nthe opening ceremony for the school\nwas held on June 17, 2018. It’d be unlikely that the advertisement for a new school to be constructed to be still hanging there after the opening, but we cannot be entirely sure. Either way, it is a start.\nNow that we know that the photo was likely taken in or around Sevastopol (who would place such an advertisement somewhere else?), we can narrow down the location. On top of that, Twitter user @firstkaffee\nnoticed\nthat an advertisement for ice in Crimea was also visible.\nTwitter-user Rytis Alekna\nsuggests\nthat having chinchillas on your shoulders may be a paid tourist attraction near the beach in Sevastopol. Googling and Yandexing in Cyrillic for “chinchillas” and “sevastopol” indeed give results,\nfor example in 2013 article in\nForPost\n, a local newspaper in Sevastopol.\nEventually, Twitter-user @zaska12345 was able to geolocate it to the Kornilova Embankment in Sevastopol, Russian-annexed Crimea. Yandex Panorama\nhas good reference visuals\nof the place, showing that even the pole and the tree match.\nNow that the location is known, time determination can start. Visual clues for the chronolocation are the mural paint (first it was grey, on the photo it is white), the advertisement (the school officially opened on June 17, 2017), and the chinchillas (present in some years going back to at least 2010 and also present in 2013, but may have become illegal over time like in Yalta\naccording to\nForPost\n). These clues have been further examined by @zaska12345, but it has not been possible to further narrow down the time frame.\nTaking the mural paint as reference, Twitter user @john_marquee was able to find\na reference image from October 2017\nshowing that the paint on the ledge was there, as well as the advertisement for the music school. The Yandex image does not show the wall painted in white, and is dated early Spring 2016, suggesting the Babiš photo was taken after that.\nAnother Twitter user, @MCantow, found a clue for the last possible time. The second tree in the background (\nhere in Spring 2016\n) was cut between October 2017 (tree visible on Google Earth) and April 2018 (tree not visible anymore on DigitalGlobe satellite imagery).\nNovinky.cz Photo\nPictured:\nPetr Protopopov\nGeolocation:\n55.709491, 37.542220\n(Moscow, Russia)\nChronolocation:\nunknown\nThis photo of Petr Protopov was published alongside an interview with him in Czech media outlet\nNovinky.cz\n. Though there is no direct relation to the case, we thought it may be useful to geolocate it nonetheless.\nThe photo can be geolocated to the Sparrow Hills (Vorobyovy Gory) in Moscow, Russia. It is a popular spot for locals and tourists alike, hence many people recognised the view and the famous Luzhniki Stadium in the background.\nIt was the first photo to be geolocated, thanks to several Twitter users, including\n@ivladdalvi\n,\n@en4rab\n,\n@thesedentaire\n,\n@john_marquee\nand\n@zaska12345\n— proving the power of the crowd.\nVKontakte Photo 1\nPictured:\nAndrej Babiš Jr. and Jelizaveta\nGeolocation:\n46.203528, 6.148187\nChronolocation:\nbefore 5 March 2018\nThe first photo was uploaded on 4 March 2018 to VK by Jelizaveta, the girlfriend of Babiš Jr. It shows the couple sitting in a restaurant with decorated ornaments, wooden seats, and a plate of oysters.\nOn Twitter, Mark Reid was quick to notice that it may be a restaurant in Geneva:\nCafé du Centre\n. Using online reference photos, he and Twitter-users\n@Sector035\nand\n@cz_smidra\nwere not only able to match the similar plates used for the oysters, but also the lights, ornaments, ceiling, and wooden chairs.\nVKontakte Photo 2\nPictured:\nAndrej Babiš Jr., Jelizaveta, Santa Claus, and unidentified minor\nGeolocation:\n47.903308, 33.400988 (Kryvyi Rih, Ukraine)\nChronolocation:\nbefore 2 January 2018\nThe second VK photo was uploaded on January 2, 2018, five days before Babiš Jr. wrote Czech police that he was kidnapped by an associate of his father.\nAn eagle-eyed Twitter-user “Lara van Gaalen” as well as Bellingcat-reader “Sergio.Beer” spotted that the shape of the window shape and style, as well the objects outside of the windows strongly resemble the location shown in VKontakte Photo 7, which is discussed below. The spacing between the vertical loudspeakers match as well, and so does the back side of the canvas banner.\nAnd picture 2 might be at the same location as picture 7. Trough the window of photo 2 you can see two rectangular shapes (looks like speakers to me). These are also visible in picture 7, above the banner.\npic.twitter.com/i2wwJlhQye\n— Lara van Gaalen (@LaravanGaalen)\nFebruary 7, 2019\nTwitter-user ”Abraxas Spa“ and Bellingcat-reader “Sergio.Beer” both\nnoticed\nthat\nBabiš Jr. and Jelizaveta are apparently also wearing similar clothes, and the lady is wearing the same bag. This suggest that it may have been the same time period as VKontakte Photo 7, which was taken between 3 December 2017 and 14 January 2018.\nVKontakte Photo 3\nPictured:\nAndrej Babiš Jr., Jelizaveta, Santa Claus and Santa’s helper\nGeolocation:\n47.905535, 33.396696 (Multiplex Union, Kryvyi Rih, Ukraine)\nChronolocation:\nunknown (likely between 24 December and 16 January, the dates which encompasses both Christmas in Ukraine and Russia)\nThe third VK photo shows the couple posing in front of a Christmas tree and a dressed-up Santa Claus (Ded Moroz, or Father Frost, as he is known in many post-Soviet countries) and his helper, the Snow Maiden (Snegurochka), inside what appears to be a building.\nWith no distinctive visual clues such as a written text, interior photos are usually harder to geolocate. Nevertheless, Twitter-user @Sector035 was able to find a photo on VK that appears to be the exact same spot. However, no location was mentioned.. So how to figure out where it was taken?\nHaley Willis also found a reference photo that was indexed as being from Kryvyi Rih in Ukraine. That was @Sector035’s initial hunch as well. Both started looking for shopping malls in the Ukrainian city, looking for red booths with tables as visible on the top left corner of the three photos. Eventually, Ms. Willis found reference photos on Foursquare, and managed to find a matching place: the Multiplex Union in Kryvyi Rih.\nFound it! Looks like the Multiplex Union (вулиця Героїв АТО, 30Г, ТРЦ Union, Kryvyi Rih). Saw this first photo going through Foursquare and recognized the banisters. Then found second photo on Google Photos.\npic.twitter.com/or4aMkxFxX\n— Haley Willis (@heytherehaIey)\nDecember 8, 2018\nVKontakte Photo 4\nPictured:\nAndrej Babiš Jr. and Jelizaveta\nGeolocation:\n44.498875, 34.167997 (Saturn IMAX, Yalta, Russian-annexed Crimea)\nChronolocation:\nbetween late October and early November 2017\nThis photo was published on VK by Jelizaveta in late October 2017.\nA Bellingcat-reader going by the name “Alex” managed to geolocate the photo. Here’s how.\nTo the right of Mr.\nBabiš Jr. are two monitors, one vertical and one horizontal. By cropping, transforming, and reverse-image-searching the image, Alex identified the image on the vertical monitor as the film poster for “Black Panther”, an American superhero film that was released in January 2018.\nAlex started comparing Black Panther posters in Russian and Ukrainian, and noticed that he white tet under the logo in the bottom of the poster varied slightly in Russian and Ukrainian. The Russian “в кино с 15 февраля” is slightly longer than the Ukrainian “скоро в кіно”. While there may be more variations, it led Alex to have a look around cinemas in Russian-annexed Crimea.\nHe thought it must be a a “modern” and “trendy” Crimean cinema, given that it looks that way. The air hockey table on the right side may hold a clue as well. Since there are not that many new or renovated cinemas in Crimea, Alex narrowed down to Yalta, as the air hockey tables in cinemas in Sevastopol did not have matching floors or air hockey tables. Eventually, the Saturn IMAX (Russian: Сатурн IMAX) in Yalta was a match, as can be seen below on\ntwo\nreference\nphotos on Google Maps.\nBased on Alex’s geolocation, it is now possible to narrow down the timeframe by looking for geotagged photos of the Saturn IMAX in Yalta, and see when the Halloween corner appeared in the building. Instagram was specifically useful, as many users\ntagged their photos\nto the Saturn IMAX. The first photo of the setting appeared on\n27 October 2017\nand the last one in\n1 November 2017\n. As Halloween is celebrated on the 30th of October, these dates do make sense to have a Halloween decor in the cinema.\nThe first photo tagged to the Saturn IMAX showing the Halloween decor appeared on Instagram on 28 October 2017.\nVKontakte Photo 5 and 6\nPictured:\nAndrej Babiš Jr. and Jelizaveta\nGeolocation:\n46.205829, 6.157553 (Geneva, Switzerland)\nChronolocation:\nbetween October 2017 and July 2018\nThe fifth and sixth photos were uploaded to Jelizaveta’s VK profile in 2018. It shows the couple standing in front of a fountain.\nIt was publicly known that Babiš Jr. was with Jelizaveta in Geneva, and for that reason it was easier to geolocate these two photos.\nHoupací Osel\n’s Jan Zabka simply wrote “fountain Swiss” into Google Images, and the first result was the “Jet d’Eau” in Geneva, Switzerland. Using reference photos, it is then possible to geolocate their exact positions to 46.205829, 6.157553 and 46.207357, 6.156086, respectively.\nWith regards to chronolocation, the stickers next to couple hold a clue (very similar to the Papadopulos case). When was the sticker there, and when was it not? Using reference photos on a variety of online platforms, it is possible to determine that the photo must have been taken between October 2017 and July 2018.\nVKontakte Photo 7\nPictured:\nAndrej Babiš Jr. and Jelizaveta\nGeolocation:\n47.903662, 33.401034 (Kryvyi Rih, Ukraine)\nChronolocation:\nbetween 3 December 2017 and 14 January 2018\nThe seventh VK photo of interest shows the couple in front of a large building.\nThe banner on the background is the first visual clue, as it allows to do a text-based search on ”бульвар чуд(…) Італійський (…)” on Google which results in a reference photo of the same poster.\nIn that way, it is possible to figure out the whole name of the show, and the place where it took place: the Kryvorizkyi Circus in Kryvyi Rih, Ukraine. The couple is tanding specifically at 47.903662, 33.401034. It was geolocated by\nHoupací Osel\n’s Jan Zabka.\nThe banner in the background allows chronolocation: the show “Бульвар чудес” (Boulevard of Miracles) only ran from December 16, 2017 to January 14, 2018, the banner reveals. Reference photos show there was a different banner until December 3, 2017.\nVKontakte Photo 8\nPictured:\nAndrej Babiš Jr. and Jelizaveta\nGeolocation:\n44.494549, 34.117881 (Yalta, Russian-annexed Crimea)\nChronolocation:\nunknown\nThe eighth and last VK photo of interest shows the couple standing in front of a caged giraffe. An orange, real-size statue of a giraffe is also visible in the background.\nSome Czech media\nmentioned\nthat this photo was taken in a zoo in Yalta, Russian-annexed Crimea.\nHoupací Osel\n’s Jan Zabka managed to confirm that this is the “Skazka” Zoo (Зоопарк “Сказка”) in Yalta, Crimea, by geolocating the photo to 44.494549, 34.117881.\nReference photos found via\nGoogle\nand\nTripAdvisor\nallow to match the details such as the fence, the wall, and the orange giraffe, to confirm the geolocation.\nWe have not been able to determine the date and time the photo was taken.\nAnalysis and Conclusion\nWhen the Czech Prime Minister published the three initial photos as a proof to the public that Babiš Jr. did not suffer while being in Crimea, a large discussion about the authenticity of the photos started among Czech citizens. For that very reason, it is important to geo- and chrono-locate the photos. Only then can we establish some facts.\nThe crowdsourcing campaign resulted in successfully geolocating 12 out of 12 photos, and getting a better sense of time by roughly chronolocating half of them.\nThe results show that Andrej Babis Jr. indeed visited the Ukrainian city of Kryvyi Rih, as well as Sevastopol and Yalta in Russian-annexed Crimea. With regards to Crimea, it is likely that this was indeed in the latter half of 2017. This backs up the claims related to the three initial photos published by the Czech Prime Minister: the “where” and “when” are correct.\nThe geolocation does not say anything about the status of Babis Jr. when he was there. It does, at least, authenticate Babis Jr. presence in Russian-annexed Crimea in late 2017. This is an important fact in a highly politicised and sensitive case in which there has been a lot of speculation.\nIt has been three months the first interview with Babiš Jr. was published. There were mass protests against the government, but it survived a no-confidence vote. Babiš Sr. is still the Prime Minister of the Czech Republic, and the turmoil surrounding the alleged kidnapping case has wound down. In fact, a\nrecent survey\nsuggests that the PM’s popularity actually increased as many voters accepted his narrative of “unethical journalists’ hunt for sensation” who harmed his “ill son” and “family as a whole.”\nCorrection 7 February 2019: An earlier version mentioned that the location of VKontakte Photo 3 had not been found yet. However, it was already\ngeolocated by\nHaley Willis on Twitter.\nUpdate 9 February 2019: A Google MyMaps of the geolocated photos has been added to the concluding section of the article. Furthermore, Bellingcat-readers and Twitter-users have found the location of VKontakte Photo 2, as well as a closer chronolocation of Facebook Photo 1. The article has been updated accordingly. An earlier version of the article also mentioned that there was reference photo of a tree in 2014. This should have been 2018.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/babis-4.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/image-1-1200x672.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/image-2-1200x773.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/image-3-1200x799.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/image-5-1200x680.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/andrej-babis-3762053_1280-1200x800.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/image-6-1200x677.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/image-7-1200x636.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtIHQf9XcAAAJlJ.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtJn9jXU8AA7vX2-1200x650.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtJpTBFVAAE-3_u.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtIHQfSWsAAYBYU.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtKjuGTUwAAc-Ky-1200x900.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtM79_3X4AA9njb-1199x1200.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtM78qeXgAIG9_s-1200x1199.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtNFW-pW0AAeniP-1200x647.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtNN0FTWoAIT6Pj-1200x802.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtNT8k0WwAAFVbe-150x150.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtNT7yYWoAIFui3-150x150.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtNTa7QXQAA4ou7-150x150.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtNT9itXcAA2TK0-150x150.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtNb2tfXQAA5fBc-1-1200x889.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtNb4AhWwAAIswR-1200x889.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Screenshot-2019-02-06-at-21.48.58-1200x751.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Screenshot-2019-02-06-at-21.50.43-1200x671.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtIHQfEWwAIfwH3.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/3_0015.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtLM2iAXQAAQIsv-1200x673.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Screenshot-2019-02-06-at-21.57.01-1200x700.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/716819-top_foto1-6yqv9.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/DtKfrTvWkAUflhW-1200x594.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Dtw0WD_XQAEUPUr.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Dtw4NkaWoAAhy7v.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Dtw2wqpW4AIa-n-.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Dtw2xvbXQAMc2zc.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/image-10.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Dtw2ycyWoAAvbJt.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/YANDEX-1200x542.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/BLACK-PANTHER-PHOTOS-1200x688.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/air-hockey-1200x612.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/reference-photo-1200x886.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Screenshot-2019-02-06-at-22.22.53.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Dtw7isMXcAEWbF6.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Dtw-LrYXcAEN8__.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Dtw-huNXgAAu9-k-1200x895.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/02/Dtw-kXkW0AAZ_7J-1200x850.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2019-03-04_geolocation-and-a-philosophers-stone-in-kashmir.json b/bellingcat_geolocation_json/2019-03-04_geolocation-and-a-philosophers-stone-in-kashmir.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4c29fd148a9c9b10a0b3f8c10aa557f73a06d14
--- /dev/null
+++ b/bellingcat_geolocation_json/2019-03-04_geolocation-and-a-philosophers-stone-in-kashmir.json
@@ -0,0 +1,67 @@
+{
+ "slug": "geolocation-and-a-philosophers-stone-in-kashmir",
+ "title": "Geolocation and a Philosopher's Stone in Kashmir",
+ "url": "https://www.bellingcat.com/resources/case-studies/2019/03/04/geolocation-and-a-philosophers-stone-in-kashmir/",
+ "date": "2019-03-04T08:32:09+00:00",
+ "modified": "",
+ "author": "Timmi Allen",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)On February 27 of this year, a MiG-21 Bison aircraft of the Indian Air Force was shot down over the contested Kashmir region of Pakistan. Soon afterward, pictures and videos showing the wreck of an Indian MiG-21 Bison at various locations were published across both traditional and social media. Bellingcat initially reported on […]",
+ "content_html": "
On February 27 of this year, a MiG-21 Bison aircraft of the Indian Air Force was shot down over the contested Kashmir region of Pakistan. Soon afterward, pictures and videos showing the wreck of an Indian MiG-21 Bison at various locations were published across both traditional and social media. Bellingcat initially reported on this incident on March 2 — and took into account the associated war of words between involved parties.
\n
The exact crash location of the MiG-21 was, however, not widely known until now. Most comments about the crash location pointed to the Samahni district in the Kashmir region; however, this is a relatively inexact location. We can assume that the MiG-21 debris is distributed in a larger area.
\n
Two photographs have been widely distributed that show an area in which a large part of the debris (front) of the MiG crashed.
\n
\n
\n
From these two photographs, we can create a panorama to make studying the area easier for conducting geolocation. This can be done by just patching together the two images in image editing software at the intersection points between the two photographs.
\n
\n
Additional images from social networks showed the same place from different perspectives.
\n
\n
\n
Although several videos of the scene can be found across social networks, the exact location was still difficult to determine as there are no visible unique buildings or man-made structures — except for just one house in the distance — that allow for an easy geolocation.
\n\n
One of the best details visible in the published images are the trees on the mountainside. While it would eventually yield a result, spending hours looking at every patch of vegetation on the hills of Kashmir across several hundred square kilometers is not a very good method.
\n
For this reason, further information was sought that would allow for a better spatial determination and a clue of where to focus our search. In finding and watching a video in which an eyewitness was questioned, the golden clue was provided. In this video, a reporter asks the eyewitness, a man named Mohammad Yasin, where they are, to which he answers “Horran Kotla.”
\n
\n
Toponyms, especially in relatively remote locations, can be very difficult to find in online mapping services due to transliteration errors and local usage differences. For example, searching “Horran Kotla” in Google refers you to the Indian village of Kotla Heran, as Google attempted to “correct” any transliteration differences to no success. However, a further search for “Horran Kotla” returned a hit in Mapcarta with a location near the reported crash site in the Kashmir region.
\n
\n
With this delimitation of the area, the search could be intensified. A key feature that can also be cross-referenced in satellite imagery is visible in the panorama: two larger boulders amongst the vegetation.
\n
\n
After trial-and-error searches on Google Earth in this area, we find these two boulders, as well as the arrangement of the trees in the scene, leading to the location of this crash site: 33°13’28.9 “N 73°57’53.3 “E.
\n
\n
Important tip: When using Google Earth, you should turn on the “Terrain” feature (bottom-left part of the interface) to cross-reference the hills, valleys, and mountains as they appear in your source material.
\n
\n
Thus, even a remote area with few identifying markers can eventually be geolocated.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nOn February 27 of this year, a MiG-21 Bison aircraft of the Indian Air Force was shot down over the contested Kashmir region of Pakistan. Soon afterward, pictures and videos showing the wreck of an Indian MiG-21 Bison at various locations were published across both traditional and social media. Bellingcat\ninitially reported on this incident on March 2\n— and took into account the associated war of words between involved parties.\nThe exact crash location of the MiG-21 was, however, not widely known until now. Most comments about the crash location pointed to the\nSamahni district in the Kashmir region\n; however, this is a relatively inexact location. We can assume that the MiG-21 debris is distributed in a larger area.\nTwo photographs have been widely distributed that show an area in which a large part of the debris (front) of the MiG crashed.\nFrom these two photographs, we can create a panorama to make studying the area easier for conducting geolocation. This can be done by just patching together the two images in image editing software at the intersection points between the two photographs.\nAdditional images from social networks showed the same place from different perspectives.\nAlthough several videos of the scene can be found across social networks, the exact location was still difficult to determine as there are no visible unique buildings or man-made structures — except for just one house in the distance — that allow for an easy geolocation.\nOne of the best details visible in the published images are the trees on the mountainside. While it would eventually yield a result, spending hours looking at every patch of vegetation on the hills of Kashmir across several hundred square kilometers is not a very good method.\nFor this reason, further information was sought that would allow for a better spatial determination and a clue of where to focus our search. In finding and watching a video in which an eyewitness was questioned, the golden clue was provided. In this video, a reporter asks the eyewitness, a man named Mohammad Yasin, where they are, to which he answers “Horran Kotla.”\nToponyms, especially in relatively remote locations, can be very difficult to find in online mapping services due to transliteration errors and local usage differences. For example, searching “Horran Kotla” in Google refers you to the Indian village of\nKotla Heran\n, as Google attempted to “correct” any transliteration differences to no success. However, a further search for “Horran Kotla”\nreturned a hit in Mapcarta\nwith a location near the reported crash site in the Kashmir region.\nWith this delimitation of the area, the search could be intensified. A key feature that can also be cross-referenced in satellite imagery is visible in the panorama: two larger boulders amongst the vegetation.\nAfter trial-and-error searches on Google Earth in this area, we find these two boulders, as well as the arrangement of the trees in the scene, leading to the location of this crash site:\n33°13’28.9 “N 73°57’53.3 “E.\nImportant tip: When using Google Earth, you should turn on the “Terrain” feature (bottom-left part of the interface) to cross-reference the hills, valleys, and mountains as they appear in your source material.\nThus, even a remote area with few identifying markers can eventually be geolocated.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/D0Zi8LnXgAAtE2o.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/D0Zi8LjX4AAX47X.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/Chrash-Site-2-Panorama.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/D0ZzusUWkAA9uvE.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/D0bojyjX0AE6c19.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/Place-Horran-Kotla.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/Kotla-Mapcarta-1.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/rocks.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/chrash-site-place-2.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/ge-terrain.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2019-03-05_how-to-use-google-earths-three-dimensional-view-feat-syria-yemen-sudan.json b/bellingcat_geolocation_json/2019-03-05_how-to-use-google-earths-three-dimensional-view-feat-syria-yemen-sudan.json
new file mode 100644
index 0000000000000000000000000000000000000000..bef261c91f018a34585032e284e1e987d119eb10
--- /dev/null
+++ b/bellingcat_geolocation_json/2019-03-05_how-to-use-google-earths-three-dimensional-view-feat-syria-yemen-sudan.json
@@ -0,0 +1,132 @@
+{
+ "slug": "how-to-use-google-earths-three-dimensional-view-feat-syria-yemen-sudan",
+ "title": "How To Use Google Earth's Three Dimensional View: Feat. Syria, Yemen, Sudan",
+ "url": "https://www.bellingcat.com/resources/how-tos/2019/03/05/how-to-use-google-earths-three-dimensional-view-feat-syria-yemen-sudan/",
+ "date": "2019-03-05T16:31:01+00:00",
+ "modified": "2019-03-05T16:39:47+00:00",
+ "author": "Benjamin Strick",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)It has become abundantly clear that geolocation and remote sensing is fundamental to knowing what happens in conflict areas where it is often too dangerous for journalists, observers and analysts to be present on the ground. This is a satellite-based case study on vetting a three-dimensional view of an area via Google Earth, […]",
+ "content_html": "
It has become abundantly clear that geolocation and remote sensing is fundamental to knowing what happens in conflict areas where it is often too dangerous for journalists, observers and analysts to be present on the ground.
\n
This is a satellite-based case study on vetting a three-dimensional view of an area via Google Earth, featuring examples from Syria, Yemen and Sudan.
\n
The satellite imagery we use for conflict analysis, geolocation, or any other form of investigation is a flat image. Aside from city-based areas, there is limited three-dimensional views of buildings or features — all of which are crucial to monitoring and geolocating events. Google Earth does have three-dimensional mountains, but even they sometimes fail to represent what is on the ground.
\n
However, it is possible to use varying angles of satellite imagery to get a three-dimensional understanding of an area.
\n
A key feature of satellite imagery is the nadir. This is the view from above the Earth directly to its center. Having an angle “off nadir” provides a variation to the standard map view. Each time a satellite captures imagery the nadir varies in its angle of shot.
\n
The images of my plant below illustrate this. The middle image is a replication of nadir angle 0 degrees. It’s a direct downward shot. The other two are varied angles, but given their presence, you have a working knowledge of the sides of the plant, as well as its top.
\n
\n
Now let’s use this information for a practical approach to geolocating serious incidents that occur during conflict.
\n
There are three important ways you can use a varied nadir to your advantage, they are:
\n\n
Identifying the sides of a building
\n
Identifying the true features of landforms
\n
Identifying the shadows of objects
\n\n
How can you access varying angles of the nadir? As explained above, each time satellite imagery is taken, the nadir will vary. If you have the Google Earth application, you are able to access past imagery.
\n
For example, in this view of Sudan’s central area, each prior image we select using the “historical imagery” option shows an alternative angle of the buildings, exposing different sides, shadows, and small side streets.
\n
\n
\n
\n
The following will include examples of what you can use to assist when viewing alternative angles of past imagery for geolocation and analysis purposes.
\n
Identifying The Sides of A Building
\n
Alternative nadir angles may expose the sides of buildings allowing analysts to get a more three-dimensional view of their design.
\n
In the first example, we look at a prison of the National Intelligence and Security Service of Sudan in Khartoum. I previously spoke with detainees of the prison, who claimed they had been tortured at the facility. When they described what it looked like, they were only able to give details as to the sides of the building, and how many levels it had.
\n
By viewing alternative angles of the building, we can match up those statements with varying views available through satellite imagery.
\n
\n
While that form of geolocation may be used to become familiar with how a building looks, this is also an effective practice for geolocating video footage.
\n
In this video from Sudan, we are able to see a number of building sides. The specific scene below shows one view that can be used:
\n
\n
There are three buildings in the image that are at varying positions. The green center building and the glass building on the right have unique features. These features can be seen on angled views to see the north facing and side panels of the buildings.
\n
Simply scrolling back in Google Earth’s satellite imagery will provide you with a clear angled view of the front of those buildings. On Google Maps, we would not be able to see this.
\n
We can’t see the face of the green building because of the shadows it casts, but we can definitely see its color on the side, which is present in the video.
\n
\n
IMPORTANT NOTE: When geolocating footage from buildings, be sure to carry out an assessment on whether or not the footage is filmed from a private dwelling. Publishing the geolocation of this may endanger the filmer should the content be sensitive.
\n
Identifying The True Features Of Landforms
\n
Using alternative nadir angles is also crucial for true identification of natural formations.
\n
In Google Earth and Google Maps, we are able to access 3D views of mountains and land formations. However, some of those formations may not reflect exactly what is present on the ground.
\n
For instance, some mountain ranges may have rocky outcrops. These formations will not be present on a 3D view, but viewing varying angles and shadows can show the true ridge or outline of a mountain by the shadows it casts.
\n
In this video, two Houthi rebels inspect a crash site in Yemen geolocated for the purposes of a GermanArms workshop.
\n
\n
The rise in the background has a unique “layering” effect to the rocks as well as a large flat-faced shelf and a number of rectangular-shaped rocks below (as seen in the red box). Geolocating this can be quite difficult as the layering is not reflected on satellite imagery (below) either in its 3D view or flat view.
\n
In the 3D view, notice the outline of the mountain. It does not represent the image of the mountain in the video.
\n
\n
\n
However, when looking at an alternate angle from past imagery, further depth is shown in the rock face (as seen below). With this view, it gives us a more “on-the-ground” feel of what that mountain would look like, to see if it makes a match for what we can glimpse in the video.
\n
\n
The corner pieces of the mountain show unique depth and shadows to confirm its position.
\n
A similar example to show why angled imagery is important for mountain geolocation comes from Afrin, Syria.
\n
In this footage from Afrin, shot through the cockpit of a T-129 helicopter, a cluster of buildings can be seen being attacked. The mountain ridge shows a unique rise of its rocky outcrop:
\n
\n
The geolocation of the mountain, confirmed with the coordinates in the bottom right of the helicopters military grid reference view, appears on Google Earth with this ridgeline.
\n
\n
The cluster of buildings destroyed are seen at a more direct nadir here.
\n
\n
However, one must notice that both the outline of the three-dimensional view, and the “top-down” view does not replicate that sharp ridge we saw in the video. What we can see in the image above is a large rock-laden outcrop. Finding that on an image where the shadows are larger illustrates its height, as seen below.
Geolocation work can be made easy by identifying large man made objects in the rear of the scene under analysis. This might be, for example, a minaret, communications tower, or even power lines.
\n
Finding those objects can often be a challenge, but can be made easier via using angled imagery or object shadows.
\n
Something as distinct as a power line pole can be helpful in geolocation, though they are often hard to pick out. Using imagery where there is a high angled view in addition with an early or late in the day image can highlight these objects as their shadows will be longer.
\n
For instance, in the satellite image below, it is very difficult to pick out the power lines.
\n
\n
Shifting to a more alternative view on a different date yields much better results with the accentuated shadows of objects.
\n
\n
In a practical example, geolocating a video of Sudan’s security forces firing into the air during a buildup of protests on December 28, 2018, was achieved by using the power lines as reference.
Notice the double pole (as seen in the image), while many of the surrounding power line poles were single poles.
\n
\n
Having geolocated a number of other videos of security forces firing sporadically in the area, the concrentation point for geolocating this video was based around the Royal Care International Hospital (مُستشفى رويال كير العالمية) in Khartoum’s east.
\n
The vehicles were driven along a road separated by black and yellow striped dividers:
\n
\n
Given that the main roads in the area used this style of road divider, the position of the double-legged powerline pole was crucial to the geolocation of this video. This was also matched with the other surrounding power line poles and trees to find a true position.
\n
Changing imagery can reveal a very clear representation of the object we are looking for. In the image below, the three screenshots of the same poles are from Google Earth and different dates. The one in the middle shows the ideal representation of the powerlines. The image on the far right is also what is shown on Google Maps.
\n
\n
This method can also be useful as far as much larger objects are concerned. For instance, you may be looking for a communications tower or a minaret from a mosque in a specific area.
\n
In the Google Earth image below we have both:
\n
\n
Using a past image, we can view the shadows which will reveal these tall objects.
\n
\nViewing exactly what the minaret looks like from top to bottom, as well as the communications tower, can be done with Google Earth imagery taken on a large angle.\n
\n
Using Angles For A Practical Approach
\n
In any open source investigation, it is much easier to find your goal with a lot of information than barely any at all. Google Earth’s past imagery is part of that information. It shows development, changes, seasons, new objects and, as discussed in this post, varying angles.
\n
Viewing those varying angles of nadir is a practical skill that can help you in all kinds of ways — from the most simple of geolocation tasks to more complicated endeavors featuring remote monitoring and sensing.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nIt has become abundantly clear that geolocation and remote sensing is fundamental to knowing what happens in conflict areas where it is often too dangerous for journalists, observers and analysts to be present on the ground.\nThis is a satellite-based case study on vetting a three-dimensional view of an area via Google Earth, featuring examples from Syria, Yemen and Sudan.\nThe satellite imagery we use for conflict analysis, geolocation, or any other form of investigation is a flat image. Aside from city-based areas, there is limited three-dimensional views of buildings or features — all of which are crucial to monitoring and geolocating events. Google Earth does have three-dimensional mountains, but even they sometimes fail to represent what is on the ground.\nHowever, it is possible to use varying angles of satellite imagery to get a three-dimensional understanding of an area.\nA key feature of satellite imagery is\nthe nadir\n. This is the view from above the Earth directly to its center. Having an angle “off nadir” provides a variation to the standard map view. Each time a satellite captures imagery the nadir varies in its angle of shot.\nThe images of my plant below illustrate this. The middle image is a replication of nadir angle 0 degrees. It’s a direct downward shot. The other two are varied angles, but given their presence, you have a working knowledge of the sides of the plant, as well as its top.\nNow let’s use this information for a practical approach to geolocating serious incidents that occur during conflict.\nThere are three important ways you can use a varied nadir to your advantage, they are:\nIdentifying the sides of a building\nIdentifying the true features of landforms\nIdentifying the shadows of objects\nHow can you access varying angles of the nadir? As explained above, each time satellite imagery is taken, the nadir will vary. If you have the\nGoogle Earth application\n, you are able to access past imagery.\nFor example, in this view of Sudan’s central area, each prior image we select using the “historical imagery” option shows an alternative angle of the buildings, exposing different sides, shadows, and small side streets.\nThe following will include examples of what you can use to assist when viewing alternative angles of past imagery for geolocation and analysis purposes.\nIdentifying The Sides of A Building\nAlternative nadir angles may expose the sides of buildings allowing analysts to get a more three-dimensional view of their design.\nIn the first example, we look at a\nprison\nof the National Intelligence and Security Service of Sudan in Khartoum. I previously spoke with detainees of the prison, who claimed they had been\ntortured at the facility\n. When they described what it looked like, they were only able to give details as to the sides of the building, and how many levels it had.\nBy viewing alternative angles of the building, we can match up those statements with varying views available through satellite imagery.\nWhile that form of geolocation may be used to become familiar with how a building looks, this is also an effective practice for geolocating video footage.\nIn this\nvideo\nfrom Sudan, we are able to see a number of building sides. The specific scene below shows one view that can be used:\nThere are three buildings in the image that are at varying positions. The green center building and the glass building on the right have unique features. These features can be seen on angled views to see the north facing and side panels of the buildings.\nSimply scrolling back in Google Earth’s satellite imagery will provide you with a clear angled view of the front of those buildings. On\nGoogle Maps\n, we would not be able to see this.\nWe can’t see the face of the green building because of the shadows it casts, but we can definitely see its color on the side, which is present in the video.\nIMPORTANT NOTE: When geolocating footage from buildings, be sure to carry out an assessment on whether or not the footage is filmed from a private dwelling. Publishing the geolocation of this may endanger the filmer should the content be sensitive.\nIdentifying The True Features Of Landforms\nUsing alternative nadir angles is also crucial for true identification of natural formations.\nIn Google Earth and Google Maps, we are able to access 3D views of mountains and land formations. However, some of those formations may not reflect exactly what is present on the ground.\nFor instance, some mountain ranges may have rocky outcrops. These formations will not be present on a 3D view, but viewing varying angles and shadows can show the true ridge or outline of a mountain by the shadows it casts.\nIn this video, two Houthi rebels inspect a crash site in\nYemen\ngeolocated for the purposes of a\nGermanArms\nworkshop.\nThe rise in the background has a unique “layering” effect to the rocks as well as a large flat-faced shelf and a number of rectangular-shaped rocks below (as seen in the red box). Geolocating this can be quite difficult as the layering is not reflected on satellite imagery (below) either in its 3D view or flat view.\nIn the 3D view, notice the outline of the mountain. It does not represent the image of the mountain in the video.\nHowever, when looking at an alternate angle from past imagery, further depth is shown in the rock face (as seen below). With this view, it gives us a more “on-the-ground” feel of what that mountain would look like, to see if it makes a match for what we can glimpse in the video.\nThe corner pieces of the mountain show unique depth and shadows to confirm its position.\nA similar example to show why angled imagery is important for mountain geolocation comes from\nAfrin, Syria\n.\nIn this\nfootage from Afrin\n, shot through the cockpit of a T-129 helicopter, a cluster of buildings can be seen being attacked. The mountain ridge shows a unique rise of its rocky outcrop:\nThe geolocation of the mountain, confirmed with the coordinates in the bottom right of the helicopters military grid reference view, appears on Google Earth with this ridgeline.\nThe cluster of buildings destroyed are seen at a more direct nadir here.\nHowever, one must notice that both the outline of the three-dimensional view, and the “top-down” view does not replicate that sharp ridge we saw in the video. What we can see in the image above is a large rock-laden outcrop. Finding that on an image where the shadows are larger illustrates its height, as seen below.\nNote: this specific geolocation was a crowdsourced for the\nItalianArms\nworkshop conducted by\nChristiaan Triebert\n,\nLeone Hadavi\n,\nTimmi Allen\nand\nThomas\n.\nIdentifying The Shadows Of Objects\nGeolocation work can be made easy by identifying large man made objects in the rear of the scene under analysis. This might be, for example, a minaret, communications tower, or even power lines.\nFinding those objects can often be a challenge, but can be made easier via using angled imagery or object shadows.\nSomething as distinct as a power line pole can be helpful in geolocation, though they are often hard to pick out. Using imagery where there is a high angled view in addition with an early or late in the day image can highlight these objects as their shadows will be longer.\nFor instance, in the satellite image below, it is very difficult to pick out the power lines.\nShifting to a more alternative view on a different date yields much better results with the accentuated shadows of objects.\nIn a practical example, geolocating a video of Sudan’s security forces firing into the air during a buildup of protests on December 28, 2018, was achieved by using the power lines as reference.\nIn the video below, a unique power line appears.\nhttps://twitter.com/BenDoBrown/status/1078719876909416448\nNotice the double pole (as seen in the image), while many of the surrounding power line poles were single poles.\nHaving geolocated a number of other videos of security forces firing sporadically in the area, the concrentation point for geolocating this video was based around the Royal Care International Hospital (مُستشفى رويال كير العالمية) in Khartoum’s east.\nThe vehicles were driven along a road separated by black and yellow striped dividers:\nGiven that the main roads in the area used this style of road divider, the position of the double-legged powerline pole was crucial to the geolocation of this video. This was also matched with the other surrounding power line poles and trees to find a true position.\nChanging imagery can reveal a very clear representation of the object we are looking for. In the image below, the three screenshots of the same poles are from Google Earth and different dates. The one in the middle shows the ideal representation of the powerlines. The image on the far right is also what is shown on\nGoogle Maps.\nThis method can also be useful as far as much larger objects are concerned. For instance, you may be looking for a communications tower or a minaret from a mosque in a specific area.\nIn the Google Earth image below we have both:\nUsing a past image, we can view the shadows which will reveal these tall objects.\nViewing exactly what the minaret looks like from top to bottom, as well as the communications tower, can be done with Google Earth imagery taken on a large angle.\nUsing Angles For A Practical Approach\nIn any open source investigation, it is much easier to find your goal with a lot of information than barely any at all. Google Earth’s past imagery is part of that information. It shows development, changes, seasons, new objects and, as discussed in this post, varying angles.\nViewing those varying angles of nadir is a practical skill that can help you in all kinds of ways — from the most simple of geolocation tasks to more complicated endeavors featuring remote monitoring and sensing.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/plantangles.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/3.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/4.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/12.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/Screen-Shot-2019-02-22-at-3.32.22-pm.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/5.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/13.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/6.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/7.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/8.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/9.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/10.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/11.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/Screen-Shot-2019-02-22-at-3.59.40-pm.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/Screen-Shot-2019-02-22-at-4.03.39-pm.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/15.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/16.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/17.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/18.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/19.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/20.png",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2019-03-19_locating-the-netherlands-most-wanted-criminal-by-scrutinising-instagram.json b/bellingcat_geolocation_json/2019-03-19_locating-the-netherlands-most-wanted-criminal-by-scrutinising-instagram.json
new file mode 100644
index 0000000000000000000000000000000000000000..7699dd44273b3526a21c69475959656b17bd19fb
--- /dev/null
+++ b/bellingcat_geolocation_json/2019-03-19_locating-the-netherlands-most-wanted-criminal-by-scrutinising-instagram.json
@@ -0,0 +1,132 @@
+{
+ "slug": "locating-the-netherlands-most-wanted-criminal-by-scrutinising-instagram",
+ "title": "Locating The Netherlands' Most Wanted Criminal By Scrutinising Instagram",
+ "url": "https://www.bellingcat.com/news/europe/2019/03/19/locating-the-netherlands-most-wanted-criminal-by-scrutinising-instagram/",
+ "date": "2019-03-19T23:22:31+00:00",
+ "modified": "2019-10-23T09:56:46+00:00",
+ "author": "Henk Van Ess",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)A fugitive convict from the Netherlands is taunting Dutch police with provocative pictures and videos from Iran, playing a game of “catch me if you can” on Instagram. “I love my freedom. And the weather,” he has stated. Bellingcat was able to reveal the last known location of the criminal with the help […]",
+ "content_html": "
A fugitive convict from the Netherlands is taunting Dutch police with provocative pictures and videos from Iran, playing a game of “catch me if you can” on Instagram. “I love my freedom. And the weather,” he has stated.
\n
Bellingcat was able to reveal the last known location of the criminal with the help of over 60 Twitter users. Here’s how we did that.
\n
Shahin Gheiybe is a convicted criminal hiding in Iran
\n
The escape
\n
The fugitive, 35-year-old Shahin Gheiybe, is on the Dutch “Most Wanted” list. On July 5, 2011, Gheiybe was sentenced for two attempted murders. In the city of ‘s-Hertogenbosch, on July 6, 2009, he tried to kill two individuals by shooting at them. He also robbed them of 175,000 euros. Gheiybe, born in Iran, was sentenced to a term of imprisonment of thirteen years.
\n
He managed to escape from prison on October 6, 2011, and has been on the run ever since. The police in the Netherlands have searched for this individual since then.
\n
The convict posts on Instagram (there are over 170 photos and videos) about his whereabouts (here is the original link to his private account, and a backup link), mostly in Iran. He has almost never tagged himself with a specific geolocation, just towns and places. Only once, for a street in Teheran, the coordinates were very precise. Gheiybe’s Instagram account was created in July 2015. It became private this month, after he landed on the list of most wanted criminals.
\n
\n
His last Instagram post is a clip from Scarface (archived post). Gheiybe’s Instagram account contains provocative material, like personal videos accusing the ruling judge in the Netherlands, claiming “the truth is very different.” He tries to explain to his followers he is a nice guy who made a mistake.
\n
A Dutch police spokesperson called this “shocking” and added, “It is unbelievable how he distorts the facts and that he cares so little about his victims and our society. Apparently he considers himself untouchable in Iran.”
\n
Problems To Solve
\n
A problem with scrutinising any of the Instagram posts is that the convict can deceive the police by using bogus location tags in Instagram, backdating the information, or otherwise hiding his real whereabouts. We looked mostly at clues where the convict is clearly visible and, at the same time, is near a traceable object (Gheiybe near a car, in a club, holding a branded knife).
\n
Using that method, we tried to reduce the risk of being deceived ourselves.
\n
Another problem centered on the ethics of impacting the investigation. If we can find a last known location — as we ultimately did — any convict undoubtedly becomes aware that they are being zeroed in on.
\n
Without crowdsourcing, any fugitive’s last address will be much harder to find. Bellingcat got the help of over 60 people who researched with us via Twitter. That research was conducted mostly on March 18, 2019.
\n
The police told the public they didn’t know the location. But the public (via crowdsourcing) ultimately did.
\n
This is an example of how transparency works both ways: It can be used by both good and bad people. We decided that we would send the police even the tiniest leads during our investigation. And that we would alert them the moment we would find something actionable, so they had at least a head start.
\n
Technical Preparations
\n
To get our hands on all the Instagram posts, we needed to prepare first. We followed these three steps:
\n
Browse Instagram On Desktop
\n
We didn’t want to fiddle with a mobile or tablet, but work on a desktop computer. To do that, we downloaded Desktop for Instagram, a Chrome plug-in.
\n
Download Videos
\n
Downloading over 170 Instagram videos and photos manually is time consuming. We used a plug-in for Chrome called Downloader for Instagram. It’s fast, downloads all material in full resolution, and also supports Instagram stories downloads.
\n
Find Hidden Connections
\n
To find the hidden connection between Instagram users, to find what our target liked, and much more, we used Helper Tools for Instagram, again a plug-in for Chrome.
\n
\n
In the picture above, for example, the convict is next to a liquor display — focusing in on specific objects can sometimes lead to finding connections. Or, at the very least, you find something about the target’s likes and lifestyle.
\n
Question 1: Whom Did He Meet?
\n
Last week, Dutch police asked the Dutch public for help with this case on TV.
\n
\n
For English followers: translation of \"Opsporing Verzocht\" (17/17) so you understand my last tweets a little better. Full explainer coming up tonight. pic.twitter.com/3aplJ1Yz6F
One question was: Who did the convict interact with? Maybe his interlocutors know his possible bogus identity, like a false passport or a false name. We concentrated on that question first.
\n
A Man And A Statue:
\n
\n
\n
\n
This picture, uploaded on November 27, 2018, shows our target play-fighting a statue, which perfectly suits our need to have an identifiable object and our target in the same photo. We did reverse image search in Google and Yandex — and, as it is often the case, Yandex was the better choice.
\n
\n
This is how we found Kish Island, some 200 kilometers away from Dubai.
\n
\n
A slightly bizarre promotional story about this island tells you what you can do there:
\n
“Women sit with cigarette in hand, wearing colorful headscarves pushed right back to reveal plenty of make-up and expensive hair-dos” and “The wide, palm tree-lined boulevards that circle the island of 100 square kilometers are full of top-end cars, including luxury American models.”
\n
On the same day, November 27, 2018, we found a video where our target was near a car with a recognizable license plate. The video shows our target getting into a red car and driving away in it.
\n
\n
The next step was for people with a great amount of patience. We tried to find hotels in Kish via Google Streetview that matched the building in the video background. In the end, we found the exact spot (pictured here in daylight):
\n
\n
Reverse imaging the license plate didn’t work at first. Sometimes, Google simply doesn’t recognize a blown up detail of a photo.
\n
\n
You can, however, approach this problem by just typing in words you can see into Google Images, without reverse searching the photo itself:
\n
\n
This leads to images of tourists driving the same red Chevrolet Camaro with the same license plate.
\n
\n
Eventually, we found the Camaro supplier by clicking on some Instagram posts featuring happy clients.
\n
Our convict didn’t steal the car. After he drove the Camaro, people were still renting it, as we found out after checking another series of posts by satisfied renters. So, we assume he had rented the vehicle as well.
\n
\n
Kish Rent Car, a local car rental place, can shine a light on the current identity of our target. We have a first possible answer to our initial question: Whom did our target meet?
\n
Finding Replies Sent To Other Instagram Accounts
\n
Did our target reply to other instagrammers? To find out, we used this formula in Google.
\n
\nWe don’t want to see postings from the target’s own account (hence we exclude site:instagram.com/shahin) but are still interested if he (hence shahin.mzr) posted comments on the Instagram posts of others (hence site:instagram.com). This is what we found:\n
\n
Around August 3, 2018, our target asks knife._.store, the Instagram account of an Iranian knife seller, if he can buy a knife and if it can be sent to him. He was prepared to pay 65,000 Thoman, which is around around $15, for the knife.
\n
\n
Later that year, on October 22, 2018, Gheyibe posted a photo of two knives on his own Instagram account. The one to the left (see photo above) is the Phrobis III Buck Gen 4 Fighting Knife M9 Scabbard USGI Military Wire Cutter. This model had gone on sale for a short moment in the autumn of 2018, at the previously mentioned knife storein Iran.
\n
We were able to find the Farsi Telegram account and exact address of the dealer, another answer to the question as to whom our target was contacting.
\n
A summary of our findings up until finding the specific knife reference on the target’s account was given to the police and later published in a Dutch newspaper, which had asked Bellingcat for assistance. I also presented the preliminary findings at a conference of the VVOJ.
In this video, Shahin Gheiybe mocks the police while standing in front of a house.
\n
More Technical Preparations
\n
We dissected the 35 seconds of video from March 9, 2019 and exported it into over 1200 separate images, each covering a different part of the area, resulting in just one photo of all of the frames the video. For that, we used a sophisticated program called Agisoft Metashape, and this method (the description is currently only available in German). You see our target more than once because he was moving in the frame.
\n
The reason behind doing this conversion is that we get a better view of all the details of the video, a method that was also used in “Inside the trenches of an information war”
\n
\n
Problems To Solve
\n
Our target posted the video on March 9, 2019, but he may have filmed it months ago.
\n
We needed to find something in the video that helped with chronolocation, i.e., finding out exactly when it was shot.
\n
Besides the when, we needed more proof of the where: There is no hint of a location and no geotag in the Instagram post itself.
\n
Staring At The Magnolias
\n
\n
We stared long and hard at the white flowers, but only thanks to botanical experts was our next lead found.
\n
The plant is called the Magnolia stellata. It bears large, showy white or pink flowers, most of them in early spring, before its leaves open. These flowers don’t last long. The magnolia from the video probably blossomed at the end of February or in early March, very near to the publication date of the video. It was probably shot in the northern part of Iran, because Magnolia stellata wouldn’t be comfortable with the climate in the south of the country. We believed it to be a wealthy neighbourhood, because of the well-maintained garden and the size of the house.
\n
The Garbage Can
\n
\n
The garbage bin in front of the house is of a color and model that is sold in Iran.
\n
\n
We were able to sharpen the Iranian (?) bin. Now compare this bin to some bins we are sure that they are from Iran. Can they be the same? What do you think? (7/7) pic.twitter.com/NlOo9iEOUR
The details from the balcony are typical of some buildings in the Iranian province of Mazandaran, local people told us via Twitter
\n
Going Through Reference Material
\n
We went back to the Instagram videos of Shahin Gheiybe looking for material that could be from Mazandaran. In one video, published on October 5, 2018, he is driving through the rain.
\n
\n
In front of him is a car, to the left of him you can see a sign of some sort of shop (1). The shop is a real estate firm in Abbasabad, Mazandaran. The license plate (2) is also local. It starts with the number 82, which is specific to the area, as Iranian journalist Ershad Alijani from France 24 Observers found out.
\n
\n
We have a new clue. In one of his videos, this still can be found. The real estate firm to the left is from Abbasabad . And the license plate is also from Abbasabad/ Tonekabon Can all #osint lovers research for a house near Abbasabad that matches the house on the right? (8/8) pic.twitter.com/eFiY61orSk
The architectural style we previously honed in on is present in Abbasabad. At this point, Twitter was going wild. Some Iranians started spontaneously hunting for magnolias and orange garbage cans in the area, by car.
\n
The mystery, however, was solved in London. Nathan Lee, a reporter of ITV, recently followed a workshop from Bellingcat with Eliot Higgins and the author of this piece. “The Bellingcat course helped a lot. I did various Google image searches of Abbasabad, Tonekabon, Amirdasht and went on the premise of it being a rich/well to do area of Iran,” he wrote. Just 15 kilometres from Abbasabad, in Amirdasht Town, Nathan spotted the house of the convicted fugitive.
Matching pillars, lamp outside directly above doorway and to the side, shrubbery, plinths outside which can be seen at the end of the video. pic.twitter.com/Wf5mlGAyEC
The houses in the area are not vacation homes, and the neighbourhood consists of people who own their own places.
\n
Once more we went back to older Instagram posting. We wanted to find out if some of the videos were shot inside the house we saw in his Instagram-feed. We found three videos that were probably made in that house because they matched the placement of windows and other details we saw on the exterior.
\n
We found similar footage on another Instagram account. That account belongs to a friend of the fugitive. She is a lawyer. The fugitive attended this friend’s wedding, according both to his and her postings. The groom is Shahin Gheiybe’s best friend. The bride claims it’s her house. Therefore, our target most likely lived at a friend’s house.
\n
\n
@bellingcat@henkvaness Short movie placed on Shahin Gheiybe instagram this morning was filmed in a female's friend house (same as the wedding pictures). See attachment for proof. pic.twitter.com/BLMbmcV5eQ
Now that we knew how the inside of the house looked, we could chronolocate this Instagram post, from Christmas 2018, in the same house. Therefore, our target has probably lived for months in that house.
\n
\n
The wanted fugitive has since confirmed that the location is right:
Can you trust the statement of a wanted convict? We don’t need to answer that question ourselves. We found the exact spot thanks to open source investigation. But what now?
\n
Epilogue
\n
Gathering all of our findings, we went to the Dutch police. They were very happy with the information and used it in their daily briefing, when they decided what to do next. But their hands are tied.
\n
The Dutch Public Prosecutor’s office then announced that they would not be able to arrest our wanted convict: “We have no extradition treaty. We studied our options, but we don’t have any legal basis to arrest him or ask the Iranian government to do that.”
\n
While this answer may be underwhelming, our quest demonstrates how social media and crowdsourcing can help pinpoint a criminal’s location. Even if said criminal thought he was playing a clever cat-and-mouse game.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nA fugitive convict from the Netherlands is\ntaunting\nDutch police with provocative pictures and videos from Iran, playing a game of “catch me if you can” on Instagram. “I love my freedom. And the weather,” he has stated.\nBellingcat was able to\nreveal\nthe last known location of the criminal with the help of over 60 Twitter users. Here’s how we did that.\nShahin Gheiybe is a convicted criminal hiding in Iran\nThe escape\nThe fugitive, 35-year-old Shahin Gheiybe, is on the Dutch “\nMost Wanted\n” list. On July 5, 2011, Gheiybe was sentenced for two attempted murders. In the city of ‘s-Hertogenbosch, on July 6, 2009, he tried to kill two individuals by shooting at them. He also robbed them of 175,000 euros. Gheiybe, born in Iran, was sentenced to a term of imprisonment of thirteen years.\nHe managed to escape from prison on October 6, 2011, and has been on the run ever since. The police in the Netherlands have searched for this individual since\nthen\n.\nThe convict posts on Instagram (there are over 170 photos and videos) about his whereabouts (here is the\noriginal\nlink to his private account, and a\nbackup\nlink), mostly in Iran. He has almost never tagged himself with a specific geolocation, just towns and places. Only once, for a street in Teheran, the coordinates were very precise. Gheiybe’s Instagram account was created in July 2015. It became\nprivate\nthis month, after he landed on the list of most wanted criminals.\nHis last Instagram post is a clip from Scarface (\narchived\npost). Gheiybe’s Instagram account contains provocative material, like personal videos accusing the ruling judge in the Netherlands, claiming “the truth is very different.” He tries to explain to his followers he is a nice guy who made a mistake.\nA Dutch police spokesperson\ncalled this “shocking”\nand added, “It is unbelievable how he distorts the facts and that he cares so little about his victims and our society. Apparently he considers himself untouchable in Iran.”\nProblems To Solve\nA problem with scrutinising any of the Instagram posts is that the convict can deceive the police by using bogus location tags in Instagram, backdating the information, or otherwise hiding his real whereabouts. We looked mostly at clues where the convict is clearly visible and, at the same time, is near a traceable object (Gheiybe near a car, in a club, holding a branded knife).\nUsing that method, we tried to reduce the risk of being deceived ourselves.\nAnother problem centered on the ethics of impacting the investigation. If we can find a last known location — as we ultimately did — any convict undoubtedly becomes aware that they are being zeroed in on.\nWithout crowdsourcing, any fugitive’s last address will be much harder to find. Bellingcat got the help of over 60 people who researched with us via\nTwitter\n. That research was conducted mostly on March 18, 2019.\nThe police told the public they didn’t know the location. But the public (via crowdsourcing) ultimately did.\nThis is an example of how transparency works both ways: It can be used by both good and bad people. We decided that we would send the police even the tiniest leads during our investigation. And that we would alert them the moment we would find something actionable, so they had at least a head start.\nTechnical Preparations\nTo get our hands on all the Instagram posts, we needed to prepare first. We followed these three steps:\nBrowse Instagram On Desktop\nWe didn’t want to fiddle with a mobile or tablet, but work on a desktop computer. To do that, we downloaded\nDesktop for Instagram,\na Chrome plug-in.\nDownload Videos\nDownloading over 170 Instagram videos and photos manually is time consuming. We used a plug-in for Chrome called\nDownloader for Instagram\n. It’s fast, downloads all material in full resolution, and also supports Instagram stories downloads.\nFind Hidden Connections\nTo find the hidden connection between Instagram users, to find what our target liked, and much more, we used\nHelper Tools for Instagram\n, again a plug-in for Chrome.\nIn the picture above, for example, the convict is next to a liquor display — focusing in on specific objects can sometimes lead to finding connections. Or, at the very least, you find something about the target’s likes and lifestyle.\nQuestion 1:\nWhom Did He Meet?\nLast week, Dutch police asked the Dutch public for help with this case on TV.\nFor English followers: translation of \"Opsporing Verzocht\" (17/17) so you understand my last tweets a little better. Full explainer coming up tonight.\npic.twitter.com/3aplJ1Yz6F\n— ʜᴇɴᴋ ᴠᴀɴ ᴇss (@henkvaness)\nMarch 19, 2019\nOne question was: Who did the convict interact with? Maybe his interlocutors know his possible bogus identity, like a false passport or a false name. We concentrated on that question first.\nA Man And A Statue:\nThis picture, uploaded on November 27, 2018, shows our target play-fighting a statue, which perfectly suits our need to have an identifiable object and our target in the same photo. We did\nreverse image search\nin Google and Yandex — and, as it is often the case, Yandex was the better choice.\nThis is how we found\nKish Island\n, some 200 kilometers away from Dubai.\nA slightly\nbizarre\npromotional story about this island tells you what you can do there:\n“Women sit with cigarette in hand, wearing colorful headscarves pushed right back to reveal plenty of make-up and expensive hair-dos” and “The wide, palm tree-lined boulevards that circle the island of 100 square kilometers are full of top-end cars, including luxury American models.”\nOn the same day, November 27, 2018, we found a video where our target was near a car with a recognizable license plate. The video shows our target getting into a red car and driving away in it.\nThe next step was for people with a great amount of patience. We tried to find hotels in Kish via Google Streetview that matched the building in the video background. In the end, we found the exact\nspot\n(pictured here in daylight):\nReverse imaging the license plate didn’t work at first. Sometimes, Google simply doesn’t recognize a blown up detail of a photo.\nYou can, however, approach this problem by just typing in words you can see into Google Images, without reverse searching the photo itself:\nThis leads to images of tourists driving the same red Chevrolet Camaro with the same license plate.\nEventually, we found the Camaro supplier by clicking on some Instagram posts featuring happy clients.\nOur convict didn’t steal the car. After he drove the Camaro, people were still renting it, as we found out after checking another series of posts by satisfied renters. So, we assume he had rented the vehicle as well.\nKish Rent Car, a local car rental place, can shine a light on the current identity of our target. We have a first possible answer to our initial question: Whom did our target meet?\nFinding Replies Sent To Other Instagram Accounts\nDid our target reply to other instagrammers? To find out, we used this formula in Google.\nWe don’t want to see postings from the target’s own account (hence we exclude site:\ninstagram.com/shahin\n) but are still interested if he (hence\nshahin.mzr\n) posted comments on the Instagram posts of others (hence\nsite:instagram.com\n). This is what we found:\nAround August 3, 2018, our target asks\nknife._.store\n, the Instagram account of an Iranian knife seller, if he can buy a knife and if it can be sent to him. He was prepared to pay 65,000 Thoman, which is around around $15, for the knife.\nLater that year, on October 22, 2018, Gheyibe posted a photo of two knives on his own Instagram account. The one to the left (see photo above) is the Phrobis III Buck Gen 4 Fighting Knife M9 Scabbard USGI Military Wire Cutter. This model had gone on sale for a short moment in the autumn of 2018, at the previously mentioned knife store\nin Iran.\nWe were able to find the Farsi Telegram account and exact address of the dealer, another answer to the question as to whom our target was contacting.\nA summary of our findings up until finding the specific knife reference on the target’s account was given to the police and later published in a\nDutch newspaper,\nwhich had asked Bellingcat for assistance. I also presented the preliminary findings at a conference of the\nVVOJ\n.\n#LROJ19\n: scoop van Bellingcat\nhttps://t.co/IQkmPrieUE\n— Bellingcat (@bellingcat)\nMarch 19, 2019\nQuestion 2: Where Is He Now?\nBecause things were going smoothly, we tackled the hardest question: Where is the target now? We concentrated on a video from March 9, 2019.\nHere is the original footage. And no – I am not sure. Why does it look like further north?\npic.twitter.com/syEgIRW87K\n— ʜᴇɴᴋ ᴠᴀɴ ᴇss (@henkvaness)\nMarch 18, 2019\nIn this video, Shahin Gheiybe mocks the police while standing in front of a house.\nMore Technical Preparations\nWe dissected the 35 seconds of video from March 9, 2019 and exported it into over 1200 separate images, each covering a different part of the area, resulting in just one photo of all of the frames the video. For that, we used a sophisticated program called Agisoft Metashape, and this\nmethod\n(the description is currently only available in German). You see our target more than once because he was moving in the frame.\nThe reason behind doing this conversion is that we get a better view of all the details of the video, a method that was also used in “\nInside the trenches of an information war\n”\nProblems To Solve\nOur target posted the video on March 9, 2019, but he may have filmed it months ago.\nWe needed to find something in the video that helped with chronolocation, i.e., finding out exactly when it was shot.\nBesides the when, we needed more proof of the where: There is no hint of a location and no geotag in the Instagram post itself.\nStaring At The Magnolias\nWe stared long and hard at the white flowers, but only thanks to\nbotanical\nexperts was our next lead found.\nThe plant is called the\nMagnolia stellata\n. It bears large, showy white or pink flowers, most of them in early spring, before its leaves open. These flowers don’t last long. The magnolia from the video probably blossomed at the end of February or in early March, very near to the publication date of the video. It was probably shot in the northern part of Iran, because\nMagnolia stellata\nwouldn’t be comfortable with the climate in the south of the country. We believed it to be a wealthy neighbourhood, because of the well-maintained garden and the size of the house.\nThe Garbage Can\nThe garbage bin in front of the house is of a color and model that is sold in Iran.\nWe were able to sharpen the Iranian (?) bin. Now compare this bin to some bins we are sure that they are from Iran. Can they be the same? What do you think? (7/7)\npic.twitter.com/NlOo9iEOUR\n— ʜᴇɴᴋ ᴠᴀɴ ᴇss (@henkvaness)\nMarch 18, 2019\nA Reflection Offers More Clues\nIf you look closely at the house in the background, you see a window:\nAnd there is a\nreflection in the window as shown here.\nThere is another house in that reflection.\nWe had to focus on a wealthy area in the north of Iran with at least three houses in the middle of what looks like the woods.\nThe Architectural Style\nAnd our public\n@bellingcat\nquest continues! Look at the architecture of the building in video compared with something\n@Tohn_Jitor\nfound in this area\nhttps://t.co/IUcE9xo2x7\nCan you help to find the building in first photo? (7/7)\npic.twitter.com/aNog8jEq9Y\n— ʜᴇɴᴋ ᴠᴀɴ ᴇss (@henkvaness)\nMarch 18, 2019\nThe details from the balcony are typical of some buildings in the Iranian province of Mazandaran, local people told us via Twitter\nGoing Through Reference Material\nWe went back to the Instagram videos of Shahin Gheiybe looking for material that could be from Mazandaran. In one video, published on October 5, 2018, he is driving through the rain.\nIn front of him is a car, to the left of him you can see a sign of some sort of shop (1). The shop is a real estate firm in Abbasabad, Mazandaran. The license plate (2) is also local. It starts with the number 82, which is specific to the area, as Iranian journalist\nErshad Alijani\nfrom\nFrance 24 Observers\nfound out.\nWe have a new clue. In one of his videos, this still can be found. The real estate firm to the left is from Abbasabad . And the license plate is also from Abbasabad/ Tonekabon Can all\n#osint\nlovers research for a house near Abbasabad that matches the house on the right? (8/8)\npic.twitter.com/eFiY61orSk\n— ʜᴇɴᴋ ᴠᴀɴ ᴇss (@henkvaness)\nMarch 18, 2019\nThe architectural style we previously honed in on is present in Abbasabad. At this point, Twitter was going wild. Some Iranians started spontaneously hunting for magnolias and orange garbage cans in the area, by car.\nThe mystery, however, was solved in London.\nNathan Lee\n, a reporter of ITV, recently followed a workshop from Bellingcat with Eliot Higgins and the author of this piece. “The Bellingcat course helped a lot. I did various Google image searches of Abbasabad, Tonekabon, Amirdasht and went on the premise of it being a rich/well to do area of Iran,” he wrote. Just 15 kilometres from Abbasabad, in Amirdasht Town, Nathan spotted the house of the convicted fugitive.\nHi\n@henkvaness\n,\n@bellingcat\n@EliotHiggins\n– I think I've found it. Amirdasht in Iran. 36.691453, 51.258657\nMatching pillars, lamp outside directly above doorway and to the side, shrubbery, plinths outside which can be seen at the end of the video.\npic.twitter.com/Wf5mlGAyEC\n— Nathan Lee (@NathanLeeTV)\nMarch 18, 2019\nIf you turn the 360 view around, to the house across the street, there is a clear view of the same orange bin we found earlier.\nWith the help of crowd sourcing, the second question was finally answered. I couldn’t find the right words, as evidenced by my typo:\nI am tremendous proud of all the\n#OSINT\npeople who helped\n@bellingcat\nthis morning in finding a fugitive criminal person. The exact location is\nhttps://t.co/FVlIweW1wF\nand the police has been notified. Thanks all! Unsung heroes are\n@ErshadAlijani\nand\n@NathanLeeTV\n! (10/10)\npic.twitter.com/bxmqtKNHkT\n— ʜᴇɴᴋ ᴠᴀɴ ᴇss (@henkvaness)\nMarch 18, 2019\nWe asked a local Iranian to find out more about the house. It is guarded.\nThe dessert; here is the entrance of the place were the criminal fugitive is hiding (12/12)\npic.twitter.com/CfpZgS5W7W\n— ʜᴇɴᴋ ᴠᴀɴ ᴇss (@henkvaness)\nMarch 18, 2019\nThe houses in the area are not vacation homes, and the neighbourhood consists of people who own their own places.\nOnce more we went back to older Instagram posting. We wanted to find out if some of the videos were shot inside the house we saw in his Instagram-feed. We found three videos that were probably made in that house because they matched the placement of windows and other details we saw on the exterior.\nWe found similar footage on another\nInstagram\naccount. That account belongs to a friend of the fugitive. She is a lawyer. The fugitive attended this friend’s wedding, according both to his and her postings. The groom is Shahin Gheiybe’s best friend. The bride claims it’s her house. Therefore, our target most likely lived at a friend’s house.\n@bellingcat\n@henkvaness\nShort movie placed on Shahin Gheiybe instagram this morning was filmed in a female's friend house (same as the wedding pictures). See attachment for proof.\npic.twitter.com/BLMbmcV5eQ\n— Erik Wassenaar (@ErikLambooy)\nMarch 21, 2019\nNow that we knew how the inside of the house looked, we could chronolocate this Instagram post, from Christmas 2018, in the same house. Therefore, our target has probably lived for months in that house.\nThe wanted fugitive has since confirmed that the location is right:\nFor what it's worth! Dutch most wanted fugitive just said about\n@bellingcat\narticle \"Yes, they found the exact location\" in a chat with one of the editors. Check also\nhttps://t.co/qwtglUn1HE\n(17/17)\npic.twitter.com/tCet2hEbiJ\n— ʜᴇɴᴋ ᴠᴀɴ ᴇss (@henkvaness)\nMarch 21, 2019\nCan you trust the statement of a wanted convict? We don’t need to answer that question ourselves. We found the exact spot thanks to open source investigation. But what now?\nEpilogue\nGathering all of our findings, we went to the Dutch police. They were very happy with the information and used it in their daily briefing, when they decided what to do next. But their hands are tied.\nThe Dutch Public Prosecutor’s office then announced that they would not be able to\narrest\nour wanted convict: “We have no extradition treaty. We studied our options, but we don’t have any legal basis to arrest him or ask the Iranian government to do that.”\nWhile this answer may be underwhelming, our quest demonstrates how social media and crowdsourcing can help pinpoint a criminal’s location. Even if said criminal thought he was playing a clever cat-and-mouse game.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/2019-03-19_13-01-50.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/Picture1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/Picture3.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/22-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/2019-03-19_19-38-31.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/Picture4.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/5-3.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/locati.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/as.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/at.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/al.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/ee.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/or.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/ob.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/ol.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/panorama2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/qq.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/qw.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/qa.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/qeeq.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/ee33.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/1313.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/03/christmas.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2019-07-02_more-europols-stop-child-abuse-photographs-geolocated.json b/bellingcat_geolocation_json/2019-07-02_more-europols-stop-child-abuse-photographs-geolocated.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e95d9a6023d40b707c3526ebe0c82e530e6f50d
--- /dev/null
+++ b/bellingcat_geolocation_json/2019-07-02_more-europols-stop-child-abuse-photographs-geolocated.json
@@ -0,0 +1,97 @@
+{
+ "slug": "more-europols-stop-child-abuse-photographs-geolocated",
+ "title": "More Europol’s “Stop Child Abuse” Photographs Geolocated",
+ "url": "https://www.bellingcat.com/news/europe/2019/07/02/more-europols-stop-child-abuse-photographs-geolocated/",
+ "date": "2019-07-02T14:39:16+00:00",
+ "modified": "2019-12-06T21:41:22+00:00",
+ "author": "Daniel Romein",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "In November 2018, Bellingcat wrote about two photographs related to child abuse crimes – photographs which were geolocated to Shenzhen in China. This investigation was crowdsourced by Europol as part of their Stop Child Abuse campaign that has been running since 2017. Images are shared on their website with the goal to identify the objects […]",
+ "content_html": "
\n
In November 2018, Bellingcat wrote about two photographs related to child abuse crimes – photographs which were geolocated to Shenzhen in China.
\n
This investigation was crowdsourced by Europol as part of their Stop Child Abuse campaign that has been running since 2017. Images are shared on their website with the goal to identify the objects in them, their country of origin, and/or the location where they were taken. Europol also shares these images on their Twitter and Facebook accounts. Sites like Reddit and Websleuths will also share these images with the purpose of discovering their origin.
\n
Both the campaign and the crowdsourcing efforts around it have not gone unnoticed by the media. The Guardian, for example, has written about Bellingcat’s involvement in this campaign. BBC News broadcast an interview with Finnish architect Olli Enne, who geolocated the two photographs from a city in East Asia, originally shared by Europol in October 2018.
\n
On 1 February 2019, Europol shared new photos. Two of these photos were shared in a separate tweet. Just 23 minutes later, a Dutch citizen named Maarten Kaskens posted a reply with a photograph of the same place and mentioned the exact location, a cluster of summer homes between Moscow and Dmitrov, in a later tweet.
\n
Soon, these photographs were removed from Europol’s website, likely because the right location was found, but also because the photographs directly led to the person who uploaded them. An account with geotagged photographs uploaded by this individual, taken in the same area, was deleted soon after Europol’s publication. The Bellingcat team investigated the photographs and the individual who uploaded them. We submitted details about the individual in question to Europol, but because of the ongoing criminal investigation into this suspect, we cannot reveal these details and can only discuss the photographs themselves.
\n
Europol shared another photograph from their 1 February 2019 publication in a separate tweet on 22 February 2019, asking if anyone was able to recognize the city visible in the background. Shortly thereafter, an Italian citizen named Lorenzo Romani wrote that he recognized the city as St. Petersburg in Russia and uploaded a skyline photograph with a bridge that roughly matches the skyline visible in Europol’s photograph. Bellingcat found the exact location of the photograph, taken from a beach area near St. Petersburg.
\n
Europol’s photographs from 1 February 2019 were also shared by Bellingcat. On the same day, several people responded to a photograph taken presumably in an East Asian city and recognized the writing on the building as Chinese and mentioned South China as a possible location because of palm trees visible in the photograph. Only the text on two billboards was clear enough for native Chinese speakers to read because of the photograph’s resolution. These billboards mentioned used car information and a brand of motorcycles from China. Further reactions to the tweet mentioned the island Hainan, Zhuhai, Shenzhen, Zhongshan and Jiangmen as possible locations. After searching in several cities in Southern China for streets with the same road lay out, street lanterns and palm trees, Bellingcat managed to find the exact location where the photograph was taken – in Jiangmen, China.
\n
What follows below is a recap of how these aforementioned photographs were geolocated.
\n
\n
Location One: “Sunny Glade”, Yazikovo, Dmitrovsky district, Russia
\n
The photograph presumably taken between Moscow and Dmitrov shows small houses and a pond in a rural area. Photographs uploaded to a website in 2003, show several similar houses and a pond – it obviously is the same area. One of these photographs, as tweeted by Maarten Kaskens, shows rocks with exactly the same paint marks as visible in one of Europol’s photos. During our research it became clear the photographs very likely were uploaded by the suspect in Europol’s case.
\n
After comparing the photographs in the album to Europol’s website and first tweet from 1 February, it became clear that another photograph shared by Europol is from the same area. The small houses visible in the photographs are Russian summer houses (commonly referred to as “dachas”). Bellingcat discovered that the person who uploaded the 2003 photo album was registered to an address in Moscow in 2012, but also noticed an address of this person in Dmitrov mentioned in a discussion forum.
\n
Left: Houses and a pond in a rural area, from Europol’s website and separate 1 February tweet. Right: Another photograph from Europol’s website and a first 1 February tweet.
\n
Left: Photograph from the 2003 photo album, showing stones with the same paint marks as one of Europol’s photographs. Right: Another photograph from the 2003 photo album, showing the same houses as another Europol photograph.
\n
Bellingcat also found a 2014 video, showing the same houses near the pond, as visible in the second photograph of Europol’s separate tweet from 1 February. Two individuals, who do not have anything to do with Europol’s case, and visible in this video have been cut out from the following screenshot.
\n
Screenshot of the 2014 video (we left out the two people pictured), matching the second photograph of Europol’s separate tweet.
\n
To confirm the exact location of these summer houses, we searched for additional evidence that supports the location of the deleted geotagged photographs. That evidence was found in a YouTube video, filmed from a car that drove from Moscow in the direction of Dmitrov, showing the route to the summer house area.
\n
At one point in the video, when the car leaves the main road, a store named “minimarket Olgovo” is visible – its location is visible on Yandex street view. Apparently, a part of the video was cut out from that moment on, as a few seconds later a small old church is visible. This church was be geolocated to a village named Yazykovo, a short distance from Olgovo.
\n
Following the road further south, the same road as is visible in the video eventually brings us to a cluster of summer houses named Solnechnaya Polyana (literally: “Sunny Glade”), owned by an organisation named SNT Solnechnaya Polyana, which runs many similar summer house communities mainly near Moscow.
\n
Two screenshots from a video that shows the route from Moscow to the summer house community from Europol’s photographs. Left: The church and houses in Yazykovo, near Olgovo in Dmitrovsky district. Right: The entrance of the Solnechnaya Polyana summer house community near Yazykovo.
\n
Eventually, the video also shows the exact location of the summer house of the suspect. This house matches with the 2003 photographs of the summer house. Bellingcat found the exact location of the summer house on satellite imagery and sent these details to Europol, but has chosen to refrain from publishing the exact location, due to the sensitivity of this criminal investigation.
\n
\n
Location Two: Olgino, Primorsky District, St. Petersburg Region, Russia
\n
We first added more contrast to the photograph, which presumably was taken in St. Petersburg, so the skyline would be more visible. This enhanced version shows four pylons on the left side of the photograph, by which Lorenzo Romani recognized the skyline of St. Petersburg. Although we found many photographs with matching skylines, we noticed something was not completely right.
\n
The most similar view we first found on Yandex streetview shows a different bridge on the left side of buildings similar to the buildings in Europol’s photograph, but that bridge has only two pylons, while in Europol’s photograph there are or there seem to be four pylons visible. The bridge with four pylons, as visible in Lorenzo’s tweet, is to the right side of these buildings. Also, the distance from the four pylons to the matching buildings in Europol’s photograph is bigger than in Yandex street view. Some of the buildings on Yandex street view are not visible on Europol’s photograph.
\n
Then, we realized two things: Europol’s photograph was likely taken from a different direction; also, the photograph was a few years old and the skyline has changed during these years. Searching through older photographs and satellite imagery to check if the bridge might have been renovated, we found out that the bridge was built from 2014 to 2016 and the two pylons were constructed in 2015.
\n
Enhanced version with more contrast of Europol’s photograph.
\n
Comparing the skyline in Europol’s photograph to Yandex streetview shows similarities, but also a few differences. The four pylons in Europol’s photograph are further away from the buildings than the bridge and some buildings seem to be missing in Europol’s photograph.
\n
Older photographs of the area show the same skyline without the bridge and several other photographs. Also, Yandex street view shows a soccer stadium to the left side of the bridge with several light poles.
\n
However, satellite imagery shows the stadium has eight light poles instead of four and the light poles stand skew instead of straight as in Europol’s photograph. Older satellite imagery makes clear that construction of the current stadium, named Gazprom Arena or Krestovsky Stadium, started in 2009 at the location of the former Kirov Stadium, which was demolished in September 2006. The Kirov Stadium had four straight light poles, matching the four poles visible in Europol’s photograph.
\n
Top: Skyline from 2010 without bridge and construction work to the left. Middle: Skyline from 2007 that closely matches Europol’s photograph skyline. Bottom: Skyline from 2018 with bridge and soccer stadium.
\n
Photographs taken from the beach area around the Thunder Stone (“Гром Камень” in Russian), a large megalith used in the 18th century as pedestal for the Bronze Horseman in St.Petersburg, seem to have the best matching skyline. In a photo album connected to a 2004 article about the Thunder Stone, a photograph that was uploaded in October 2004 shows almost the same skyline as visible in Europol’s photograph, including the four light poles of the old soccer stadium. An overlay of Europol’s photograph over the 2004 photograph shows the striking similarity of the skyline.
\n
Top: 2004 photograph of the “Thunder Stone” beach area showing a very similar skyline as Europol’s photograph. Bottom: Overlay of Europol’s photograph over the 2004 Thunder Stone beach area photograph.
\n
In Europol’s photograph it seems a few taller buildings are visible that are not visible in the 2004 Thunder Stone beach area photograph. This can be explained because construction of a residential complex named Morskoy Kaskad (literally: “Sea Cascade” in Russian) started in 2003 and was finished in 2005. Therefore, Europol’s photograph very likely was taken between spring and autumn 2005, or in the spring or summer of 2006.
\n
\n
Location Three: Jiangmen, Xinhui district, Guangdong province, China
\n
One of the locations published by Europol on 1 February, which was not tweeted about separately, is accompanied by the question, “Do you recognize this building in Asia?”
\n
In response to this tweet, as well as to Bellingcat’s tweet, several people responded by saying that they recognize the text displayed on several billboards on the building as Chinese and that the location was very likely to be in South China, because of the palm trees visible in the photograph. Only the text on two billboards is clear enough for Chinese speakers to read because of the photograph’s low resolution. The billboards mention used cars and a brand of motorcycles from China. As previously mentioned, further reactions to the tweets specifically mention the island Hainan, Zhuhai, Shenzhen, Zhongshan and Jiangmen as possible locations in South China.
\n
Building in South China with two readable billboards, displaying the name of a brand of motorcycles from China (bottom left) and “used car information” (bottom right). The image obviously has been edited/censored by Europol, because very likely a child was visible in the original image.
\n
Twitter user Ben Lokshin, among others, deciphered the motorcycle brand billboard as “中裕摩托“, written in simplified Chinese characters and translated as Zhongyu Motorcycles. Another Twitter user named Konrad Iturbe pointed out this brand is sold all over China, but also mentioned that only Jiangmen would have this kind of architecture, road layout, and palm trees.
\n
A Twitter user named James noticed a photograph of a different building with the same first four Chinese characters displayed on a billboard – a logo for a motorcycle repair service department. That repair service deparment, however, is on Taoyuan west road 35 in Mabazhen in the Qujiang district, Shaoguan, Guangdong province, more than 100 kilometers north of the other aforementioned cities in Southern China. Both Google as Baidu, showing yet another building with the same four Chinese characters on a billboard, give Hiuanshi 3rd road, Penjiang district in Jiangmen city, Guangdong province in Southern China as the first search result first “中裕摩托” (Zhongyu motorcycles).
\n
Nathan Ruser from the Australian Strategic Policy Institute, among others, recognized the text on the second billboard as simplified Chinese characters “二手车信息”, translated as “used car information.”
\n
Zhongyu motorcycle repair service department in Mabazhen, Qujiang district, Shaoguan, Guangdong province, China (left), Zhongyu motorcycle accessories in Penjiang district, Jiangmen, China (right).
\n
Further research into the brand name “Zhongyu motorcycles” made clear the company that produced these motorcycles went bankrupt in 2006 and the owner fled China, despite being ranked first in 2005 as motorcycle production in China and that everyone in Jiangmen knew the factory closed down.
\n
The location of the factory, which is only correctly displayed on the Chinese version of Google maps, was on Jinlu Road in the Jianghai District of Jiangmen City. The Zhongyu motorcycle group was established in 1994 and produced 600,000 motorcycles a year on the 420-acre industrial park in Jiangmen. A 2002 archived version of Zhongyu motorcycles’ website mentions the engines were produced in the Wuhe indsutrial zone in Daze town in the Xinhui district of Jiangmen.
\n
Because Zhongyu went bankrupt in 2006, we suspected the Europol photograph to be at least ten years old, as after 2006 very likely no new advertisements of this motorcycle brand would appear anymore on billboards. Since these motorcycles were produced in Jiangmen, we knew there was a chance that Europol’s photograph was taken in Jiangmen, despite that these motorcycles likely were sold everywhere in China – or at least everywhere in South China. Our research started in various cities in South China and we especially focused to the layout of the road visible in the photograph: a two-way road with three lanes each, separated by a centre median strip with palm trees and lamp posts with four lights.
\n
After an intensive research, we finally found such a road and a similar building in the Xinhui district of Jiangmen, named Sanhe Avenue. However, the building from Europol’s photograph was not found along this road. As satellite imagery in Jiangmen doesn’t go back further in history than 2008 and Europol’s photograph was likely taken in or before 2006, we realized there was a chance the building was demolished before 2008. As we expected to find the building along this north-south road, on the west side of the road based on the direction of the sunlight, Timmi Allen of our Bellingcat team created a 3D model of the building to understand how the building would look like on satellite imagery.
\n
3D reconstruction model of Europol’s Asian city photograph, showing recognizable features of the building.
\n
Via this reconstruction model we eventually found the building on Ganzhou Avenue East 88 in the Xinhui district of Jiangmen, China. According to geotags on Baidu and Google, the building houses, among other businesses, a used car information center and a business hotel, showing the same entrance on the hotel website as Europol’s photograph.
\n
Baidu’s street view from 2015 even shows the exact same lam post and blue one-way traffic sign on the centre median strip with palm trees. Examination of satellite imagery clearly shows Europol’s photograph was taken before 2014, as a building to the right of Europol’s building has been demolished before 2014 and a new building to the left was built before 2014. Also the height of the trees visible in Europol’s photograph fit the 2008 satellite imagery.
\n
The building from Europol’s Asian city photograph on Google Earth satellite imagery of 2018 (top left) and satellite imagery from 2008 (top right). Baidu’s street view from 2015 of the same building and road visible in Europol’s photograph.
\n
The photograph that Europol published was taken from behind a window of the building on the opposite side of the street. Europol’s photograph shows a black fence, big windows and two white pillars on the other side of the window. Also a white roof is visible on the bottom side of the photograph. The photograph clearly shows this part of the building has a round shape. A white cross and a car on the left side are visible outside in the area which seems a parking place. Baidu’s streetview shows a part of the building with a round shape and a roof above the entrance. On the second floor of this part of the building, a fence is visible behind the only window that is not covered by a curtain.
\n
The street view to the third floor of the other side of this building shows a black fence with exactly the same shape as in Europol’s photograph. However, since a large part of the roof above the entrance is visible, it’s more likely the photograph has been taken from the second floor, from one of the rooms that, on street view, has its windows covered by curtains. Photographs taken in meeting rooms of the building, which in the meantime have been deleted, also show the exact same type of fence.
\n
The building where Europol’s photograph was taken. A similar fence is visible on the second floor (left) and the exact type of fence on the third floor (right). Large windows, white pillars and a white roof above the entrance are similar to Europol’s photograph.
\n
Photographs taken from meeting rooms in the building where Europol’s photograph was taken show the exact same type of black fence as in Europol’s photograph.
\n
Further investigation into the building where Europol’s photograph was taken, led to the surprising conclusion that another photograph, published by Europol on 1 February 2019, was taken inside that same building. This photograph, showing a hall of what looks like a shopping center and several red tables and benches with “Coca Cola” written on them, is accompanied by the question “Do you know where this food court is located?”
\n
An article about the Jianghui Times City (archived here), published in July 2013, shows a photograph of a central hall of this shopping mall, which shows many similarities to Europol’s food court photograph. Another article about the same shopping center published in 2014 (archived here), confirms the location.
\n
Another photograph published by Europol on February 1, 2019 shows a hall inside what looks like a shopping mall, with several red tables and benches visible in the photograph with the Coca Cola logo displayed on them (left). Photograph of the hall of the Jianghui Times City shopping mall, the building from where Europol’s Asian city photograph had been taken from (right).
\n
Comparison between another photograph of the main hall of the Jianghui Times City shopping mall to Europol’s food court photograph of 1 February 2019.
\n
\n
Conclusion And Recommendations To Europol \n
\n
This analysis demonstrates that Europol’s “Stop Child Abuse” online crowdsourcing campaign is quite succesful. It’s not only that objects in photographs get recognizedoften, so that they can be linked to an origin country making it possible for Europol to inform law enforcement of said country in order to further investigate the case. Sometimes, often after intensive research, locations of photographs eventually get geolocated precisely, possibly increasing the likelihood that such a case can be solved.
\n
However, we have a few recommendations to Europol in their crowdsourcing campaign. The first recommendation is to be careful with sharing photographs that can be easily found with reverse image search, as this could immediately lead to a victim or even a potential suspect. The risk for vigilante justice can be high, particularly if a social media profile belonging to a potential suspect or person of interest is uncovered. When it came to Location One, it was relatively easy to find one of the photographs with reverse image search, which happened just less than half an hour after one of the photographs was tweeted by Europol.
\n
The second recommendation is this: If several photographs from the same case are displayed on Europol’s website and Europol is aware of these photographs as being linked, it can help to make the connection between the photographs clear. Both in the case of Location One and Location Three it became clear Europol uploaded two different photographs that are part of the same case, without specifying it. It’s possible that this information could help solve the case possibly even quicker.
\n
Our final recommendation: Before a photograph is shared, it would be good to be realistic about the chances for recognition. Sometimes the quality of the photographs of objects seems too low to be able to be recognized and sometimes the location seems as if it could really be anywhere and could not be geolocated without any background information. This is why we recommend sharing as much information about a certain location as possible, in case that information is available. Our next article in the series of Europol’s “Stop Child Abuse” will show an example of a photograph that like would never have been geolocated if no additional information was found about the image.
\n
\n
Research by Daniel Romein, Timmi Allen, Klement Anders, Noor Nahas and “Bo”, review/editing by Natalia Antonova.
\n
",
+ "content_text": "In\nNovember 2018\n, Bellingcat wrote about two photographs related to child abuse crimes – photographs which were geolocated to Shenzhen in China.\nThis investigation was crowdsourced by Europol as part of their\nStop Child Abuse\ncampaign that has been running since 2017. Images are shared on their website with the goal to identify the objects in them, their country of origin, and/or the location where they were taken. Europol also shares these images on their\nTwitter\nand\nFacebook\naccounts. Sites like\nReddit\nand\nWebsleuths\nwill also share these images with the purpose of discovering their origin.\nBoth the campaign and the crowdsourcing efforts around it have not gone unnoticed by the media.\nThe Guardian\n, for example, has written about Bellingcat’s involvement in this campaign.\nBBC News\nbroadcast an interview with Finnish architect Olli Enne, who geolocated the two photographs from a city in East Asia, originally shared by Europol in\nOctober 2018\n.\nOn\n1 February 2019\n, Europol shared\nnew photos\n. Two of these photos were shared in a\nseparate tweet\n. Just 23 minutes later, a Dutch citizen named\nMaarten Kaskens\nposted a reply with a photograph of the same place and mentioned\nthe exact location\n, a cluster of summer homes between Moscow and Dmitrov, in a later tweet.\nSoon, these photographs were removed from Europol’s website, likely because the right location was found, but also because the photographs directly led to the person who uploaded them. An account with geotagged photographs uploaded by this individual, taken in the same area, was deleted soon after Europol’s publication. The Bellingcat team investigated the photographs and the individual who uploaded them. We submitted details about the individual in question to Europol, but because of the ongoing criminal investigation into this suspect, we cannot reveal these details and can only discuss the photographs themselves.\nEuropol shared another photograph from their 1 February 2019 publication in a separate tweet on\n22 February 2019\n, asking if anyone was able to recognize the city visible in the background. Shortly thereafter, an Italian citizen named\nLorenzo Romani\nwrote that he recognized the city as St. Petersburg in Russia and uploaded a skyline photograph with a bridge that roughly matches the skyline visible in Europol’s photograph. Bellingcat found the exact location of the photograph, taken from a beach area near St. Petersburg.\nEuropol’s photographs from 1 February 2019 were also\nshared by Bellingcat\n. On the same day, several people responded to a\nphotograph\ntaken presumably in an East Asian city and recognized the writing on the building as Chinese and mentioned South China as a possible location because of palm trees visible in the photograph. Only the text on two billboards was clear enough for native Chinese speakers to read because of the photograph’s resolution. These billboards mentioned used car information and a brand of motorcycles from China. Further reactions to the tweet mentioned the island Hainan, Zhuhai, Shenzhen, Zhongshan and Jiangmen as possible locations. After searching in several cities in Southern China for streets with the same road lay out, street lanterns and palm trees, Bellingcat managed to find the exact location where the photograph was taken – in Jiangmen, China.\nWhat follows below is a recap of how these aforementioned photographs were geolocated.\nLocation One: “Sunny Glade”, Yazikovo, Dmitrovsky district, Russia\nThe photograph presumably taken between Moscow and Dmitrov shows small houses and a pond in a rural area. Photographs uploaded to a website in 2003, show several similar houses and a pond – it obviously is the same area. One of these photographs, as tweeted by\nMaarten Kaskens\n, shows rocks with exactly the same paint marks as visible in one of Europol’s photos. During our research it became clear the photographs very likely were uploaded by the suspect in Europol’s case.\nAfter comparing the photographs in the album to Europol’s website and first tweet from 1 February, it became clear that another photograph shared by Europol is from the same area. The small houses visible in the photographs are Russian summer houses (commonly referred to as “dachas”). Bellingcat discovered that the person who uploaded the 2003 photo album was registered to an address in Moscow in 2012, but also noticed an address of this person in Dmitrov mentioned in a discussion forum.\nLeft: Houses and a pond in a rural area, from Europol’s website and separate 1 February tweet.\nRight: Another photograph from Europol’s website and a first 1 February tweet.\nLeft: Photograph from the 2003 photo album, showing stones with the same paint marks as one of Europol’s photographs.\nRight: Another photograph from the 2003 photo album, showing the same houses as another Europol photograph.\nBellingcat also found a 2014 video, showing the same houses near the pond, as visible in the second photograph of Europol’s separate tweet from 1 February. Two individuals, who do not have anything to do with Europol’s case, and visible in this video have been cut out from the following screenshot.\nScreenshot of the 2014 video (we left out the two people pictured), matching the second photograph of Europol’s separate tweet.\nTo confirm the exact location of these summer houses, we searched for additional evidence that supports the location of the deleted geotagged photographs. That evidence was found in a YouTube video, filmed from a car that drove from Moscow in the direction of Dmitrov, showing the route to the summer house area.\nAt one point in the video, when the car leaves the main road, a store named “minimarket Olgovo” is visible – its location is visible on\nYandex street view\n. Apparently, a part of the video was cut out from that moment on, as a few seconds later a\nsmall old church\nis visible. This church was be geolocated to a village named\nYazykovo\n, a short distance from Olgovo.\nFollowing the road further south, the same road as is visible in the video eventually brings us to a cluster of summer houses named\nSolnechnaya Polyana\n(literally: “Sunny Glade”), owned by an organisation named\nSNT Solnechnaya Polyana\n, which runs many similar summer house communities mainly near Moscow.\nTwo screenshots from a video that shows the route from Moscow to the summer house community from Europol’s photographs.\nLeft: The church and houses in Yazykovo, near Olgovo in Dmitrovsky district.\nRight: The entrance of the Solnechnaya Polyana summer house community near Yazykovo.\nEventually, the video also shows the exact location of the summer house of the suspect. This house matches with the 2003 photographs of the summer house. Bellingcat found the exact location of the summer house on satellite imagery and sent these details to Europol, but has chosen to refrain from publishing the exact location, due to the sensitivity of this criminal investigation.\nLocation Two: Olgino, Primorsky District, St. Petersburg Region, Russia\nWe first added more contrast to the photograph, which presumably was taken in St. Petersburg, so the skyline would be more visible. This enhanced version shows four pylons on the left side of the photograph, by which\nLorenzo Romani\nrecognized the skyline of St. Petersburg. Although we found many photographs with matching skylines, we noticed something was not completely right.\nThe most similar view we first found on\nYandex streetview\nshows a different bridge on the left side of buildings similar to the buildings in Europol’s photograph, but that bridge has only two pylons, while in Europol’s photograph there are or there\nseem\nto be four pylons visible. The bridge with four pylons, as visible in Lorenzo’s tweet, is to the right side of these buildings. Also, the distance from the four pylons to the matching buildings in Europol’s photograph is bigger than in Yandex street view. Some of the buildings on Yandex street view are not visible on Europol’s photograph.\nThen, we realized two things: Europol’s photograph was likely taken from a different direction; also, the photograph was a few years old and the skyline has changed during these years. Searching through older photographs and satellite imagery to check if the bridge might have been renovated, we found out that the bridge was built from 2014 to 2016 and the two pylons were constructed in 2015.\nEnhanced version with more contrast of Europol’s photograph.\nComparing the skyline in Europol’s photograph to Yandex streetview shows similarities, but also a few differences. The four pylons in Europol’s photograph are further away from the buildings than the bridge and some buildings seem to be missing in Europol’s photograph.\nOlder photographs of the area show the same skyline without the bridge and several other photographs. Also, Yandex street view shows a soccer stadium to the left side of the bridge with several light poles.\nHowever,\nsatellite imagery\nshows the stadium has eight light poles instead of four and the light poles stand skew instead of straight as in Europol’s photograph. Older satellite imagery makes clear that construction of the current stadium, named Gazprom Arena or\nKrestovsky Stadium\n, started in 2009 at the location of the former\nKirov Stadium\n, which was demolished in September 2006. The Kirov Stadium had four straight light poles, matching the four poles visible in Europol’s photograph.\nTop: Skyline from 2010 without bridge and construction work to the left.\nMiddle: Skyline from 2007 that closely matches Europol’s photograph skyline.\nBottom: Skyline from 2018 with bridge and soccer stadium.\nPhotographs taken from the beach area around the\nThunder Stone\n(“\nГром Камень\n” in Russian), a large megalith used in the 18\nth\ncentury as pedestal for the Bronze Horseman in St.Petersburg, seem to have the best matching skyline. In a\nphoto album\nconnected to a\n2004 article\nabout the Thunder Stone, a photograph that was uploaded in\nOctober 2004\nshows almost the same skyline as visible in Europol’s photograph, including the four light poles of the old soccer stadium. An overlay of Europol’s photograph over the 2004 photograph shows the striking similarity of the skyline.\nTop: 2004 photograph of the “Thunder Stone” beach area showing a very similar skyline as Europol’s photograph.\nBottom: Overlay of Europol’s photograph over the 2004 Thunder Stone beach area photograph.\nIn Europol’s photograph it seems a few\ntaller buildings\nare visible that are not visible in the 2004 Thunder Stone beach area photograph. This can be explained because construction of a residential complex named\nMorskoy Kaskad\n(literally: “Sea Cascade” in Russian) started in 2003 and was finished in 2005. Therefore, Europol’s photograph very likely was taken between spring and autumn 2005, or in the spring or summer of 2006.\nLocation Three: Jiangmen, Xinhui district, Guangdong province, China\nOne of the locations\npublished by Europol\non\n1 February\n, which was not tweeted about separately, is accompanied by the question, “Do you recognize this building in Asia?”\nIn response to this tweet, as well as to\nBellingcat’s tweet\n, several people responded by saying that they recognize the text displayed on several billboards on the building as Chinese and that the location was very likely to be in South China, because of the palm trees visible in the photograph. Only the text on two billboards is clear enough for Chinese speakers to read because of the photograph’s low resolution. The billboards mention used cars and a brand of motorcycles from China. As previously mentioned, further reactions to the tweets specifically mention the island Hainan, Zhuhai, Shenzhen, Zhongshan and Jiangmen as possible locations in South China.\nBuilding in South China with two readable billboards, displaying the name of a brand of motorcycles from China (bottom left) and “used car information” (bottom right). The image obviously has been edited/censored by Europol, because very likely a child was visible in the original image.\nTwitter user\nBen Lokshin\n, among others, deciphered the motorcycle brand billboard as “中裕摩托“, written in simplified Chinese characters and translated as Zhongyu Motorcycles. Another Twitter user named\nKonrad Iturbe\npointed out this brand is sold all over China, but also mentioned that only Jiangmen would have this kind of architecture, road layout, and palm trees.\nA Twitter user named\nJames\nnoticed a\nphotograph\nof a different building with the same first four Chinese characters displayed on a billboard – a logo for a motorcycle repair service department. That\nrepair service deparment\n, however, is on\nTaoyuan west road 35 in Mabazhen\nin the Qujiang district, Shaoguan, Guangdong province, more than 100 kilometers north of the other aforementioned cities in Southern China. Both\nGoogle\nas\nBaidu\n, showing yet another building with the same four Chinese characters on a billboard, give\nHiuanshi 3\nrd\nroad\n, Penjiang district in Jiangmen city, Guangdong province in Southern China as the first search result first “中裕摩托” (Zhongyu motorcycles).\nNathan Ruser\nfrom the\nAustralian Strategic Policy Institute\n, among others, recognized the text on the second billboard as simplified Chinese characters “二手车信息”, translated as “used car information.”\nZhongyu motorcycle repair service department in Mabazhen, Qujiang district, Shaoguan, Guangdong province, China (left),\nZhongyu motorcycle accessories in Penjiang district, Jiangmen, China (right).\nFurther research into the brand name “Zhongyu motorcycles” made clear the company that produced these motorcycles went\nbankrupt\nin\n2006\nand the owner\nfled\nChina, despite being ranked\nfirst in 2005\nas motorcycle production in China and that everyone in Jiangmen knew the factory closed down.\nThe\nlocation of the factory\n, which is only correctly displayed on the Chinese version of Google maps, was on Jinlu Road in the Jianghai District of Jiangmen City. The\nZhongyu motorcycle group\nwas established in 1994 and produced 600,000 motorcycles a year on the 420-acre industrial park in Jiangmen. A\n2002 archived version\nof Zhongyu motorcycles’ website mentions the engines were produced in the\nWuhe indsutrial zone\nin Daze town in the Xinhui district of Jiangmen.\nBecause Zhongyu went bankrupt in 2006, we suspected the Europol photograph to be at least ten years old, as after 2006 very likely no new advertisements of this motorcycle brand would appear anymore on billboards. Since these motorcycles were produced in Jiangmen, we knew there was a chance that Europol’s photograph was taken in Jiangmen, despite that these motorcycles likely were sold everywhere in China – or at least everywhere in South China. Our research started in various cities in South China and we especially focused to the layout of the road visible in the photograph: a two-way road with three lanes each, separated by a centre median strip with palm trees and lamp posts with four lights.\nAfter an intensive research, we finally found\nsuch a road and a similar building\nin the Xinhui district of Jiangmen, named\nSanhe Avenue\n. However, the building from Europol’s photograph was not found along this road. As satellite imagery in Jiangmen doesn’t go back further in history than 2008 and Europol’s photograph was likely taken in or before 2006, we realized there was a chance the building was demolished before 2008. As we expected to find the building along this north-south road, on the west side of the road based on the direction of the sunlight, Timmi Allen of our Bellingcat team created a\n3D model of the building\nto understand how the building would look like on satellite imagery.\n3D reconstruction model of Europol’s Asian city photograph, showing recognizable features of the building.\nVia this reconstruction model we eventually found the\nbuilding\non\nGanzhou Avenue East 88\nin the Xinhui district of Jiangmen, China. According to geotags on Baidu and Google, the building houses, among other businesses, a used car information center and a\nbusiness hotel\n, showing the same entrance on the hotel website as Europol’s photograph.\nBaidu’s street view from 2015 even shows the exact same lam post and blue one-way traffic sign on the centre median strip with palm trees. Examination of satellite imagery clearly shows Europol’s photograph was taken before 2014, as a building to the right of Europol’s building has been demolished before 2014 and a new building to the left was built before 2014. Also the height of the trees visible in Europol’s photograph fit the 2008 satellite imagery.\nThe building from Europol’s Asian city photograph on Google Earth satellite imagery of 2018 (top left) and satellite imagery from 2008 (top right). Baidu’s street view from 2015 of the same building and road visible in Europol’s photograph.\nThe photograph that Europol published was taken from behind a window of the building on the opposite side of the street. Europol’s photograph shows a black fence, big windows and two white pillars on the other side of the window. Also a white roof is visible on the bottom side of the photograph. The photograph clearly shows this part of the building has a round shape. A white cross and a car on the left side are visible outside in the area which seems a parking place. Baidu’s\nstreetview\nshows a part of the building with a round shape and a roof above the entrance. On the second floor of this part of the building, a fence is visible behind the only window that is not covered by a curtain.\nThe street view to the third floor of the other side of this building shows\na black fence\nwith exactly the same shape as in Europol’s photograph. However, since a large part of the roof above the entrance is visible, it’s more likely the photograph has been taken from the second floor, from one of the rooms that, on street view, has its windows covered by curtains. Photographs taken\nin meeting rooms\nof the building, which in the meantime have been deleted, also show the exact same type of fence.\nThe building where Europol’s photograph was taken. A similar fence is visible on the second floor (left)\nand the exact type of fence on the third floor (right). Large windows, white pillars and a white roof above the entrance are similar to Europol’s photograph.\nPhotographs taken from meeting rooms in the building where Europol’s photograph was taken\nshow the exact same type of black fence as in Europol’s photograph.\nFurther investigation into the building where Europol’s photograph was taken, led to the surprising conclusion that another photograph, published by Europol on\n1 February 2019\n, was taken inside that same building. This photograph, showing a hall of what looks like a shopping center and several red tables and benches with “Coca Cola” written on them, is accompanied by the question “Do you know where this food court is located?”\nAn article about the\nJianghui Times City\n(archived\nhere\n), published in July 2013, shows a photograph of a central hall of this shopping mall, which shows many similarities to Europol’s food court photograph. Another article about\nthe same shopping center\npublished in 2014 (archived\nhere\n), confirms the location.\nAnother photograph published by Europol on February 1, 2019 shows a hall inside what looks like a shopping mall, with several red tables and benches visible in the photograph with the Coca Cola logo displayed on them (left). Photograph of the hall of the Jianghui Times City shopping mall, the building from where Europol’s Asian city photograph had been taken from (right).\nComparison between another photograph of the main hall of the Jianghui Times City shopping mall\nto Europol’s food court photograph of 1 February 2019.\nConclusion And Recommendations To Europol\nThis analysis demonstrates that Europol’s “Stop Child Abuse” online crowdsourcing campaign is quite succesful. It’s not only that objects in photographs get\nrecognized\noften\n, so that they can be linked to an origin country making it possible for Europol to inform law enforcement of said country in order to further investigate the case. Sometimes, often after intensive research, locations of photographs eventually get geolocated precisely, possibly increasing the likelihood that such a case can be solved.\nHowever, we have a few recommendations to Europol in their crowdsourcing campaign. The first recommendation is to be careful with sharing photographs that can be easily found with reverse image search, as this could immediately lead to a victim or even a potential suspect. The risk for vigilante justice can be high, particularly if a social media profile belonging to a potential suspect or person of interest is uncovered. When it came to Location One, it was relatively easy to find one of the photographs with reverse image search, which happened just less than half an hour after one of the photographs was tweeted by Europol.\nThe second recommendation is this: If several photographs from the same case are displayed on Europol’s website and Europol is aware of these photographs as being linked, it can help to make the connection between the photographs clear. Both in the case of Location One and Location Three it became clear Europol uploaded two different photographs that are part of the same case, without specifying it. It’s possible that this information could help solve the case possibly even quicker.\nOur final recommendation: Before a photograph is shared, it would be good to be realistic about the chances for recognition. Sometimes the quality of the photographs of objects seems too low to be able to be recognized and sometimes the location seems as if it could really be anywhere and could not be geolocated without any background information. This is why we recommend sharing as much information about a certain location as possible, in case that information is available. Our next article in the series of Europol’s “Stop Child Abuse” will show an example of a photograph that like would never have been geolocated if no additional information was found about the image.\nResearch by Daniel Romein, Timmi Allen, Klement Anders, Noor Nahas and “Bo”,\nreview/editing by Natalia Antonova.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/summer-house-collection-1200x525.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/summer-house-album-1200x449.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/Lake-1200x602.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/video-location-1200x334.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/enhanced-1200x263.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/comparison.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/skylines-1200x424.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/thunder_stone_2004_skyline_and_Europol_overlay-1200x768.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/Building_and_text.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/Zhongyu_compare-1200x400.jpeg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/3D-model-compare-1200x436.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/Jiangmen_building_compare.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/Times_hotel-1200x385.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/meeting_rooms-1200x424.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/mallchina_coke-1200x437.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/07/2018-6-c17-compare-1200x451.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2019-09-11_europols-child-abuse-image-geolocated-in-ukraine-a-forgotten-story-hidden-behind-a-landscape.json b/bellingcat_geolocation_json/2019-09-11_europols-child-abuse-image-geolocated-in-ukraine-a-forgotten-story-hidden-behind-a-landscape.json
new file mode 100644
index 0000000000000000000000000000000000000000..83029bee3eb2ab1466a1267000bbfcd715d7ddac
--- /dev/null
+++ b/bellingcat_geolocation_json/2019-09-11_europols-child-abuse-image-geolocated-in-ukraine-a-forgotten-story-hidden-behind-a-landscape.json
@@ -0,0 +1,112 @@
+{
+ "slug": "europols-child-abuse-image-geolocated-in-ukraine-a-forgotten-story-hidden-behind-a-landscape",
+ "title": "Europol’s Child Abuse Image Geolocated In Ukraine: A Forgotten Story Hidden Behind A Landscape",
+ "url": "https://www.bellingcat.com/news/2019/09/11/europols-child-abuse-image-geolocated-in-ukraine-a-forgotten-story-hidden-behind-a-landscape/",
+ "date": "2019-09-11T08:27:15+00:00",
+ "modified": "2019-09-17T11:55:22+00:00",
+ "author": "Carlos Gonzales",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Research by Carlos Gonzales, Daniel Romein, Timmi Allen and “Bo” The following report contains reference to a child modelling studio producing child sexual abuse material in 2001. All names related to the studio are fictitious. The original source did not contain any explicit material. All the images accessed and used during the investigation were already […]",
+ "content_html": "
\n
Research by Carlos Gonzales, Daniel Romein, Timmi Allen and “Bo” \n
\n
The following report contains reference to a child modelling studio producing child sexual abuse material in 2001. All names related to the studio are fictitious. The original source did not contain any explicit material. All the images accessed and used during the investigation were already censored, but for the avoidance of doubt, it must be noted that the researchers did not obtain, look or download any explicit content. The original source was shared with Europol before the publication of this report and cannot be revealed for protection of the victims and as to not impede the investigations. Although the main objective of the article is to show the method by which an image listed by Europol was geolocated, Bellingcat have decided to publish some details found in the investigation to create awareness of the subject and to support Europol’s #StopChildAbuse campaign.
\n
Europol currently holds more than 40 million images of child sexual abuse from across the world. In June 2017, Europol launched a crowdsourcing campaign called Stop Child Abuse – Trace an Object. Censored extracts from explicit images are regularly published on their website and members of the public are asked to help tracing their location or country of origin. These tips are then used to inform the competent law enforcement authority to further investigate the lead and to assist in the identification of the offender and the victim.
\n
As of 23 March 2019, Europol has received more than 23,000 tips, which led to the identification of eight victims and the prosecution of one offender. In recent reports, Bellingcat wrote about their contribution to geolocating images in China and Russia.
\n
For several months, the image shown in Figure 1 was live on Europol’s website. From now on we will refer to this image as “ImageC5” in the text. Bellingcat team members and other Twitter users previously pointed out the difficulty of geolocating this image. The main reason: There are no mountains, landmarks, roads, posters, signs, brand names, or any remarkable object featured on it. Instead, there is a rather desolate landscape flooded with weeds and few rural buildings in very low resolution. What we did not know then was the dark story ImageC5 hid, the number of victims affected by it and how the story itself would allow us to geolocate this image to Kalahliya, a Ukrainian village 47 km southwest of Odessa.
\n
Figure 1: ImageC5 as listed by Europol on their website #StopChildAbuse #TraceAnObject (left). The image was geolocated to a village called Kalahliya, which is southwest of Odessa in Ukraine. The red mark indicates the approximate location from which the original photo was taken from; just on the northeast corner of a group of building ruins (right).
\n
\nThe Initial Research
\n
Our first observations about the image were as follows:
\n
\n
A small gradient in the terrain with a few sudden drops across the land
\n
Rural houses and buildings with metal roofs (single and multiple pitch)
\n
Tall dried weeds and broken branches as well as reeds and grasslands
\n
An orange/brick construction with a dark roof far on the right side of the image
\n
The image was edited by Europol to protect the victim. The censoring process was done in such a way that almost 10% of the landscape was reconstructed. Small fractions from other points on the image were merged into the cropped area. This caused the repetition of items in several points of the landscape.
\n
\n
\nImportant Definitions And Key Facts
\n
Child Sexual Abuse Material (CSAM), also called child pornography, refers to any representation, by whatever means, of a child engaged in real or simulated explicit sexual activities or any representation of the sexual parts of a child for primarily sexual purposes. In 2014, Russia was reported to host the second highest amount of CSAM in the world, accounting for 24%. In 2016, Ukraine was the first in Eastern Europe. The traffic of children for sexual exploitation to Moscow and St. Petersburg from Moldova and Ukraine has also been documented.
\n
The type of houses observed on ImageC5 suggested a very small village which seems to be a rural Soviet construction style, as opposed to western European.
\n
ImageC5 did not feature either any type of tropical, coniferous, or desert vegetation. Therefore Asia, the Middle East and dense forested areas in Europe and North America were not considered. Grasslands are located across Moldova, Ukraine and a thin belt along the border between Russia and Kazakhstan (See Figure 2). Intersecting all these geographical and architectural observations with the information about Children Sexual Abuse in the region, the initial search was chosen to be within the area marked in red on the map in Figure 3, particularly those villages with links to major cities such as Moscow and Kyiv.
\n
Figure 2: Vegetation map of the former Soviet Republics. Grasslands are found across Ukraine and a thin Belt into Russia.
\n
Figure 3: Initial region selected to start the search for ImageC5. The area included West Moldova, South and East Ukraine, and the area around the Ukraine-Russia Border, Moscow, and the River Volga.
\n
It was initially believed the brick construction on the far right side of the picture could have been a Soviet-era railway water tower or a small train station. Railway routes, travellers’ videos, and hundreds of photos were checked to understand the landscape, as well as the typical layout and architecture of small villages.
\n
The first slight landscape similarities with ImageC5 were found in eastern Ukraine within the Sumy oblast. Unfortunately, with no defined landmarks to look for and the limited availability of imagery and street views outside main cities, further visual checks were unproductive. No evidence found was significantly relevant as to draw any conclusion. The place could have been anywhere in the marked region if it existed there at all.
\n
\nA Breakthrough In The Search
\n
The next step was to find more details about Child Sexual Abuse cases in the suspected region that could then be associated with ImageC5. In early June, after following a lead from a regional children exploitation report, a comprehensive investigative file was found containing details of a child modelling studio trafficking children from Moldova to Ukraine for the production of CSAM.
\n
Important Note:
\n
The source of such files cannot be revealed for protection of the victims and as to not impede the investigations. The original source was shared with Europol before the publication of this report. All the images accessed and used by the Bellingcat Team during the investigation were already heavily censored by the original source of the files, but for the avoidance of doubt, we will once again note that the researchers did not obtain, look or download any explicit content. All names related to the studio used in this report are fictitious.
\n
What information did we find in the files and how did we use it to geolocate ImageC5?
\n
The investigative file described the operation of the studio and their associated websites. Most of the CSAM created by the studio was reported as having been produced in the Odessa region in Ukraine. The file also included a list of unknown locations where at least 60 models were photographed in 2001 for Website2.
\n
All unknown locations appeared classified as Coast, Inland or Ruins, followed by a sequential number, their respective group of censored thumbnails, and a brief description of the surrounding landscape or visible buildings. No specific geolocation was given.
\n
To our surprise, in the group called “Location Ruins (1 & 8)”, a thumbnail containing the same ImageC5, as published by Europol, was found. A direct link between ImageC5 with an official Child Sexual Abuse case was established.
\n
Other censored thumbnails found next to ImageC5 showed several building ruins and a red building in the background, which, according to the file description, could have been a church in a nearby village. The search for a desolate landscape had evolved into a search for more identifiable landmarks – the building ruin and “the red church.”
\n
Having new information about other potential unknown locations related to the case, our strategy was to “encircle” ImageC5’s location in the Odessa region by a simultaneous hunt for the red church, Location Ruins (1 & 8), Ruins (2,3 & 7), Inland (3,6 & 7), Inland (8) and Coast (1 & 6). This selection of censored thumbnails included the most recognisable items for each one of the locations and constituted the key to finding ImageC5. A diagram with all the material used is presented in Figure 4.
\n
Figure 4: Selection of unknown locations listed in the investigative files and related to ImageC5. All images accessed and used by Bellingcat during the investigation were originally censored by the source of the files.
\n
\n“The Red Church” And Landscape Features
\n
Initial efforts were made to find the red church by exploring the area using Google Earth and Yandex. The amount of churches in Odessa oblast (approx. 33.000 km2) was too large to scrutinise them one by one and many did not have photos easily available for checks.
\n
We also noticed that in some cases the colours and features of churches were changed, due to upgrades and repairs. This made the initial search for the church extremely difficult. Although we were initially unsuccessful, valuable landscape and architecture information was gathered. Different points across the region were analysed with similarities found in several areas; for example, Anan’iv Pershyi and Vyzyrka. Based on this and other examples, it was likely the location to be found along the coastal area of Odessa.
\n
Using Google imagery, a dimensional analysis was done on several churches to establish a reference geometric pattern (see Figure 5):
\n
\n
The height of the bell tower was the same as the altar tower, excluding the domes.
\n
The width of the altar tower was approximately double of the bell tower.
\n
Exterior made of red bricks. But we noted that it could have been repainted.
\n
Both towers could feature either round, semi-round, pitched or onion type domes.
\n
In church architecture, altar towers are normally oriented to the East. In almost all cases in Odessa there was no exception. This meant Ruins (1 & 8) were very likely to be located southeast of the church. A very important check to do when finding potential geolocations.
\n
\n
Figure 5: Dimensional characteristics of similar church in Odessa oblast.
\n
\nThe Ruins
\n
Group Ruins (1 & 8) featured what it seemed to be the ruins of a small industrial facility. Scattered precast pillars could be seen on the ground. There were walls with faded pink tones, large holes, and bullet marks on many of them. The holes seemed similar to those left by a projectile impact. Presumably, these ruins were the result of previous conflicts in the area.
\n
But where to find this type of ruin in an area of 40,000 km2 including Odessa oblast and west Moldova? Furthermore, do the ruins still exist or have the sites been cleared up to give way for new developments? Also, which conflict could have inflicted such damage on a rural structure before 2001?
\n
Using Google imagery, examples of ruins were gathered to establish a visual reference as shown in Figure 6.
\n
Figure 6: Example of ruins found in the Odessa region. Establishing a reference to locate Ruins (1& 8).
\n
Two main events were considered: WWII and the Transnistria conflict in 1992. WWII front lines were studied according to literature as per the maps in Figure 7. In 1941, a system of 3 defence rings was established by the Soviets to protect Odessa; with the outermost ring located at a 50km mark from the city. From March to April 1944, the Red Army launched offensives to liberate Odessa. In August 1944, as part of the Jassy–Kishinev operation to reclaim the Moldavian SSR, the Red Army engaged a large German-Romanian contingent with heavy battles starting with the Akkerman landing.
\n
In both conflicts, WWII and Transnistria, there were fronts localised on several points along the Dniester River and Dniester Estuary. This line would become a focus of attention later in our search.
\n
Figure 7: Battle fronts in 1941 and 1944 during WWII were used as a reference to find potential matching ruin sites. The Dniester Estuary would become important in our search.
\n
\n
Geolocating ImageC5 By Finding Other Locations Related To The Case
\n
1. Location Ruins (2,3 & 7) – The Greenhouse
\n
Ruins (2,3 &7) appeared to be part of a period property or castle featuring a glass roof structure typical of those found in old greenhouses. The studio was reported to operate from several apartments in the city of Odessa. Their approximate location was investigated and pinned as shown on the map in Figure 8. It was believed that one or several of the locations could have been nearby their studios. Hence, we started to explore the area with Google Maps along Road F. From north to south, we looked for abandoned buildings, castles and ruins where any of the studio’s sets could have been located.
\n
A landmark symbol led us to the ruins of greenhouse Marazli, originally built in 1889. Pictures found in Yandex dated back to 2016 were gathered and analysed. Many features in its interior matched those found in the censored thumbnails of ruins (2, 3 & 7), especially the angle of the glass roof structure, supporting beams, door features, windows and stone masonry (See Figure 9 ). The location served its purpose until 1990s, but various events ultimately contributed to its abandoned state. The source reported the location was used by the studio to produce several CSAM between July and August 2001. No signs of location Ruins (1 & 8) or the red church were noticed in the surrounding areas.
\n
Figure 8: Approximate location of the apartments owned by the studio in the city of Odessa. The F location allowed our investigation to get closer to a child abuse site: the greenhouse.
\n
Figure 9: A 1889 greenhouse in Odessa City. Pictures found in Yandex and dated 2016 were gathered and analysed. Many features in its interior matched those found in the censored thumbnails of ruins (2, 3 & 7). The source reported the location was used by the studio to produce several CSAM between July and August 2001.
A thumbnail within the Inland (3 & 6) group featured a river or lake. A transmission tower in the background could be seen just opposite the bank from where the shot was taken. Another shot, believed to have been taken in the opposite direction to the first, featured another transmission tower, dried grasslands, and small trees in the background.
\n
Using Google Earth, narrow bodies of water with these characteristics along the coast were investigated. On the south end of Oleksandrivske Reservoir, a peninsula with transmission lines on its west side was identified. Across the reservoir, another transmission tower and cables crossing above the water were noticed. By setting the time bar to 05/06/2008, the grass on the peninsula showed up as dried and small trees could be seen clearer now on the east side of the peninsula, just behind the transmission tower. A match was found (see Figure 10). The source reported this location was used by the studio to produce several CSAM between July and August 2001.
\n
Checking the rest of the reservoir for more clues, a white building with a slim tree in the front was noticed east of the previous location. These elements matched those displayed on location Inland (7). After analysing the image, it was determined the shot was taken from the south shore of the Oleksandrivske Reservoir (see Figure 10). The source reported this location was used by the studio to produce CSAM in October 2001. No signs of Ruins (1 & 8) or the red church were identified in the surrounding areas.
\n
Figure 10: Oleksandrivske Reservoir. Power transmission lines and landscape matched those featuring in thumbnail images for location Inland (3,6,7). The source reported these locations were used by the studio to produce several CSAM between July and October 2001.
\n
\n3. Location Inland (8) – Kuyalnik Estuary
\n
The unknown location Inland (8), featured a water body and sand around the shores. Despite the low quality of the censored thumbnails, a pink/red tone to the sand was noticed. On the left side of the image, a long cliff which ended in the horizon line was visible. Looking at the background, no buildings or mountains were seen. The body of water seemed to be quite elongated.
\n
In our initial searches around Odessa, we had noticed Kuyalnik Reservoir, which had pink sands. We re-visited the images of Kuyalnik for a reinspection. The result: a Google maps image featuring the same cliff on the west shore of the reservoir.
\n
The cliff exhibited the same angular lines converging downwards and touching the horizon. Dark pink sand was also noticed as well as curved water marks. Another location verified. The source reported the location was used by the studio to produce CSAM in September 2001. No signs of Ruins (1 & 8) or the red church were identified in the surrounding areas.
\n
Figure 11: Kuyalnik Estuary. The source reported the location was used by the studio to produce CSAM in September 2001.
Many of the censored images within group Coast (1 & 6) featured a peninsula with a slim tree on it. The images seemed to have been taken on a shore very close to a grassy cliff.
\n
Rocks bigger than 1.5 meters in diameter, which had become detached from the cliff, appeared in almost every shot from different angles. All main elements were numbered (1 to 10) and organised by scene (A to D). The tree and the peninsula were selected as landmarks with more potential to be identified in satellite images.
\n
Having checked other bodies of water on the northeast side of the coastal region of Odessa, the southwest area around the Dniester Estuary was to be explored using Google Earth. On the east coast of the estuary at the level of the village Mykolaivka, a small peninsula was found. This peninsula had several trees on it but only one produced a long shadow relative to the size of the peninsula and matching the relative position to the one shown in the censored thumbnails.
\n
From the 2014 satellite image, large rocks could be seen all along the shore. After careful analysis, all rocks and trees were identified. Another image geolocated. The source reported the location was used by the studio to produce several CSAM between the end of July and beginning of August 2001.
\n
Figure 12: Location Coast (1& 6). All main rock elements on thumbnail images identified along the east coast of the Dniester Estuary at the level of the village Mykolaivka. The source reported the location was used by the studio to produce several CSAM between the end of July and beginning of August 2001.
\n
\n5. The Final Search And Verification
\n
After geolocating four different sites related to the case, a 50 km radial perimeter was established around the city of Odessa. This perimeter coincided with one of the battle fronts that took place during WWII on the east shore of the Dniester Estuary. Hence, it was believed that some type of ruins, if linked to the war at all, were likely to be found around this area.
\n
Figure 13: Geolocation of the selected group of images marked in red.
\n
The main road connecting the village of Mykolaivka and Ovidiopol (T1625) was checked for any signs of the red church or ruins. When doing a visual inspection with Google Earth around the village of Kalahliya (Калаглія / Калаглия ), a church was noticed at coordinates 46.281507, 30.358500 on June 2. Its name: Церква Миколая Чудотворця, i.e. the Church of St Nicholas. All the features matched our references. A Twitter user, Lorenzo Romani, also confirmed finding the same church on June 22 through someone in his network.
\n
Figure 14: St. Nicholas Church, located in the village Kalahliya. The church matched all features recognised on the image related to ImageC5.
\n
But was this truly the right place? Were there any ruins around? Checks were done in the immediate surroundings southeast of the church but there were no signs of any recognisable ruin structure. Then, the historical view of Google Earth was changed. A compound in ruins was visible in other years, with 2007 being the earliest image available.
\n
Despite the lack of street views in Google Earth and Yandex for this location, Images and YouTube videos of the village were scrutinised in order to identify all the buildings visible on ImageC5. A drone video found offered the best view of the site and it helped us not only to identify each of the elements but also to estimate the position of the camera (See Figure 15 and Figure 16).
\n
Figure 15: ImageC5 buildings identified on a YouTube drone view of the village Kalahliya
\n
Figure 16: Northeast corner of the ruins as the approximate camera position from where the ImageC5 landscape photograph was taken.
\n
Further checks were done on two of the buildings featuring on ImageC5: the single pitch roof building (marked in yellow) and the brick construction on the far right (marked in red). We found videos recorded from Road T1625 which connects Kalahliya with Ovidiopol and passes just behind these buildings. The type of windows, roof, masonry colours as well as relative position matched those shown on satellite views as well as ImageC5 (see Figure 17).
\n
Figure 17: Videos and satellite images used to analyse main building construction features such as masonry colour, roof, and windows type for selected buildings. All elements matched those on ImageC5.
\n
Satellite images and all thumbnails in group Ruins (1 & 8) were carefully analysed to extract information such as the positions of trees, shadows sizes, sunlight direction, vegetation, exterior paint colours, and the relative position of debris. A solar calculation was done for early September 2001 as representative data for the month when the CSAM was generated.
\n
Putting all of the puzzle pieces together, a basic sketch of the ruins was done to understand more about the site and to indicate the relative points where the thumbnail photos were taken from (See Figure 18).
\n
The size of the construction and the presence of gates to the southeast and northwest ends of the main structure reconfirm the possibility that the building was an old, small-sized industrial/army facility. The reconstruction should be taken only as an intuitive guide only. With all elements identified, the site has been verified and ImageC5 has been now officially geolocated. The source reported the location was used by the studio to produce several CSAM between August and September 2001. A summary of the geolocation process followed is offered in Figure 19.
\n
Figure 18: Basic reconstruction of ruins based on solar data for September 2001, satellite images as well as the image thumbnails found in the investigative files. The source reported the location was used by the studio to produce several CSAM between August and September 2001.
\n
Figure 19: A summary of the strategy followed to geolocate ImageC5 after the finding of the investigative files. Geolocating other images within the set allowed the team for a relative quick converge into the final findings.
\n
\n
“Studio A” And Their Associated Websites
\n
Although our main goal was to share information on how ImageC5 was geolocated, we also want to offer a brief summary of the operation of the studio and their websites as reported by the source. Though almost 20 years have passed and there is little reference to it on the internet, this studio impacted the world of CSAM production and contributed to its development. It is a clear example of how purveyors of CSAM adopt the guise of benefactors. First – they pay off the victims’ typically impoverished families. Second – they develop a child educational and entertainment programme designed to quell the parents’ and the public’s distrust in the studio. Finally, they take advantage of then-underdeveloped CSAM legislation in Ukraine and an emerging worldwide internet. This is a business model that is still being followed by many modern criminal organizations.
\n
This following section was extracted from the files and by no means expresses the views of the authors – it is rather meant to raise awareness of how such criminal enterprises operate. For legal reasons and for protection of the victims and offenders, we cannot disclose the source. All names used here are fictitious.
\n
Studio A was a company founded by Denis M. & Sergey P. in early 2000s with funds coming from several European financiers. Denis was an IT expert and Sergey was a photographer known for nude “artistic” work with teens. Both individuals met in the summer 2000 and made several arrangements with other webmasters and studios to set up Website1. Sergey would generate the photographic content and be the main model finder; whereas Denis would be in charge of the technical side. By September of the same year, Website1 was fully operational, with material being produced by Sergey in Ukraine. All websites associated operated on a subscription basis. The studio operation included transportation, admin, and security crews. The latter were mainly previous law enforcement officers.
\n
Most of the underage children featuring in their productions were scouted around schools in eastern Europe – mainly Moldova. They usually came from impoverished backgrounds and were offered “modelling work” to help their families with money equivalent to a month’s salary of their parents. Knowing the conditions required by the studio, many parents reportedly agreed to send their children from Moldova to Ukraine with Studio A. The victims lived in different apartments owned by the studio across the city and received computer, arts and English lessons when they were not “modelling.” In efforts to further legitimize the criminal enterprise, the studio organized outings with the parents. Also, according to interviews reported in the files, Denis had falsely claimed on several occasions that the business was “socially responsible” and part of a charitable project in partnership with a worldwide human rights organization.
\n
According to the files, Sergey, who managed the business independently, had a serious addiction to heavy drugs. According to victims’ testimonies, as per the files, Sergey not only photographed victims naked but would then physically sexually assault them and offer them addictive substances. What ultimately happened to Denis and Sergey’s partnership remains unclear, but it is reported that tensions between Denis and Sergey increased and they parted ways in early 2001. Denis renamed the company Studio A* and Sergey kept the original enterprise. Studio A. Sergey convinced a number of parents and children to stay him, while another of other children exploited by Website1 moved on to Studio A*.
\n
Denis then created another website, Website2. The initial material used for this website came from Russia-based projects that featured nude underage girls. Later, Denis started to create new material for which he recruited new victims in Moldova. Sergey kept most of the material created for Website1 and sold it to other parties on the web. Sergey kept working for other studios and also created a new brand, Studio C. Both Website1 and Website2 went down starting in 2002. According to Denis, Website2 was hijacked and redirected to another page that apparently contained more explicit abuse material. According to the files, Denis placed responsibility for that on organizations based in Russia.
\n
After the 9/11 attacks in the U.S., federal authorities and international agencies started to investigate all sort of irregularities on the web more closely. It is likely that Ukrainian authorities at this point came under more pressure, while the police had information about the frequent trafficking of minors from Moldova to Ukraine. In 2002, two women travelling with four children were detained in a train station close to the Ukraine/Moldova border for interrogation. They were carrying false documentation to take these minors abroad. During questioning, all underage girls mentioned they were paid an average of $50 per session in exchange for taking part in explicit photography produced by two individuals: Denis & Sergey. The police proceeded to raid all studios and seized all computer equipment, drugs, and CSAM. Both Denis and Sergey had fled by then, presumably alerted by their corrupt contacts on the police force.
\n
Denis moved to Switzerland and then to Moldova, whereas Sergey was reported to have lived in Moldova, Germany and the U.S. The Ukrainian government initiated a case against both individuals in 2002 for production, sale, and distribution of CSAM. Surprisingly, charges against Denis were dropped in 2003. Ukrainian legislation at that time did not have any provisions for CSAM production.
\n
According to the source, Denis relaunched Website1, but without any nude material. He also created a new studio called Studio B and closed Studio A*. By 2004, both websites Website1 and Website2 were down. In 2007, Denis married one of the girls that worked for the studios – it is unclear how old she was at the time and when she was first recruited – and the ceremony reportedly took place in Moldova. A documentary film was also produced to portray Denis and girls who worked for the studio in more recent years. This production was never commercialised.
\n
Most of the CSAM Sergey produced remains undocumented due to missing data. Sergey was put on a wanted list by police in Germany in relation to abuse of minors as well as for CSAM dissemination. There is no information available as to determine whether Sergey was prosecuted or not.
\n
Some key facts after analysing the information presented in the investigative files:
\n
\n
ImageC5 was produced for Website2
\n
At least 60 victims between the ages of 8 and 17 were hired to work for Website2
\n
More than 7300 images were produced in less than a year
\n
The website published approximately 270 sets of images
\n
60% of the CSAM was produced in studio facilities, while 40% was produced in outdoor locations
\n
On average, 3% of the production was done in each one of the Ukrainian outdoor locations with an 8% maximum produced in Russia.
\n
There were seven “photographers” working for the page including Denis and other unidentified individuals associated with the studio.
\n
The source reported the studio employed assistants, drivers, teachers and security staff. The latter have had previously worked in law enforcement.
\n
$250,000 were received on the personal accounts of Denis and Sergey for the operation carried in Ukraine.
\n
\n
\nConclusion & Recommendations
\n
ImageC5 on its own contained no relevant information. The initial geolocation process for this image proved to be unproductive via employing traditional methods. More details were needed to be able to take the research further.
\n
This time, an in-depth literature review on child sexual abuse cases supplied the information required to narrow down the search. Without knowing the details of the case, ImageC5 would have been never geolocated. By finding other locations related to the case, valuable landscape, architectural, war zone, and terrain data was collected. The iteration process of the search converged quicker towards the location where ImageC5 photograph was taken from.
\n
After an intensive research, Images are normally geolocated accurately. A great deal of information related to the cases is usually gathered through the process. Not all details found in our investigations are shared – due to the need to protect victims. On this case, the investigative files found came from an anonymous source which Europol has been informed about.
\n
Eighteen years have passed since the discovery of the dark operation of Studio A. The case was known to both Ukrainian and international authorities. If data and privacy laws permit, It could be helpful to see that for any child abuse case already processed by law enforcement bodies in Europe, all CSAM images seized should be entered into a centralised database. Law enforcement agencies could then run algorithms for an internal image reverse search to assist future geolocations and investigations. In this way, images for which there is already a background case would be left out of any crowdsourcing campaign. Moreover, if the amount of data collected is large, maybe the algorithm could be capable of comparing landscapes and other information between old and new images, thus drawing a preliminary geographical search area or creating leads for new cases.
\n
The Studio A case is a clear example of child sexual abuse and human trafficking. The operation persuaded stakeholders of their “socially responsible” approach to leverage their business. The production of CSAM was allowed by an archaic legal framework and corruption, but also by the families of the victims, who exploited their children for financial gain. This business model has been copied by many crime organizations disguised as “modelling studios.” After many years, the situation in Ukraine has changed – but it has changed for the worse. The problem has been aggravated by the development of technology and faster communications. There are new forms of criminality and an intensified production and dissemination of child abuse material in the Ukraine-Moldova-Belarus-Russia region.
\n
In sharing our investigations and search methods we believe we can contribute with authorities to bring offenders to justice, and that we can make the #StopChildAbuse network stronger while raising awareness about the problem. Our intention is to encourage members of the public and OSINT professionals to be vigilant and pass any known information to the competent authorities. Although some images published by Europol might seem old, new cases might be related to those pictures and any lead or input into their database could mean the difference between continuous abuse or rescuing the victims. At Bellingcat, we are strongly committed to these efforts.
\n
",
+ "content_text": "Research by Carlos Gonzales, Daniel Romein, Timmi Allen and “Bo”\nThe following report contains reference to a child modelling studio producing child sexual abuse material in 2001. All names related to the studio are fictitious. The original source did not contain any explicit material. All the images accessed and used during the investigation were already censored, but for the avoidance of doubt, it must be noted that the researchers did not obtain, look or download any explicit content. The original source was shared with Europol before the publication of this report and cannot be revealed for protection of the victims and as to not impede the investigations. Although the main objective of the article is to show the method by which an image listed by Europol was geolocated, Bellingcat have decided to publish some details found in the investigation to create awareness of the subject and to support Europol’s #StopChildAbuse campaign.\nEuropol currently holds more than\n40 million images\nof child sexual abuse from across the world. In June 2017, Europol launched a crowdsourcing campaign called\nStop Child Abuse – Trace an Object\n. Censored extracts from explicit images are regularly published on their website and members of the public are asked to help tracing their location or country of origin. These tips are then used to inform the competent law enforcement authority to further investigate the lead and to assist in the identification of the offender and the victim.\nAs of 23 March 2019, Europol has received more than 23,000 tips, which led to the identification of eight victims and the prosecution of one offender. In\nrecent reports\n, Bellingcat wrote about their contribution to geolocating images in China and Russia.\nFor several months, the image shown in Figure 1 was live on\nEuropol’s website.\nFrom now on we will refer to this image as “ImageC5” in the text. Bellingcat team members and other Twitter users previously pointed out the difficulty of geolocating this image. The main reason: There are no mountains, landmarks, roads, posters, signs, brand names, or any remarkable object featured on it. Instead, there is a rather desolate landscape flooded with weeds and few rural buildings in very low resolution. What we did not know then was the dark story ImageC5 hid, the number of victims affected by it and how the story itself would allow us to geolocate this image to\nKalahliya\n, a Ukrainian village 47 km southwest of Odessa.\nFigure 1: ImageC5 as listed by Europol on their website #StopChildAbuse #TraceAnObject (left). The image was geolocated to a village called Kalahliya, which is southwest of Odessa in Ukraine. The red mark indicates the approximate location from which the original photo was taken from; just on the northeast corner of a group of building ruins (right).\nThe Initial Research\nOur first observations about the image were as follows:\nA small gradient in the terrain with a few sudden drops across the land\nRural houses and buildings with metal roofs (single and multiple pitch)\nTall dried weeds and broken branches as well as reeds and grasslands\nAn orange/brick construction with a dark roof far on the right side of the image\nThe image was edited by Europol to protect the victim. The censoring process was done in such a way that almost 10% of the landscape was reconstructed. Small fractions from other points on the image were merged into the cropped area. This caused the repetition of items in several points of the landscape.\nImportant Definitions And Key Facts\nChild Sexual Abuse Material\n(\nCSAM\n), also called child pornography, refers to any representation, by whatever means, of a child engaged in real or simulated explicit sexual activities or any representation of the sexual parts of a child for primarily sexual purposes. In 2014, Russia was reported to host the second highest amount\nof\nCSAM\nin the world, accounting for\n24%\n. In 2016, Ukraine was the\nfirst in Eastern Europe\n. The traffic of children for sexual exploitation to Moscow and St. Petersburg from Moldova and Ukraine has also been\ndocumented\n.\nThe type of houses observed on ImageC5 suggested a very small village which seems to be a rural Soviet construction style, as opposed to western European.\nImageC5 did not feature either any type of tropical, coniferous, or desert vegetation. Therefore Asia, the Middle East and dense forested areas in Europe and North America were not considered.\nGrasslands\nare located across Moldova, Ukraine and a thin belt along the border between Russia and Kazakhstan (See Figure 2). Intersecting all these geographical and architectural observations with the information about Children Sexual Abuse in the region, the initial search was chosen to be within the area marked in red on the map in Figure 3, particularly those villages with links to major cities such as Moscow and Kyiv.\nFigure 2: Vegetation map of the former Soviet Republics. Grasslands are found across Ukraine and a thin Belt into Russia.\nFigure 3: Initial region selected to start the search for ImageC5. The area included West Moldova, South and East Ukraine, and the area around the Ukraine-Russia Border, Moscow, and the River Volga.\nIt was initially believed the brick construction on the far right side of the picture could have been a Soviet-era railway water tower or a small train station. Railway routes, travellers’ videos, and hundreds of photos were checked to understand the landscape, as well as the typical layout and architecture of small villages.\nThe first slight landscape similarities with ImageC5 were found in eastern Ukraine within the Sumy oblast. Unfortunately, with no defined landmarks to look for and the limited availability of imagery and street views outside main cities, further visual checks were unproductive. No evidence found was significantly relevant as to draw any conclusion. The place could have been anywhere in the marked region if it existed there at all.\nA Breakthrough In The Search\nThe next step was to find more details about\nChild Sexual Abuse\ncases in the suspected region that could then be associated with ImageC5. In early June, after following a lead from a\nregional children exploitation report\n, a comprehensive investigative file was found containing details of a child modelling studio trafficking children from Moldova to Ukraine for the production of\nCSAM\n.\nImportant Note:\nThe source of such files cannot be revealed for protection of the victims and as to not impede the investigations. The original source was shared with Europol before the publication of this report. All the images accessed and used by the Bellingcat Team during the investigation were already heavily censored by the original source of the files, but for the avoidance of doubt, we will once again note that the researchers did not obtain, look or download any explicit content. All names related to the studio used in this report are fictitious.\nWhat information did we find in the files and how did we use it to geolocate ImageC5?\nThe investigative file described the operation of the studio and their associated websites. Most of the\nCSAM\ncreated by the studio was reported as having been produced in the Odessa region in Ukraine. The file also included a list of unknown locations where at least 60 models were photographed in 2001 for Website2.\nAll unknown locations appeared classified as Coast, Inland or Ruins, followed by a sequential number, their respective group of censored thumbnails, and a brief description of the surrounding landscape or visible buildings. No specific geolocation was given.\nTo our surprise, in the group called “Location Ruins (1 & 8)”, a thumbnail containing the same ImageC5, as published by Europol, was found. A direct link between ImageC5 with an official\nChild Sexual Abuse\ncase was established.\nOther censored thumbnails found next to ImageC5 showed several building ruins and a red building in the background, which, according to the file description, could have been a church in a nearby village. The search for a desolate landscape had evolved into a search for more identifiable landmarks – the building ruin and “the red church.”\nHaving new information about other potential unknown locations related to the case, our strategy was to “encircle” ImageC5’s location in the Odessa region by a simultaneous hunt for the red church, Location Ruins (1 & 8), Ruins (2,3 & 7), Inland (3,6 & 7), Inland (8) and Coast (1 & 6). This selection of censored thumbnails included the most recognisable items for each one of the locations and constituted the key to finding ImageC5. A diagram with all the material used is presented in Figure 4.\nFigure 4: Selection of unknown locations listed in the investigative files and related to ImageC5. All images accessed and used by Bellingcat during the investigation were originally censored by the source of the files.\n“The Red Church” And Landscape Features\nInitial efforts were made to find the red church by exploring the area using Google Earth and Yandex. The amount of churches in Odessa oblast (approx. 33.000 km\n2\n) was too large to scrutinise them one by one and many did not have photos easily available for checks.\nWe also noticed that in some cases the colours and features of churches were changed, due to upgrades and repairs. This made the initial search for the church extremely difficult. Although we were initially unsuccessful, valuable landscape and architecture information was gathered. Different points across the region were analysed with similarities found in several areas; for example, Anan’iv Pershyi and Vyzyrka. Based on this and other examples, it was likely the location to be found along the coastal area of Odessa.\nUsing Google imagery, a dimensional analysis was done on several churches to establish a reference geometric pattern (see Figure 5):\nThe height of the bell tower was the same as the altar tower, excluding the domes.\nThe width of the altar tower was approximately double of the bell tower.\nExterior made of red bricks. But we noted that it could have been repainted.\nBoth towers could feature either round, semi-round, pitched or onion type domes.\nIn\nchurch architecture\n, altar towers are normally oriented to the East. In almost all cases in Odessa there was no exception. This meant Ruins (1 & 8) were very likely to be located southeast of the church. A very important check to do when finding potential geolocations.\nFigure 5: Dimensional characteristics of similar church in Odessa oblast.\nThe Ruins\nGroup Ruins (1 & 8) featured what it seemed to be the ruins of a small industrial facility. Scattered precast pillars could be seen on the ground. There were walls with faded pink tones, large holes, and bullet marks on many of them. The holes seemed similar to those left by a projectile impact. Presumably, these ruins were the result of previous conflicts in the area.\nBut where to find this type of ruin in an area of 40,000 km\n2\nincluding Odessa oblast and west Moldova? Furthermore, do the ruins still exist or have the sites been cleared up to give way for new developments? Also, which conflict could have inflicted such damage on a rural structure before 2001?\nUsing Google imagery, examples of ruins were gathered to establish a visual reference as shown in Figure 6.\nFigure 6: Example of ruins found in the Odessa region. Establishing a reference to locate Ruins (1& 8).\nTwo main events were considered: WWII and the\nTransnistria conflict\nin 1992. WWII front lines were studied according to literature as per the maps in Figure 7. In 1941, a system of\n3 defence rings\nwas established by the Soviets to\nprotect Odessa\n; with the outermost ring located at a 50km mark from the city. From March to April 1944, the Red Army launched offensives to liberate Odessa. In August 1944, as part of the\nJassy–Kishinev\noperation to reclaim the Moldavian SSR, the Red Army engaged a large German-Romanian contingent with heavy battles starting with the\nAkkerman landing\n.\nIn both conflicts, WWII and Transnistria, there were fronts localised on several points along the Dniester River and\nDniester Estuary\n. This line would become a focus of attention later in our search.\nFigure 7: Battle fronts in 1941 and 1944 during WWII were used as a reference to find potential matching ruin sites. The Dniester Estuary would become important in our search.\nGeolocating ImageC5 By Finding Other Locations Related To The Case\n1. Location Ruins (2,3 & 7) – The\nG\nreenhouse\nRuins (2,3 &7) appeared to be part of a period property or castle featuring a glass roof structure typical of those found in old greenhouses. The studio was reported to operate from several apartments in the city of Odessa. Their approximate location was investigated and pinned as shown on the map in Figure 8. It was believed that one or several of the locations could have been nearby their studios. Hence, we started to explore the area with Google Maps along Road F. From north to south, we looked for abandoned buildings, castles and ruins where any of the studio’s sets could have been located.\nA landmark symbol led us to the ruins of\ngreenhouse Marazli\n, originally built in\n1889\n. Pictures found in Yandex dated back to 2016 were gathered and analysed. Many features in its interior matched those found in the censored thumbnails of ruins (2, 3 & 7), especially the angle of the glass roof structure, supporting beams, door features, windows and stone masonry (See Figure 9 ). The location served its purpose until 1990s, but various events ultimately contributed to its abandoned state. The source reported the location was used by the studio to produce several\nCSAM\nbetween July and August 2001. No signs of location Ruins (1 & 8) or the red church were noticed in the surrounding areas.\nFigure 8: Approximate location of the apartments owned by the studio in the city of Odessa. The F location allowed our investigation to get closer to a child abuse site: the greenhouse.\nFigure 9: A 1889 greenhouse in Odessa City. Pictures found in Yandex and dated 2016 were gathered and analysed. Many features in its interior matched those found in the censored thumbnails of ruins (2, 3 & 7). The source reported the location was used by the studio to produce several CSAM between July and August 2001.\n2. Location Inland (3,6 & 7) – Oleksandrivske Reservoir\nA thumbnail within the Inland (3 & 6) group featured a river or lake. A transmission tower in the background could be seen just opposite the bank from where the shot was taken. Another shot, believed to have been taken in the opposite direction to the first, featured another transmission tower, dried grasslands, and small trees in the background.\nUsing Google Earth, narrow bodies of water with these characteristics along the coast were investigated. On the south end of Oleksandrivske Reservoir, a peninsula with transmission lines on its west side was identified. Across the reservoir, another transmission tower and cables crossing above the water were noticed. By setting the time bar to 05/06/2008, the grass on the peninsula showed up as dried and small trees could be seen clearer now on the east side of the peninsula, just behind the transmission tower. A match was found (see Figure 10). The source reported this location was used by the studio to produce several\nCSAM\nbetween July and August 2001.\nChecking the rest of the reservoir for more clues, a white building with a slim tree in the front was noticed east of the previous location. These elements matched those displayed on location Inland (7). After analysing the image, it was determined the shot was taken from the south shore of the Oleksandrivske Reservoir (see Figure 10). The source reported this location was used by the studio to produce\nCSAM\nin October 2001. No signs of Ruins (1 & 8) or the red church were identified in the surrounding areas.\nFigure 10: Oleksandrivske Reservoir. Power transmission lines and landscape matched those featuring in thumbnail images for location Inland (3,6,7). The source reported these locations were used by the studio to produce several CSAM between July and October 2001.\n3. Location Inland (8) – Kuyalnik Estuary\nThe unknown location Inland (8), featured a water body and sand around the shores. Despite the low quality of the censored thumbnails, a pink/red tone to the sand was noticed. On the left side of the image, a long cliff which ended in the horizon line was visible. Looking at the background, no buildings or mountains were seen. The body of water seemed to be quite elongated.\nIn our initial searches around Odessa, we had noticed Kuyalnik Reservoir, which had pink sands. We re-visited the images of Kuyalnik for a reinspection. The result: a Google maps image featuring the same cliff on the west shore of the reservoir.\nThe cliff exhibited the same angular lines converging downwards and touching the horizon. Dark pink sand was also noticed as well as curved water marks. Another location verified. The source reported the location was used by the studio to produce\nCSAM\nin September 2001. No signs of Ruins (1 & 8) or the red church were identified in the surrounding areas.\nFigure 11: Kuyalnik Estuary. The source reported the location was used by the studio to produce CSAM in September 2001.\n4. Location Coast (1 & 6) – Mykolaivka Village (Dniester Estuary)\nMany of the censored images within group Coast (1 & 6) featured a peninsula with a slim tree on it. The images seemed to have been taken on a shore very close to a grassy cliff.\nRocks bigger than 1.5 meters in diameter, which had become detached from the cliff, appeared in almost every shot from different angles. All main elements were numbered (1 to 10) and organised by scene (A to D). The tree and the peninsula were selected as landmarks with more potential to be identified in satellite images.\nHaving checked other bodies of water on the northeast side of the coastal region of Odessa, the southwest area around the\nDniester Estuary\nwas to be explored using Google Earth. On the east coast of the estuary at the level of the village\nMykolaivka\n, a small peninsula was found. This peninsula had several trees on it but only one produced a long shadow relative to the size of the peninsula and matching the relative position to the one shown in the censored thumbnails.\nFrom the 2014 satellite image, large rocks could be seen all along the shore. After careful analysis, all rocks and trees were identified. Another image geolocated. The source reported the location was used by the studio to produce several\nCSAM\nbetween the end of July and beginning of August 2001.\nFigure 12: Location Coast (1& 6). All main rock elements on thumbnail images identified along the east coast of the Dniester Estuary at the level of the village Mykolaivka. The source reported the location was used by the studio to produce several CSAM between the end of July and beginning of August 2001.\n5. The\nFinal Search And Verification\nAfter geolocating four different sites related to the case, a 50 km radial perimeter was established around the city of Odessa. This perimeter coincided with one of the battle fronts that took place during WWII on the east shore of the\nDniester Estuary\n. Hence, it was believed that some type of ruins, if linked to the war at all, were likely to be found around this area.\nFigure 13: Geolocation of the selected group of images marked in red.\nThe main road connecting the village of\nMykolaivka\nand\nOvidiopol\n(T1625) was checked for any signs of the red church or ruins. When doing a visual inspection with Google Earth around the village of\nKalahliya\n(\nКалаглія\n/ Калаглия ), a church was noticed at coordinates 46.281507, 30.358500 on June 2. Its name:\nЦерква Миколая Чудотворця\n, i.e. the Church of St Nicholas. All the features matched our references. A Twitter user,\nLorenzo Romani,\nalso confirmed finding the same church on June 22 through someone in his network.\nFigure 14: St. Nicholas Church, located in the village Kalahliya. The church matched all features recognised on the image related to ImageC5.\nBut was this truly the right place? Were there any ruins around? Checks were done in the immediate surroundings southeast of the church but there were no signs of any recognisable ruin structure. Then, the historical view of Google Earth was changed. A compound in ruins was visible in other years, with 2007 being the earliest image available.\nDespite the lack of street views in Google Earth and Yandex for this location, Images and YouTube videos of the village were scrutinised in order to identify all the buildings visible on ImageC5. A\ndrone video\nfound offered the best view of the site and it helped us not only to identify each of the elements but also to estimate the position of the camera (See Figure 15 and Figure 16).\nFigure 15: ImageC5 buildings identified on a YouTube drone view of the village Kalahliya\nFigure 16: Northeast corner of the ruins as the approximate camera position from where the ImageC5 landscape photograph was taken.\nFurther checks were done on two of the buildings featuring on ImageC5: the single pitch roof building (marked in yellow) and the brick construction on the far right (marked in red). We found videos recorded from Road T1625 which connects Kalahliya with Ovidiopol and passes just behind these buildings. The type of windows, roof, masonry colours as well as relative position matched those shown on satellite views as well as ImageC5 (see Figure 17).\nFigure 17: Videos and satellite images used to analyse main building construction features such as masonry colour, roof, and windows type for selected buildings. All elements matched those on ImageC5.\nSatellite images and all thumbnails in group Ruins (1 & 8) were carefully analysed to extract information such as the positions of trees, shadows sizes, sunlight direction, vegetation, exterior paint colours, and the relative position of debris. A solar calculation was done for early September 2001 as representative data for the month when the\nCSAM\nwas generated.\nPutting all of the puzzle pieces together, a basic sketch of the ruins was done to understand more about the site and to indicate the relative points where the thumbnail photos were taken from (See Figure 18).\nThe size of the construction and the presence of gates to the southeast and northwest ends of the main structure reconfirm the possibility that the building was an old, small-sized industrial/army facility. The reconstruction should be taken only as an intuitive guide only. With all elements identified, the site has been verified and ImageC5 has been now officially geolocated. The source reported the location was used by the studio to produce several\nCSAM\nbetween August and September 2001. A summary of the geolocation process followed is offered in Figure 19.\nFigure 18: Basic reconstruction of ruins based on solar data for September 2001, satellite images as well as the image thumbnails found in the investigative files. The source reported the location was used by the studio to produce several CSAM between August and September 2001.\nFigure 19: A summary of the strategy followed to geolocate ImageC5 after the finding of the investigative files. Geolocating other images within the set allowed the team for a relative quick converge into the final findings.\n“Studio A” And Their Associated Websites\nAlthough our main goal was to share information on how ImageC5 was geolocated, we also want to offer a brief summary of the operation of the studio and their websites as reported by the source. Though almost 20 years have passed and there is little reference to it on the internet, this studio impacted the world of\nCSAM\nproduction and contributed to its development. It is a clear example of how purveyors of\nCSAM\nadopt the guise of benefactors. First – they pay off the victims’ typically impoverished families. Second – they develop a child educational and entertainment programme designed to quell the parents’ and the public’s distrust in the studio. Finally, they take advantage of then-underdeveloped\nCSAM\nlegislation in Ukraine and an emerging worldwide internet. This is a business model that is still being followed by many modern criminal organizations.\nThis following section was extracted from the files and by no means expresses the views of the authors – it is rather meant to raise awareness of how such criminal enterprises operate. For legal reasons and for protection of the victims and offenders, we cannot disclose the source. All names used here are fictitious\n.\nStudio A\nwas a company founded by Denis M. & Sergey P. in early 2000s with funds coming from several European financiers. Denis was an IT expert and Sergey was a photographer known for nude “artistic” work with teens. Both individuals met in the summer 2000 and made several arrangements with other webmasters and studios to set up\nWebsite1\n. Sergey would generate the photographic content and be the main model finder; whereas Denis would be in charge of the technical side. By September of the same year,\nWebsite1\nwas fully operational, with material being produced by Sergey in Ukraine. All websites associated operated on a subscription basis. The studio operation included transportation, admin, and security crews. The latter were mainly previous law enforcement officers.\nMost of the underage children featuring in their productions were scouted around schools in eastern Europe – mainly Moldova. They usually came from impoverished backgrounds and were offered “modelling work” to help their families with money equivalent to a month’s salary of their parents. Knowing the conditions required by the studio, many parents reportedly agreed to send their children from Moldova to Ukraine with Studio A. The victims lived in different apartments owned by the studio across the city and received computer, arts and English lessons when they were not “modelling.” In efforts to further legitimize the criminal enterprise, the studio organized outings with the parents. Also, according to interviews reported in the files, Denis had falsely claimed on several occasions that the business was “socially responsible” and part of a charitable project in partnership with a worldwide human rights organization.\nAccording to the files, Sergey, who managed the business independently, had a serious addiction to heavy drugs. According to victims’ testimonies, as per the files, Sergey not only photographed victims naked but would then physically sexually assault them and offer them addictive substances. What ultimately happened to Denis and Sergey’s partnership remains unclear, but it is reported that tensions between Denis and Sergey increased and they parted ways in early 2001. Denis renamed the company\nStudio A\n* and Sergey kept the original enterprise.\nStudio A\n. Sergey convinced a number of parents and children to stay him, while another of other children exploited by\nWebsite1\nmoved on to\nStudio A*\n.\nDenis then created another website,\nWebsite2\n. The initial material used for this website came from Russia-based projects that featured nude underage girls. Later, Denis started to create new material for which he recruited new victims in Moldova. Sergey kept most of the material created for\nWebsite1\nand sold it to other parties on the web. Sergey kept working for other studios and also created a new brand,\nStudio C\n. Both\nWebsite1\nand\nWebsite2\nwent down starting in 2002. According to Denis,\nWebsite2\nwas hijacked and redirected to another page that apparently contained more explicit abuse material. According to the files, Denis placed responsibility for that on organizations based in Russia.\nAfter the 9/11 attacks in the U.S., federal authorities and international agencies started to investigate all sort of irregularities on the web more closely. It is likely that Ukrainian authorities at this point came under more pressure, while the police had information about the frequent trafficking of minors from Moldova to Ukraine. In 2002, two women travelling with four children were detained in a train station close to the Ukraine/Moldova border for interrogation. They were carrying false documentation to take these minors abroad. During questioning, all underage girls mentioned they were paid an average of $50 per session in exchange for taking part in explicit photography produced by two individuals: Denis & Sergey. The police proceeded to raid all studios and seized all computer equipment, drugs, and\nCSAM\n. Both Denis and Sergey had fled by then, presumably alerted by their corrupt contacts on the police force.\nDenis moved to Switzerland and then to Moldova, whereas Sergey was reported to have lived in Moldova, Germany and the U.S. The Ukrainian government initiated a case against both individuals in 2002 for production, sale, and distribution of\nCSAM\n. Surprisingly, charges against Denis were dropped in 2003. Ukrainian legislation at that time did not have any provisions for\nCSAM\nproduction.\nAccording to the source, Denis relaunched\nWebsite1\n, but without any nude material. He also created a new studio called\nStudio B\nand closed\nStudio A*.\nBy 2004, both websites\nWebsite1\nand\nWebsite2\nwere down. In 2007, Denis married one of the girls that worked for the studios – it is unclear how old she was at the time and when she was first recruited – and the ceremony reportedly took place in Moldova. A documentary film was also produced to portray Denis and girls who worked for the studio in more recent years. This production was never commercialised.\nMost of the\nCSAM\nSergey produced remains undocumented due to missing data. Sergey was put on a wanted list by police in Germany in relation to abuse of minors as well as for\nCSAM\ndissemination. There is no information available as to determine whether Sergey was prosecuted or not.\nSome key facts after analysing the information presented in the investigative files:\nImageC5 was produced for Website2\nAt least 60 victims between the ages of 8 and 17 were hired to work for Website2\nMore than 7300 images were produced in less than a year\nThe website published approximately 270 sets of images\n60% of the\nCSAM\nwas produced in studio facilities, while 40% was produced in outdoor locations\nOn average, 3% of the production was done in each one of the Ukrainian outdoor locations with an 8% maximum produced in Russia.\nThere were seven “photographers” working for the page including Denis and other unidentified individuals associated with the studio.\nThe source reported the studio employed assistants, drivers, teachers and security staff. The latter have had previously worked in law enforcement.\n$250,000 were received on the personal accounts of Denis and Sergey for the operation carried in Ukraine.\nConclusion & Recommendations\nImageC5 on its own contained no relevant information. The initial geolocation process for this image proved to be unproductive via employing traditional methods. More details were needed to be able to take the research further.\nThis time, an in-depth literature review on child sexual abuse cases supplied the information required to narrow down the search. Without knowing the details of the case, ImageC5 would have been never geolocated. By finding other locations related to the case, valuable landscape, architectural, war zone, and terrain data was collected. The iteration process of the search converged quicker towards the location where ImageC5 photograph was taken from.\nAfter an intensive research, Images are normally geolocated accurately. A great deal of information related to the cases is usually gathered through the process. Not all details found in our investigations are shared – due to the need to protect victims. On this case, the investigative files found came from an anonymous source which Europol has been informed about.\nEighteen years have passed since the discovery of the dark operation of\nStudio A\n. The case was known to both Ukrainian and international authorities. If data and privacy laws permit, It could be helpful to see that for any child abuse case already processed by law enforcement bodies in Europe, all\nCSAM\nimages seized should be entered into a centralised database. Law enforcement agencies could then run algorithms for an internal image reverse search to assist future geolocations and investigations. In this way, images for which there is already a background case would be left out of any crowdsourcing campaign. Moreover, if the amount of data collected is large, maybe the algorithm could be capable of comparing landscapes and other information between old and new images, thus drawing a preliminary geographical search area or creating leads for new cases.\nThe\nStudio A\ncase is a clear example of child sexual abuse and human trafficking. The operation persuaded stakeholders of their “socially responsible” approach to leverage their business. The production of\nCSAM\nwas allowed by an archaic legal framework and corruption, but also by the families of the victims, who exploited their children for financial gain. This business model has been copied by many crime organizations disguised as “modelling studios.” After many years, the situation in Ukraine has changed – but it has changed for the worse. The problem has been aggravated by the development of technology and faster communications. There are new forms of criminality and an intensified production and dissemination of child abuse material in the Ukraine-Moldova-Belarus-Russia region.\nIn sharing our investigations and search methods we believe we can contribute with authorities to bring offenders to justice, and that we can make the #StopChildAbuse network stronger while raising awareness about the problem. Our intention is to encourage members of the public and OSINT professionals to be vigilant and pass any known information to the competent authorities. Although some images published by Europol might seem old, new cases might be related to those pictures and any lead or input into their database could mean the difference between continuous abuse or rescuing the victims. At Bellingcat, we are strongly committed to these efforts.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F1.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F2.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F3.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F4.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F5.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F6.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F7.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F8.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F9.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F10.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F11.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F12.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F13.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F14.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F15.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F16.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F17.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F18-847x1200.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/09/F19.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2019-12-05_two-europol-stopchildabuse-images-geolocated-part-i-madagascar.json b/bellingcat_geolocation_json/2019-12-05_two-europol-stopchildabuse-images-geolocated-part-i-madagascar.json
new file mode 100644
index 0000000000000000000000000000000000000000..e4d791b1931e7f9168946fb0b8ff6cd830420735
--- /dev/null
+++ b/bellingcat_geolocation_json/2019-12-05_two-europol-stopchildabuse-images-geolocated-part-i-madagascar.json
@@ -0,0 +1,122 @@
+{
+ "slug": "two-europol-stopchildabuse-images-geolocated-part-i-madagascar",
+ "title": "Two Europol StopChildAbuse Images Geolocated: Part I — Madagascar",
+ "url": "https://www.bellingcat.com/news/2019/12/05/two-europol-stopchildabuse-images-geolocated-part-i-madagascar/",
+ "date": "2019-12-05T20:23:59+00:00",
+ "modified": "2019-12-17T20:27:00+00:00",
+ "author": "Carlos Gonzales",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Europol is currently in possession of more than 40 million images of child sexual abuse from across the world. In June 2017, it launched a crowdsourcing campaign called Stop Child Abuse – Trace an Object. Through this campaign, censored extracts from explicit images are regularly published on their website and members of the public are […]",
+ "content_html": "
\n
Europol is currently in possession of more than 40 million images of child sexual abuse from across the world. In June 2017, it launched a crowdsourcing campaign called Stop Child Abuse – Trace an Object. Through this campaign, censored extracts from explicit images are regularly published on their website and members of the public are asked to help tracing their location or country of origin.
\n
These tips are then used to inform a competent law enforcement authority to further investigate the lead and to assist in the identification of the offender and the victim. As of October 25, 2019, Europol has received more than 24,000 tips, which have led to the identification of nine victims and the prosecution of two offenders.
\n
In a previous article, Bellingcat wrote about the discovery of a link between one of these images and the story of a “child modelling studio” trafficking children for the production of Child Sexual Abuse Material (CSAM) in Odesa, Ukraine. On Twitter, Europol acknowledged our investigative methods which, according to them, significantly contribute to tackling online abuse.
\n
In this new article, we will show how two new batches of images published by Europol, Group ImageG13 and Group ImageC20, have been respectively geolocated to Madagascar and Cambodia. Both locations were found to be in tourist destinations, close to hotel facilities and very likely related to sex tourism. We will present a summary of the process we followed to confirm these locations and the likely time when the pictures were taken. This is Part I of the investigation and it will deal with the case in Madagascar. The upcoming Part II will address the case in Cambodia.
\n
Figure 1: Group ImageG13 (left) and Group ImageC20 (right) as published by Europol on their website
\n\n
\n
Nosy Kely, Madagascar
\n
1: The Location
\n
Group ImageG13 was geolocated to coordinates (-20.301825, 44.266543) in Nosy Kely, a small village south of the town centre of Morondava, a city in Madagascar.
\n
Figure 2: Group ImageG13 A & B as published by Europol on their website #StopChildAbuse #TraceAnObject. The group of images was geolocated to the village of Nosy Kely, Madagascar; just south of the town of Morondava. The red pointer indicates the approximate spot where the images were taken from. Google satellite image, 2004
\n
\n
2: Initial Observations
\n
– After analysing ImageG13-A & B from Figure 2, it was determined the two could be joined to form a panoramic view of the landscape. The images were edited to obtain Figure 3. We will call this edited version ImageG13 in the text for the purpose of clarity.
\n
Figure 3: Group ImageG13 edited and joined to obtain a wider view of the beach
\n
– The landscape shown on ImageG13 was believed to be a beachfront.
\n
– A row of pine trees and a group of palms were observed to the left and the right of ImageG13 respectively. The landscape showed no sign of high elevations or mountains on the horizon.
\n
– The area covered by grey sand was large and split by a line of water along the beach. Three options were initially considered: water remaining during a low tide, flowing water coming from a river mouth or a combination of both. A few tree branches were also observed in the water body. These were likely to be mangroves as seen in Figure 4. This increased the possibility of a river mouth in the vicinity.
\n
– A group of houses were visible along the beachfront, believed to be of rural construction. In front of one of these groups, a dark brown object was observed on the sand. This was likely to be a boat which suggested the possibility of fishing activity in the area. If this was a touristic place, there were likely to be hotels in the area.
\n
Figure 4: Branches in the water body were observed. These were believed to be mangroves. A dark object on the sand in front of the houses was thought to be a boat which indicated possible fishing activity in the area
\n
– Footprints across the sand area were visible. Close to one of the footprints, a small solid body believed to be a stone was observed as seen in Figure 5. For such a small body to cast a shadow approximately twice its height, the sun was likely to be at low altitude. ImageG13 was likely taken either in the early hours of the morning or late afternoon; with the beach likely oriented east or west respectively. The sun’s direction was also confirmed by the shadows cast by a group of houses. These observations would constitute an important element to consider when screening the coasts to explore in the following section.
\n
Figure 5: Footprint marks across the sand area were visible. The relatively large shadow cast by a small stone on the sand suggested the image was taken in the early morning or late afternoon; with the beach likely oriented east or west respectively
\n
\n
\n
3: The Screening Process
\n
The presence of palm trees strongly suggested the location to be tropical or subtropical. Via doing some research on the pine trees, we arrived at the conclusion that they likely belong to a genus called “Casuarina” which is native to Australia, the Indian subcontinent, Southeast Asia and the islands of the western Pacific Ocean. This type of tree was also introduced to other regions — mainly in West Africa, as shown on the map inFigure 7.
\n
Figure 6: Comparison between the pine trees [technically, she-oaks] featured on ImageG14 and an example found during the search. The tree is classified under the genus Casuarina
\n
Figure 7: Genus Casuarina is native to Australia, the Indian subcontinent, Southeast Asia, and the islands of the western Pacific Ocean. The genus was also introduced in other regions such as West Africa
\n
Based on this map and the initial observations in the previous section; three preliminary areas for exploration were selected in Figure 7 as a seed to start our iteration process:
\n
\n
The south coast of Australia
\n
The Mozambique channel
\n
The west coast of India
\n
\n
The main features sought in beaches within these regions were:
\n
\n
North-west or south-east orientation
\n
Casuarina trees along shores with very flat terrain and no mountains nearby
\n
Likely presence of river mouths close to the beach
\n
Low tides cycle
\n
Presence of mangroves to the front of the beach
\n
Rural architecture of houses or hotels
\n
\n\n
\n
4: Geolocating ImageG13
\n
Visually similar landscapes were found in the three regions. However, after two weeks of checking hundreds of Google Maps photos and satellite imagery on Google Earth as well as hotels on booking.com and images of beaches in Google, we found that the west coast of Madagascar shared the largest number of strong similarities with ImageG13. In specific, the coast of the Toliara Province seemed like a possible match.
\n
A collage with a selection of these images was built in the process to assist visual recognition (See Figure 8). The type of double pitch houses, casuarina trees, river mouths, colour of sands, mangroves and boats largely matched those observed in ImageG13. Given the level of confidence the visual evidence offered, It was decided Madagascar’s coast was a potential match and warranted further investigation.
\n
Figure 8: Image collage with landscape and architecture similarities found on the west coast of Madagascar. Zones with similar houses, palms, casuarina trees, mangroves and boats were found within the Toliara province
\n
\n
A Note On Child Sexual Abuse in Madagascar
\n
In trying to narrow down the search further, geographic and demographic data on child sexual abuse in Madagascar was consulted in UN and UNICEF reports, as well a case stories and assessments from ECPAT. A brief summary of the situation is offered here:
\n
\n
According to the World Bank, Madagascar is among the poorest countries in the world with 75% of the population living on less than US$1.90 per day.
\n
The Toliara province is one of the poorest in the country.
\n
The number of children exploited and prostituted continues to increase.
\n
Child exploitation and sexual abuse imagery affect all provinces but some specific areas can be highlighted:
Tourist: Nosy Be, Diégo-Suarez, Mangily, Foulpointe, Sainte Marie and Fort Dauphin. Nosy Be is a particularly noticeable centre for prostitution/exploitation according to many sources.
\n\n
\n
Domestic sex tourists outnumber international sex tourists.
\n
Sex tourism targeting children has increased in coastal cities.
\n
58% of the total number of tourists are French.
\n
Two recent cases of child sexual abuse and CSAM production have been reported:
\n
\n
Case 1: A 60-year-old Frenchman abused a 13-year-old girl in Toamasina. The main perpetrator was arrested in France in June 2013.
\n
Case 2: In 2016, two girls were sexually exploited in Nosy Be by a 59-year-old Frenchman, who police believe is part of a larger network of criminals involved in the production and distribution of child sexual abuse material. He is now in prison, awaiting extradition to France.
\n
All the towns named above were plotted on a map of Madagascar to see if some fell in the Toliara Region (See Figure 9). Effectively three of them were but could not be matched to ImageG13. Given the level of poverty and the increasing trend in sex tourism on the island, it was highly possible for the photograph to have been taken in a touristic town.
\n
Figure 9: A map of Madagascar. Towns highlighted in red are reported to be affected by high number of cases of child exploitation and the production of CSAM
\n
More images and videos were analysed for each one of the most touristic coastal places in the Toliara province: Anakao, Ifaty, Andavadoaka, Morombe, Belo Sur Mer and Morondava.
\n
A drone video filmed in Morondava showed a first close match for some of the elements featured on ImageG13. In particular; a group of houses, pine trees and a white poster located in the village of Nosy Kely.
\n
Figure 10: A YouTube drone video offered a view of Morondava Beach in the Village of Nosy Kely. A group of houses and a white poster at the front of one of the hotels in the area were preliminary identified
\n
A satellite image from 2004 clearly revealed a matching distribution of buildings and pine trees along the beach. Further verification of the area was carried out. A Flickr Photo from 2007 and a YouTube video revealed an exact match for all of the buildings featured in ImageG13 with the exception of the unit marked in blue as shown in Figure 11 and Figure 12. The shadows generated by a group of houses largely matched too. The approximate location of the camera was also determined.
\n
Having identified all these elements, the location was considered verified.
\n
Figure 11: A satellite view from 2004 compared to ImageG13. All of the buildings matched with the exception of the one marked in blue which seemed missing
\n
Figure 12: Images of Morondava from 2007 and 2013 revealed an exact match for all the buildings featured on ImageG13. The building unit marked in blue could not be identified. Below: The shadows generated by the group of houses largely match those observed in ImageG13. The palm trees relative location with respect to the houses is also a match
\n
Figure 13: The approximate position of the camera was determined to verify the location. Image credit: Timmi Allen
\n\n\n
\n
5: When Was ImageG13 Likely Taken?
\n
The earliest satellite image available for Nosy Kely dates back to 2004. On that image, the group of houses projecting shadows on each other on ImageG13, are presented lined up in a staggered organised pattern (See Figure 14). From Google Maps, the small beach location has currently more than ten hotels and restaurants in service. Given the design of the houses and the proximity to other hotels in the area; it was believed the group of houses in question could have belonged to a hotel facility.
\n
Figure 14: Houses featured on ImageG13 are seen lined up in a staggered pattern on satellite image from 2004. The group was believed to be part of a former hotel facility
\n
Based on the physical condition of the houses in 2013, a preliminary estimate of 20 years was established for ImageG13. A report from theMinistère de l’Agriculture, de l’Elevage et de la Pêche provided a comprehensive list of accommodation and catering establishments in Morondava for the year 2000. Among this list, there was a hotel called Le Royal Toera. Imagery gathered from different websites revealed the group of houses indeed belonged to this hotel (See Figure 15).
Analysing the images found on the architects and travel agents’ websites, all 16 bungalows in the grounds of the hotel were identified. They were numbered from 1-16 and grouped according to their orientation, location and style as shown in Figure 15 and Figure 16. The restaurant and the pool were also identified.
\n
On the west side, the hotel originally featured a pool with a bridge. Around this area, there were lamp posts which helped to determine the relative position of each of the bungalows in the analysis. Although many of the buildings found in the pictures were missing on the satellite imagery for unknown reasons; all the information was merged to produce a sketch in order to understand how the original hotel layout could have looked like (See Figure 17).
\n
Figure 15: Buildings in the grounds of the hotel were identified using images of the hotel between 1997-2000. Only eight bungalows out of the original 10 in the north group are visible on the satellite view from 2004. Here, the restaurant and other 6 bungalows from the south group are all missing. Bungalows 1 & 2 feature no roof in ImageG13
\n
Figure 16: Buildings in the grounds of the hotel were identified using images of the hotel between 1997-2000. The original pool and its bridge were also identified on satellite image. On the latter, the restaurant and the south group of bungalows are clearly missing
\n
Figure 17: Sketch with an approximation of how the original layout of the hotel could have looked like. The diagram is based on imagery found of the facilities and satellite images from 2004
\n
From Figure 15, it can be noticed that ImageG13 features bungalows 1-10. However, bungalows 1 & 2 seem to have no roof.
\n
Also, according to the field of view and position of the camera calculated in the previous section, the restaurant unit should have been visible on ImageG13 as well, but instead, the whole building is missing. Bungalows 1-2, the restaurant and bungalows 11-16 are missing on the satellite view from March 2004. In order to know when ImageG13 was taken, it was necessary to understand what sequence of events changed the layout of the hotel so drastically.
\n\n
\n
Establishing A Timeline
\n
Between the 11th and the 16th of March 2000, a German tourist wrote in his travel journal about his arrival to the Royal Toera Hotel and made the following observations:
\n
\n
Confirmed the hotel had 16 bungalows.
\n
The swimming pool and restaurant were fully operational.
\n
There were palm trees all around the facility but they were not higher than 1.5 m.
\n
There was no beach immediately in front of the hotel, but a rough sea instead.
\n
Workers had been repairing a breakwater/promenade structure at the waterfront that was affected by cyclones in the previous weeks.
\n
In discussion with the owners of one of the local restaurants, he described how sex tourism was affecting the town. Some tourists looked for girls as young as 14 years old.
\n
\n
We discussed earlier that bungalows 1-2 are shown with no roof in ImageG13. Given the fact that the German tourist described the relatively good conditions of the hotel facilities including the restaurant, pool and the 16 bungalows after recent cyclones; ImageG13 was most likely taken after March 2000 but before the bungalows 1-2 disappeared completely as shown in satellite images from the 25th of March 2004.
\n
Although not conclusive but rather informative, it was considered important to highlight a couple of findings and key facts which will be discussed later in the report:
\n
\n
A Curriculum Vitae from a former employee of the hotel indicates he stopped working for the Royal Toera in May 2002.
\n
In 2008, a TripAdvisor user received information about the Royal Toera having burned down several years ago. A local source who cannot be named affirmed that the fire in question only destroyed the restaurant of the hotel. The same source indicated that sometime after this event the south group of bungalows and the pool were destroyed by a cyclone. Unfortunately, no specific dates were recalled.
\n
Between April and June 2002, Apavou Hotels Groupstopped advertising their operations in Madagascar and Mayotte (including the Royal Toera). According to the news, the management contract of the hotel should have run until 2005.
\n
Presidential elections were held in Madagascar on 16th of December 2001, leading to violence between supporters of two candidates. The conflict lasted until July 2002.
\n
Prostitution represents a serious problem in the village. In 2008, Morondava was reported to be the home to at least 200 sex workers, from which at least 50% were in fact minor victims of exploitation. Night clubs and bars in Nosy Kely were frequented by sex workers to meet tourists or Malagasy visitors.
\n
\n
Morondava is an area at high risk of flooding and severe coastal erosion. Different environmental factors including cyclones, strong tides, heavy rains, drought and human activities have created an imbalance in the flow of the nearby rivers, generating a severe effect of marine erosion of the coastal sector. The mouth of the river where the Royal Toera Hotel was located is particularly sensitive to maritime submersion and flooding. More information can be found in this paper and a project for protection and mitigation of the littoral in Morondava called PALM.
\n
In an article from The Guardian from 2016, the situation above is discussed. An image in the article depicts the ruins of what they called “The remains of a hotel and swimming pool that was flooded several years ago by rising sea tides in Morondava”. It was determined the structures described belonged to the Royal Toera Hotel (See Figure 18).
\n
Figure 18: The Royal Toera photographed by The Guardian in 2016. The hotel was reported to have been flooded but no specific dates are given in the article
\n
Despite the fact that no specific dates were obtained via the sources consulted, the destruction of the hotel was most probably linked to flooding and cyclone events in the area. Southwest Indian Ocean cyclone seasons 2000-01, 2001-02, 2002-03 and 2003-04 were investigated and a summary of the activity around Madagascar is shown in Figure 19. From all the cyclones and storms investigated, only Cyprien in 2002, as well as Elita and Gafilo in 2004, are reported to have caused major damage in Morondava.
\n
Figure 19: Season summary maps for cyclones in the Southwest Indian Ocean around Madagascar
\n
In Table 1, a timeline has been created with a summary of the information gathered about the country and the hotel operation, as well as cyclones and storms which are reported to have caused severe flooding in Morondava between 2000-2004. The chronological order of the observations strongly suggests that ImageG13 might have been taken between January 2002 and March 2004.
\n
Table 1: Construction of a timeline for cyclones, storms, events and observations related to the Royal Toera Hotel between 2000-2004
\n\n\n
\n
Verification Of The Proposed Timeline
\n
On ImageG13, bungalows 3 to 8 show palm trees on their north side. One of these trees was clearly identified in other images produced between 2005 and 2017. For simplification, this palm has been labelled “B8”.
\n
Given that all images were taken from the distance at a similar angle and although the growth rate of a tree is probably not linear as it depends on many factors; a simplified calculation was proposed to determine an approximate year when ImageG13 was produced.
\n
The height of B8 was measured for each one of the shots as a linear proportion of the bungalows’ height (H) (See Figure 20). The results were plotted for each year and a linear equation obtained to calculate the year when B8’s height was 1.2 H, as observed in ImageG13. The resulting year, 2002, fits into the timeline presented in Table 1.
\n
Figure 20: Bungalows 3-8 pictured in different years. A simplified linear equation (based on the observed growth rate of Palm tree B8) was used to estimate the year when Palm B8 had the height observed on ImageG13. The resulting year, 2002, fits into the timeline of events
\n\n\n
\n
Conclusion: Part I ImageG13/Nosy Kely, Madagascar
\n
The geolocation process for ImageG13 took approximately 3 weeks in total. Further investigations were carried out to determine the year when the image was taken, which then extended the investigation for a further 2 months.
\n
The location was identified to be in Nosy Kely, Madagascar; near a now partially destroyed hotel, formerly known as Royal Toera. Despite no specific reports or images about damages caused to thehotel were found during this investigation, a timeline was created with relevant events to determine how the hotel could have been destroyed.
\n
All buildings, which used to belong to the hotel, were identified and drawn on a sketch using images available on the internet. Ten of these buildings are featured on ImageG13 with two of them presenting no roof.
\n
It is very likely that the hotel was impacted by the severe tropical storm Cyprien in January 2002. As a result, the roof of bungalows 1 & 2 might have been damaged. In 2004, after the landing of much stronger cyclones like Elita and Gafilo, a dam collapsed upstream a nearby river and caused severe flooding in the area. This most probably damaged the rest of the south and west wings of the hotel, including the pool as seen on Satellite images taken 2 weeks after the cyclone Gafilo.
\n
Apavou Hotels Group stopped advertising the Royal Toera Hotel in 2002. While in the middle of a serious political crisis in the country and hit badly by two cyclones in the same year, the management of the hotel might have decided to implement changes in the business that either justified the end of contracts and employment of staff; or, in a worst case scenario, the final closure of the hotel. This subject deserves further investigation.
\n
A dimensional analysis applied to a palm tree, observed on ImageG13 and other images of the same location for different years, helped to support the timeline proposed.
\n
Based on the evidence discussed, it has been determined that ImageG13 was most likely produced between January 2002 and March 2004. Given the number of hotels and bars in the area during this period and the reported levels of exploitation of minors, it is also very likely the image was related to sex tourism.
\n
This report will continue with the second part of the investigation, which is due shortly:
Research by Carlos Gonzales, Daniel Romein, Timmi Allen and “Bo”
\n
",
+ "content_text": "Europol is currently in possession of more than\n40 million images\nof child sexual abuse from across the world. In June 2017, it launched a crowdsourcing campaign called\nStop Child Abuse – Trace an Object\n. Through this campaign, censored extracts from explicit images are regularly published on their website and members of the public are asked to help tracing their location or country of origin.\nThese tips are then used to inform a competent law enforcement authority to further investigate the lead and to assist in the identification of the offender and the victim.\nAs of October 25, 2019\n, Europol has received more than 24,000 tips, which have led to the identification of nine victims and the prosecution of two offenders.\nIn a previous\narticle\n, Bellingcat wrote about the discovery of a link between one of these images and the story of a “child modelling studio” trafficking children for the production of Child Sexual Abuse Material (CSAM) in Odesa, Ukraine. On\nTwitter\n, Europol acknowledged our investigative methods which, according to them, significantly contribute to tackling online abuse.\nIn this new article, we will show how two new batches of images\npublished by Europol\n, Group ImageG13 and Group ImageC20, have been respectively geolocated to Madagascar and Cambodia. Both locations were found to be in tourist destinations, close to hotel facilities and very likely related to sex tourism. We will present a summary of the process we followed to confirm these locations and the likely time when the pictures were taken. This is Part I of the investigation and it will deal with the case in Madagascar. The upcoming Part II will address the case in Cambodia.\nFigure 1: Group ImageG13 (left) and Group ImageC20 (right) as published by Europol on their website\nNosy Kely, Madagascar\n1: The Location\nGroup ImageG13 was geolocated to coordinates\n(-20.301825, 44.266543\n) in Nosy Kely, a small village south of the town centre of Morondava, a city in Madagascar.\nFigure 2: Group ImageG13 A & B as published by Europol on their website #StopChildAbuse #TraceAnObject. The group of images was geolocated to the village of Nosy Kely, Madagascar; just south of the town of Morondava. The red pointer indicates the approximate spot where the images were taken from. Google satellite image, 2004\n2: Initial Observations\n– After analysing ImageG13-A & B from\nFigure 2\n, it was determined the two could be joined to form a panoramic view of the landscape. The images were edited to obtain\nFigure 3\n. We will call this edited version ImageG13 in the text for the purpose of clarity.\nFigure 3: Group ImageG13 edited and joined to obtain a wider view of the beach\n– The landscape shown on ImageG13 was believed to be a beachfront.\n– A row of pine trees and a group of palms were observed to the left and the right of ImageG13 respectively. The landscape showed no sign of high elevations or mountains on the horizon.\n– The area covered by grey sand was large and split by a line of water along the beach. Three options were initially considered: water remaining during a low tide, flowing water coming from a river mouth or a combination of both. A few tree branches were also observed in the water body. These were likely to be mangroves as seen in\nFigure 4\n. This increased the possibility of a river mouth in the vicinity.\n– A group of houses were visible along the beachfront, believed to be of rural construction. In front of one of these groups, a dark brown object was observed on the sand. This was likely to be a boat which suggested the possibility of fishing activity in the area. If this was a touristic place, there were likely to be hotels in the area.\nFigure 4: Branches in the water body were observed. These were believed to be mangroves. A dark object on the sand in front of the houses was thought to be a boat which indicated possible fishing activity in the area\n– Footprints across the sand area were visible. Close to one of the footprints, a small solid body believed to be a stone was observed as seen in\nFigure 5\n. For such a small body to cast a shadow approximately twice its height, the sun was likely to be at low altitude. ImageG13 was likely taken either in the early hours of the morning or late afternoon; with the beach likely oriented east or west respectively. The sun’s direction was also confirmed by the shadows cast by a group of houses. These observations would constitute an important element to consider when screening the coasts to explore in the following section.\nFigure 5: Footprint marks across the sand area were visible. The relatively large shadow cast by a small stone on the sand suggested the image was taken in the early morning or late afternoon; with the beach likely oriented east or west respectively\n3: The Screening Process\nThe presence of palm trees strongly suggested the location to be tropical or subtropical. Via doing some research on the pine trees, we arrived at the conclusion that they likely belong to a genus called “\nCasuarina\n” which is\nnative\nto Australia, the Indian subcontinent, Southeast Asia and the islands of the western Pacific Ocean. This type of tree was also introduced to other regions — mainly in West Africa, as shown on the\nmap\nin\nFigure 7\n.\nFigure 6: Comparison between the pine trees [technically, she-oaks] featured on ImageG14 and an example found during the search. The tree is classified under the genus Casuarina\nFigure 7: Genus Casuarina is native to Australia, the Indian subcontinent, Southeast Asia, and the islands of the western Pacific Ocean. The genus was also introduced in other regions such as West Africa\nBased on this map and the initial observations in the previous section; three preliminary areas for exploration were selected in\nFigure 7\nas a seed to start our iteration process:\nThe south coast of Australia\nThe Mozambique channel\nThe west coast of India\nThe main features sought in beaches within these regions were:\nNorth-west or south-east orientation\nCasuarina trees along shores with very flat terrain and no mountains nearby\nLikely presence of river mouths close to the beach\nLow tides cycle\nPresence of mangroves to the front of the beach\nRural architecture of houses or hotels\n4: Geolocating ImageG13\nVisually similar landscapes were found in the three regions. However, after two weeks of checking hundreds of Google Maps photos and satellite imagery on Google Earth as well as hotels on booking.com and images of beaches in Google, we found that the west coast of Madagascar shared the largest number of strong similarities with ImageG13. In specific, the coast of the\nToliara Province\nseemed like a possible match.\nA collage with a selection of these images was built in the process to assist visual recognition (See\nFigure 8\n). The type of double pitch houses, casuarina trees, river mouths, colour of sands, mangroves and boats largely matched those observed in ImageG13. Given the level of confidence the visual evidence offered, It was decided Madagascar’s coast was a potential match and warranted further investigation.\nFigure 8: Image collage with landscape and architecture similarities found on the west coast of Madagascar. Zones with similar houses, palms, casuarina trees, mangroves and boats were found within the Toliara province\nA Note On Child Sexual Abuse in Madagascar\nIn trying to narrow down the search further, geographic and demographic data on child sexual abuse in Madagascar was consulted in\nUN\nand\nU\nNICEF\nreports, as well a case stories and assessments from\nECPAT\n. A brief summary of the situation is offered here:\nAccording to the\nWorld Bank\n, Madagascar is among the poorest countries in the world with 75% of the population living on less than US$1.90 per day.\nThe Toliara province is one of the poorest in the country.\nThe number of children exploited and prostituted continues to increase.\nChild exploitation and sexual abuse imagery affect all provinces but some specific areas can be highlighted:\nUrban: Antananarivo, Mahajanga, Toamasina, Toliara, Antsiranana\nMining: Ilakaka, Moramanga and Tamatave.\nTourist: Nosy Be, Diégo-Suarez, Mangily, Foulpointe, Sainte Marie and Fort Dauphin. Nosy Be is a particularly noticeable centre for prostitution/exploitation according to many sources.\nDomestic sex tourists outnumber international sex tourists.\nSex tourism targeting children has increased in coastal cities.\n58% of the total number of tourists are French.\nTwo recent cases of child sexual abuse and CSAM production have been reported:\nCase 1:\nA 60-year-old Frenchman abused a 13-year-old girl in Toamasina. The main perpetrator was arrested in France in June 2013.\nCase 2\n: In 2016, two girls were sexually exploited in Nosy Be by a 59-year-old Frenchman, who police believe is part of a larger network of criminals involved in the production and distribution of child sexual abuse material. He is now in prison, awaiting extradition to France.\nAll the towns named above were plotted on\na map of Madagascar\nto see if some fell in the Toliara Region (See\nFigure 9\n). Effectively three of them were but could not be matched to ImageG13. Given the level of poverty and the increasing trend in sex tourism on the island, it was highly possible for the photograph to have been taken in a touristic town.\nFigure 9: A map of Madagascar. Towns highlighted in red are reported to be affected by high number of cases of child exploitation and the production of CSAM\nMore images and videos were analysed for each one of the\nmost touristic coastal places\nin the Toliara province: Anakao, Ifaty, Andavadoaka, Morombe, Belo Sur Mer and Morondava.\nA\ndrone video\nfilmed in Morondava showed a first close match for some of the elements featured on ImageG13. In particular; a group of houses, pine trees and a\nwhite poster\nlocated in the village of\nNosy Kely\n.\nFigure 10: A YouTube drone video offered a view of Morondava Beach in the Village of Nosy Kely. A group of houses and a white poster at the front of one of the hotels in the area were preliminary identified\nA satellite image from 2004 clearly revealed a matching distribution of buildings and pine trees along the beach. Further verification of the area was carried out. A\nFlickr Photo\nfrom 2007 and a\nYouTube video\nrevealed an exact match for all of the buildings featured in ImageG13 with the exception of the unit marked in blue as shown in\nFigure 11\nand\nFigure 12\n. The shadows generated by a group of houses largely matched too. The approximate location of the camera was also determined.\nHaving identified all these elements,\nthe location was considered verified\n.\nFigure 11: A satellite view from 2004 compared to ImageG13. All of the buildings matched with the exception of the one marked in blue which seemed missing\nFigure 12: Images of Morondava from 2007 and 2013 revealed an exact match for all the buildings featured on ImageG13. The building unit marked in blue could not be identified. Below: The shadows generated by the group of houses largely match those observed in ImageG13. The palm trees relative location with respect to the houses is also a match\nFigure 13: The approximate position of the camera was determined to verify the location. Image credit: Timmi Allen\n5: When Was ImageG13 Likely Taken?\nThe earliest satellite image available for Nosy Kely dates back to 2004. On that image, the group of houses projecting shadows on each other on ImageG13, are presented lined up in a staggered organised pattern (See\nFigure 14\n). From Google Maps, the small beach location has currently more than ten hotels and restaurants in service. Given the design of the houses and the proximity to other hotels in the area; it was believed the group of houses in question could have belonged to a hotel facility.\nFigure 14: Houses featured on ImageG13 are seen lined up in a staggered pattern on satellite image from 2004. The group was believed to be part of a former hotel facility\nBased on the physical condition of the houses in 2013, a preliminary estimate of 20 years was established for ImageG13. A report from the\nMinistère de l’Agriculture, de l’Elevage et de la Pêche\nprovided a comprehensive list of accommodation and catering establishments in Morondava for the year 2000. Among this list, there was a hotel called\nLe Royal Toera\n. Imagery gathered from different websites revealed the group of houses indeed belonged to this hotel (See\nFigure 15\n).\nAccording to the\narchitects\n, the hotel project was completed in 1997. In January 2000,\nApavou Hotels Group\nstarted managing\nthe Royal Toera.\nDifferent sources including the operator and a\ntravel agency website\n; described the facilities as having:\nOne restaurant\nOne pool\nSixteen bungalows\nAnalysing the images found on the architects and travel agents’ websites, all 16 bungalows in the grounds of the hotel were identified. They were numbered from 1-16 and grouped according to their orientation, location and style as shown in\nFigure 15\nand\nFigure 16\n. The restaurant and the pool were also identified.\nOn the west side, the hotel originally featured a pool with a bridge. Around this area, there were lamp posts which helped to determine the relative position of each of the bungalows in the analysis. Although many of the buildings found in the pictures were missing on the satellite imagery for unknown reasons; all the information was merged to produce a sketch in order to understand how the original hotel layout could have looked like (See\nFigure 17\n).\nFigure 15: Buildings in the grounds of the hotel were identified using images of the hotel between 1997-2000. Only eight bungalows out of the original 10 in the north group are visible on the satellite view from 2004. Here, the restaurant and other 6 bungalows from the south group are all missing. Bungalows 1 & 2 feature no roof in ImageG13\nFigure 16: Buildings in the grounds of the hotel were identified using images of the hotel between 1997-2000. The original pool and its bridge were also identified on satellite image. On the latter, the restaurant and the south group of bungalows are clearly missing\nFigure 17: Sketch with an approximation of how the original layout of the hotel could have looked like. The diagram is based on imagery found of the facilities and satellite images from 2004\nFrom\nFigure 15\n, it can be noticed that ImageG13 features bungalows 1-10. However, bungalows 1 & 2 seem to have no roof.\nAlso, according to the field of view and position of the camera calculated in the previous section, the restaurant unit should have been visible on ImageG13 as well, but instead, the whole building is missing. Bungalows 1-2, the restaurant and bungalows 11-16 are missing on the satellite view from March 2004.\nIn order to know when ImageG13 was taken\n,\nit was necessary to understand what sequence of events changed the layout of the hotel so drastically\n.\nEstablishing A Timeline\nBetween the 11th and the 16th of March 2000, a German tourist wrote in\nhis travel journal\nabout his arrival to the\nRoyal Toera Hotel\nand made the following observations:\nConfirmed the hotel had 16 bungalows.\nThe swimming pool and restaurant were fully operational.\nThere were palm trees all around the facility but they were not higher than 1.5 m.\nThere was no beach immediately in front of the hotel, but a rough sea instead.\nWorkers had been repairing a breakwater/promenade structure at the waterfront that was affected by cyclones in the previous weeks.\nIn discussion with the owners of one of the local restaurants, he described how sex tourism was affecting the town. Some tourists looked for girls as young as 14 years old.\nWe discussed earlier that bungalows 1-2 are shown with no roof in ImageG13. Given the fact that the German tourist described the relatively good conditions of the hotel facilities including the restaurant, pool and the 16 bungalows after recent cyclones;\nImageG13 was most likely taken after March 2000 but before the bungalows 1-2 disappeared completely as shown in satellite images from the 25th of March 2004\n.\nAlthough not conclusive but rather informative, it was considered important to highlight a couple of findings and key facts which will be discussed later in the report:\nA Curriculum Vitae from a former employee of the hotel indicates he stopped working for the\nRoyal Toera\nin May 2002.\nIn 2008,\na TripAdvisor user\nreceived information about the\nRoyal Toera\nhaving burned down several years ago. A local source who cannot be named affirmed that the fire in question only destroyed the restaurant of the hotel. The same source indicated that sometime after this event the south group of bungalows and the pool were destroyed by a cyclone. Unfortunately, no specific dates were recalled.\nBetween April and June 2002,\nApavou Hotels Group\nstopped advertising\ntheir operations in Madagascar and Mayotte (including the\nRoyal Toera\n). According to the\nnews\n, the management contract of the hotel should have run until 2005.\nPresidential elections\nwere held in Madagascar on 16th of December 2001, leading to violence between supporters of two candidates. The conflict lasted until July 2002.\nProstitution represents a serious problem in the village. In 2008, Morondava was\nreported\nto be the home to at least 200 sex workers, from which at least 50% were in fact minor victims of exploitation. Night clubs and bars in Nosy Kely were frequented by sex workers to meet tourists or Malagasy visitors.\nMorondava is an area at high risk of flooding and severe coastal erosion. Different environmental factors including cyclones, strong tides, heavy rains, drought and human activities have created an imbalance in the flow of the nearby rivers, generating a severe effect of marine erosion of the coastal sector. The mouth of the river where the\nRoyal Toera Hotel\nwas located is particularly sensitive to maritime submersion and flooding. More information can be found in\nthis paper\nand a project for protection and mitigation of the littoral in Morondava called\nPALM\n.\nIn an\narticle from The Guardian\nfrom 2016, the situation above is discussed. An image in the article depicts the ruins of what they called “The remains of a hotel and swimming pool that was flooded several years ago by rising sea tides in Morondava”. It was determined the structures described belonged to the\nRoyal Toera Hotel\n(See\nFigure 18\n).\nFigure 18: The Royal Toera photographed by The Guardian in 2016. The hotel was reported to have been flooded but no specific dates are given in the article\nDespite the fact that no specific dates were obtained via the sources consulted, the destruction of the hotel was most probably linked to flooding and cyclone events in the area. Southwest Indian Ocean cyclone seasons\n2000-01\n,\n2001-02\n,\n2002-03\nand\n2003-04\nwere investigated and a summary of the activity around Madagascar is shown in\nFigure 19\n. From all the cyclones and storms investigated, only Cyprien in\n2002\n, as well as Elita and Gafilo in\n2004\n, are reported to have caused major damage in Morondava.\nFigure 19: Season summary maps for cyclones in the Southwest Indian Ocean around Madagascar\nIn\nTable 1\n, a timeline has been created with a summary of the information gathered about the country and the hotel operation, as well as\ncyclones and storms\nwhich are reported to have caused severe flooding in Morondava between 2000-2004.\nThe chronological order of the observations strongly suggests that ImageG13 might have been taken between January 2002 and March 2004.\nTable 1: Construction of a timeline for cyclones, storms, events and observations related to the Royal Toera Hotel between 2000-2004\nVerification Of The Proposed Timeline\nOn ImageG13, bungalows 3 to 8 show palm trees on their north side. One of these trees was clearly identified in other images produced between 2005 and 2017. For simplification, this palm has been labelled “B8”.\nGiven that all images were taken from the distance at a similar angle and although the growth rate of a tree is probably not linear as it depends on many factors; a simplified calculation was proposed to determine an approximate year when ImageG13 was produced.\nThe height of B8 was measured for each one of the shots as a linear proportion of the bungalows’ height (H) (See\nFigure 20\n). The results were plotted for each year and a linear equation obtained to calculate the year when B8’s height was 1.2 H, as observed in ImageG13.\nThe resulting year, 2002, fits into the timeline presented in Table 1\n.\nFigure 20: Bungalows 3-8 pictured in different years. A simplified linear equation (based on the observed growth rate of Palm tree B8) was used to estimate the year when Palm B8 had the height observed on ImageG13. The resulting year, 2002, fits into the timeline of events\nConclusion: Part I ImageG13/Nosy Kely, Madagascar\nThe geolocation process for ImageG13 took approximately 3 weeks in total. Further investigations were carried out to determine the year when the image was taken, which then extended the investigation for a further 2 months.\nThe location was identified to be in Nosy Kely, Madagascar; near a now partially destroyed hotel, formerly known as\nRoyal Toera\n. Despite no specific reports or images about damages caused to the\nhotel were found during this investigation, a timeline was created with relevant events to determine how the hotel could have been destroyed.\nAll buildings, which used to belong to the hotel, were identified and drawn on a sketch using images available on the internet. Ten of these buildings are featured on ImageG13 with two of them presenting no roof.\nIt is very likely that the hotel was impacted by the severe tropical storm Cyprien in January 2002. As a result, the roof of bungalows 1 & 2 might have been damaged. In 2004, after the landing of much stronger cyclones like Elita and Gafilo, a dam collapsed upstream a nearby river and caused severe flooding in the area. This most probably damaged the rest of the south and west wings of the hotel, including the pool as seen on Satellite images taken 2 weeks after the cyclone Gafilo.\nApavou Hotels Group\nstopped advertising the\nRoyal Toera Hotel\nin 2002. While in the middle of a serious political crisis in the country and hit badly by two cyclones in the same year, the management of the hotel might have decided to implement changes in the business that either justified the end of contracts and employment of staff; or, in a worst case scenario, the final closure of the hotel. This subject deserves further investigation.\nA dimensional analysis applied to a palm tree, observed on ImageG13 and other images of the same location for different years, helped to support the timeline proposed.\nBased on the evidence discussed, it has been determined that ImageG13 was most likely produced between January 2002 and March 2004\n.\nGiven the number of hotels and bars in the area during this period and the reported levels of exploitation of minors, it is also very likely the image was related to sex tourism.\nThis report will continue with the second part of the investigation, which is due shortly:\nPart II – ImageC20 Siem Reap, Cambodia\nResearch by Carlos Gonzales, Daniel Romein, Timmi Allen and “Bo”",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/1-scaled.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/3-scaled.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/4.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/5.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/6.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/7.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/8-scaled.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/9.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/10-scaled.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/11-scaled.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/12-scaled.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/13.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/14.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/15.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/16.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/17.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/18.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/19.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/Table-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/20.png",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2019-12-17_two-europol-stopchildabuse-images-geolocated-part-ii-cambodia.json b/bellingcat_geolocation_json/2019-12-17_two-europol-stopchildabuse-images-geolocated-part-ii-cambodia.json
new file mode 100644
index 0000000000000000000000000000000000000000..602399fa60d3cb4349ba76f933edcf058de7e48a
--- /dev/null
+++ b/bellingcat_geolocation_json/2019-12-17_two-europol-stopchildabuse-images-geolocated-part-ii-cambodia.json
@@ -0,0 +1,157 @@
+{
+ "slug": "two-europol-stopchildabuse-images-geolocated-part-ii-cambodia",
+ "title": "Two Europol StopChildAbuse Images Geolocated: Part II — Cambodia",
+ "url": "https://www.bellingcat.com/news/2019/12/17/two-europol-stopchildabuse-images-geolocated-part-ii-cambodia/",
+ "date": "2019-12-17T14:23:30+00:00",
+ "modified": "2019-12-18T08:44:00+00:00",
+ "author": "Carlos Gonzales",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "In the previous part of this investigation, we showed how a batch of images published by Europol, Group ImageG13, was geolocated to a small coastal village in southwest Madagascar. Based on the evidence discussed, it was determined that ImageG13 was most likely produced between January 2002 and March 2004. Given the number of hotels and […]",
+ "content_html": "
\n
In the previous part of this investigation, we showed how a batch of images published by Europol, Group ImageG13, was geolocated to a small coastal village in southwest Madagascar. Based on the evidence discussed, it was determined that ImageG13 was most likely produced between January 2002 and March 2004. Given the number of hotels and bars in the area during this period and the reported levels of exploitation of minors, it is also very likely the image was related to sex tourism.
\n
Part II of the investigation analyzes another batch of images, ImageC20, which have been geolocated to Siem Reap, Cambodia. Similarly to the previous case, the location was found close to a hotel building and is most probably related to sex tourism too.
\n
On this occasion, the city district investigated was found to be visited by thousands of local and international tourists during a yearly festival. The amount of images recorded by visitors allowed us to build a timeline to chronolocate ImageC20 by observing a nearby construction site as a reference. The preparation of a 3D model in conjunction with solar data and shadow analysis allowed us to estimate the approximate year, month and time of the day when ImageC20 was taken.
\n
\n
Siem Reap, Cambodia
\n
1. The Location:
\n
Group ImageC20 was geolocated to coordinates (13°21’24.90″N 103°51’31.60″E) in Siem Reap, Cambodia at the level of the Wat Bo village in the city centre.
\n
Figure 21: Group ImageC20 A,B & C as published by Europol on their website #StopChildAbuse #TraceAnObject. The images were geolocated to Siem Reap, Cambodia near the Wat Bo village (bottom). The red mark indicates approximately where the pictures were taken: an empty plot of land now occupied by a shopping centre and situated next to a hotel formerly known as the “Angkor Sayana Hotel & Spa”.
\n
\n
2. Initial Observations
\n
–ImageC20-A,B & C from Figure 21 were assumed to be taken on the same plot of land. In order to determine the site layout, the main elements on each image were identified and numbered. An analysis based on these elements and their respective shadows allowed us to establish a connection between the three images.
\n
– The wall featured on ImageC20-C was believed to be a continuation of the wall [7] featured on ImageC20-B. By assuming this, different sections of a building are also visible behind the fences and bushes. A sketch of a hypothetical building [1] behind the fence, served as the connection between the three pieces as shown in Figure 22. All of the puzzle pieces were put together indicating that the offender did the three shots pointing towards the same wall.
\n
Figure 22: Group ImageC20 rearranged to define the land plot orientation based on sunlight direction, shadows and other elements seen in the images. A building was believed to be located behind the fence and some of its parts appeared on both ImageC20 A & C
\n
– On the right side of ImageC20-C, just behind the fence, a window [2] located just between a tree [6*] and the bushes was visible. On the top left corner of ImageC20-A, a building overhang [3] was noticed. A shadow seemed to be projected on another nearby building surface [19], indicating that the building [1] or another object that was not visible was probably taller than the adjacent houses.
\n
– ImageC20-A & B feature a building with pink roofing sheets [4]. They also show a pile of long, straight wooden poles [5] lying on the ground between trees [6] and a fencing wall [7].
\n
– The borders of the land plot suggested an improvised construction. The bottom wall section [7], made of bricks, seems to have lost its render and painting. Above and along the same wall, a top section made from galvanised-type roofing sheets [9,12] was noticed. Some of the sections of the lower brick wall featured metal spikes [11].
\n
– Power line cables and a post [14] were noticed on the top left corner of ImageC20-C. This indicated the plot of land was likely to be next to a road. Based on the multiple number of cables hanging, the property was likely to be in a large village. More details of the electricity post [14] were found by zooming in and reducing the brightness of the original ImageC20-C. The post seemed to be made of concrete and of trapezoidal shape. It also featured several reinforcing concrete ribs as shown in Figure 25. This element was of extreme importance when carrying out the screening process detailed in the following section.
\n
– The original title posted by Europol on ImageC20-C was the following: “This is most likely a village in Asia, do you recognise where is it located?”. This statement was used to centre the initial search efforts in Asia. Further analysis on vegetation led the investigation team to believe it could be in southeast Asia.
\n
– Thick and high tree trunks [18] elevate in a straight line behind the posts and cables. There are also several trees of the same type within the land plot [6,6*].
\n
– Also, on the left side of ImageC20-C, an improvised tent was visible. The tent seemed to be made of a blue-green tarpaulin held together by a wooden structure [16]. On one side of the tent, there were clothes hanging as if they were drying after having been washed. On the other side of the tent, to the left, a red parasol was visible [17]. The whole configuration indicated the possibility of people living or working under this tent.
\n
–Behind the steel roofing sheet fence [12], the tip of a similar blue tent structure is visible[13].
\n
–Based on all this information, a preliminary sketch of the site was produced to assist the geolocation process. Using this sketch, general urban layouts from satellite imagery could be quickly assessed as potential leads for a geolocation.
\n
\n
Figure 23: Preliminary sketch of the site produced to assist the geolocation process. The layout is based on the initial observations described and the connection of ImageC20-A,B&C as described previously.
\n
\n
3. The Screening Process
\n
Initial efforts focused on finding urban centres in southeast Asia that could lead to a village featuring the following elements:
\n
–Houses with pink roofing sheets type [4].
\n
–Concrete electricity posts type [14].
\n
–Tall trees type [18] along the roads.
\n
The following countries were checked for potential geolocation matches: Thailand, Singapore, Vietnam, Indonesia, Laos and Cambodia. After a full day of navigating via street view in Google Earth through the principal cities of these countries, only Siem Reap was shortlisted as a potential candidate. This city presented the largest number of similarities to the source photograph, including vegetation, buildings, fencing and, most importantly, electricity posts type [14]. Some of these similarities are listed in Figure 24–Figure 27 in the exact order they were found.
\n
\n
Figure 24: Top: A section of ImageC20-B. Bottom: Pink roofing sheets was the main material observed from satellite on many of the houses in Siem Reap
\n
\n
Figure 25: Left: edited image of the electricity post featured on ImageC20-C. Right: matching sample found on the main roads of Siem Reap. The posts are of trapezoidal shape with two partially hollow faces. The post is reinforced with ribs up to approximately 3/5 of its total height
\n
\n
Figure 26: Middle: A section of ImageC20-B showing a fence. Top and Bottom: Fencing samples observed in houses across Siem Reap. They feature similar brickwork, height and coping to those featuring in ImageC20-B. In some low-income residential areas and construction sites, galvanised steel sheets are a popular solution to increase the height of walls for security reasons. Some of them feature zig-zag spikes made by cutting the sheets as noticed on ImageC20. The type of young palm trees were very similar.
\n
\n
Figure 27: Similar trees to those featuring in the background of ImageC20 were mainly found along the Siem Reap River which crosses the city from north to south. Two adjacent main roads run along its path: “River Road” (northbound) and “Pokambor Avenue” (southbound).
\n
\n
4. Geolocating Group ImageC20
\n
All the findings listed in the screening process strongly indicated that group ImageC20 was likely to be geolocated in Siem Reap. A satellite image of the city revealed a large concentration of trees type [18] along the Siem Reap river which has been marked in yellow in Figure 28. Parallel to the river, there are two main roads called “River Road” (northbound) and “Pokambor Avenue” (southbound). Both have been fully fitted with concrete posts type [14] as seen in 2013 Google maps street view.
\n
Figure 28: A dense belt of trees type [18] was identified along the Siem Reap River. Two main roads run along the river: “River Road” (northbound) and “Pokambor Avenue” (southbound). Both became the main search area.
We took a section of the original ImageC20-C featuring a post [14] and a “tree fork” [15] next to it. This image was zoomed and flipped horizontally to recreate the relative view a person standing on the road would see as shown in Figure 29. The result was used as a visual pattern to check trees next to posts along 5 km, up and down the river. \n
Using Google street view, a match was found on River Road at the level of the “Wat Bo” village (coordinates 13°21’24.90″N 103°51’31.60″E); just between a hotel formerly known as the “Angkor Sayana Hotel & Spa” and an empty plot of land which is now occupied by a shopping centre as shown in Figure 21. The empty plot of land, and its surroundings, largely matched the sketch drawn in Figure 23 in the previous section.
\n
\n
Figure 29: A visual pattern was used to check every pair of tree-posts located along 10 km of road along the river
\n
\n
Now that a match for the tree fork and post was found, further verifications of the site were needed. By exploring Road 22 adjacent to the plot of land, more matching elements were identified. The triangular roof elevation covered in pink roofing sheets [4], the fencing wall [7] and the overhang [3] are visible on street view from 2013 (See Figure 30).
\n
\n
Figure 30: Verification of additional matching elements from ImageC20. The pink roof sheets building, the original fencing wall, as well as the building overhang were identified using Google street view
\n
Observing the hotel from Samdech Tep Vong Street and Road 22, it was determined that the overhang [3] seen on ImageC20-A is an awning attached around the building’s perimeter above the third floor (See Figure 31). Doing further visual checks, the middle brown window on the ground floor of the hotel matched the relative position, style and colour of window [2] seen on ImageC20-C (See Figure 32).
\n
Having identified all these elements, the location was considered verified.
\n
\n
\n
Figure 31: Google street view of the back of the hotel from Road 22 and Samdech Tep Vong Street in 2013. The awning of the hotel above the third floor was already fitted to the hotel when ImageC20 was taken
\n
Figure 32: Verification of additional matching elements from ImageC20. One of the side elevation windows on the ground floor match the relative position, height and style of the one featuring on ImageC20-C
\n
\n
\n
5. When Was ImageC20 Likely Taken?
\n
From the previous section, we know ImageC20-C was taken in a plot of land featuring several trees and a section of a building which belonged to the Angkor Sayana Hotel.
\n
The earliest satellite image available for the location dates back to April 2004. According to this image, the hotel did not exist by then. The surrounding land was occupied by houses and trees.
\n
Another satellite image from December 2015 shows the hotel; but this time, the adjacent property to the south was fully occupied by a concrete structure. Hence, ImageC20 could not have been taken at this point in time since trees are missing on the image.
\n
The investigation was then limited to the period 2004-2015. However, more information was needed to build a more detailed timeline based on the evolution of the hotel construction and its surroundings.
\n
\n
Figure 33: Satellite images of corner of River Road & Samdech Tep Vong in different years. In 2004 the hotel building (marked in red) did not exist. In 2015 the adjacent property south of the hotel was fully occupied with a concrete structure and all trees were missing. Therefore, ImageC20 could have only be taken during the period 2004-2015
\n
The original name of the Hotel was Angkor Sayana Hotel & Spa but it was recently rebranded as Pierre Hotel. The former managing company was called INUNISSION Group Co., Ltd and acquired the hotel around 2011. In January 2013, the Phnom Pehn Post reported that Innussion had the lease of the hotel for at least 10 years. The hotel opened its doors in January 2012 and this is confirmed by multiple users on TripAdvisor.
\n
An image of the Hotel from 2013 features the pool and a wall with spikes in the background (See Figure 34). This wall is located exactly where the galvanised roofing sheets [12] were on ImageC20-C. For obvious aesthetic reasons, this means that ImageC20 was likely taken before the hotel was fully finished and operating in January 2012.
\n
\n
Figure 34: Google street view of the hotel from Road 22 in 2013 (left). A TripAdvisor picture of the pool in 2013 (right). Both were compared with ImageC20-C. The brown wooden windows were already fitted to the building when ImageC20 was taken. Also the missing galvanised roofing sheets on the wall in 2013 suggests ImageC20-C was taken during the construction of the hotel before it started to operate.
\n
\n
From Figure 31 and Figure 32, there is evidence indicating that the hotel was built at least up to the third floor and fitted with awnings [3] when ImageC20 was taken. Windows around the ground level were likely already in place.
\n
Hence, the estimation of the date for ImageC20 was done based on answers to the following questions: at which point, between April 2004 and January 2012, that is, during the construction of the Angkor Sayana Hotel, was the following criteria more likely to be met?
\n
Criteria for the hotel:
\n
\n
Built up to the third floor or above with awnings [3] attached.
\n
Windows [2] fitted in the ground floor.
\n
Perimeter fence [12] constructed with galvanised sheets.
\n
\n
Criteria for the adjacent property:
\n
\n
A blue tent [16] on the front corner adjacent to the hotel.
\n
A red parasol [17] in front of the blue tent on the street side.
\n
A young mango tree close to the tent.
\n
Tall trees along the boundary with the hotel.
\n
\n
\n
Establishing A Timeline
\n
In Siem Reap, the rainy season normally runs from June until November. Flooding of the Siem Reap River is a frequent problem in the city. The town centre was reported to be severely affected by flooding in 2009, 2010 and 2011. Also, Bon Om Touk, a Cambodian Water Festival, is celebrated in November at the level of Wat Bo Bridge close to the Hotel Angkor Sayana.
\n
Using Flickr, Google maps, YouTube and news articles, images mainly related to these events were investigated. Those showing the hotel were used to build a chronological sequence of changes in the construction of the building between 2005 and 2011 (See Figure 35 to Figure 46). A timeline with the main observations issummarised in Table 2.
\n
The majority of the structure and façade of the hotel was finished between 2004 and 2007, with many of the outer accessories fitted by the end of this period (see Figure 35 to Figure 38).
\n
In the second half of 2006 the building was surrounded by scaffolding and green netting (See Figure 36 and 37). This green netting is not observed on ImageC20. Furthermore, imagery of different construction sites as found in Siem Reap revealed that wooden poles are commonly used for scaffolding. The wooden poles [5] laying on the ground on ImageC20-B (see Figure 22), were likely part of the scaffold structure used in the hotel construction. The rear awnings [3] were fitted most probably whilst the scaffolding was in place. Therefore, ImageC20 was likely taken after the scaffolding was decommissioned.
Figure 35: April 2005. View of the construction of the Hotel Angkor Sayana from across the Wat Bo Bridge. The construction of the third floor just started
Figure 36: November 2006. View of the construction of the Hotel Angkor Sayana from across the Wat Bo Bridge. The Hotel was already fitted with a roof . A section of the awning is visible. However, wooden scaffolding and green netting are still in place. No windows are fitted yet to the building
Figure 37: December 2006. View from the corner River Road with Road 22. The hotel was still surrounded with wooden scaffolding and green netting. The latter seems to fall onto the front corner of neighbouring plot of land. This is not featured on ImageC20. A row of tall trees is seen along the fencing dividing the two properties
\n
On February 2007, after scaffolds were decommissioned, the building featured no windows (see Figure 38). By April 2008, the majority of the floors were fitted with wooden frames. However, the ground floor on the north side did not have all of the windows fitted until August 2008 (see Figure 39 and 41).
Figure 38: February 2007. View of the construction of the Hotel Angkor Sayana from across the Wat Bo Bridge. The hotel no longer features scaffolding and green netting around it. No fitted wooden windows are yet visible. The awning above the third floor is clearly visible
Figure 39: April 2008. View of the Hotel Angkor Sayana from Pokambor Avenue (bottom) compared to a view from Samdech Tep Vong Street in 2013 (top). The ground floor of the hotel is not finished. The north wall and its windows are missing. The opposite side feature on ImageC20 could have been in the same conditions. The first floor seems to have the windows frames fitted though
\n
\n
On April 2008, the fencing of the hotel appeared to be made of galvanised steel sheets (see Figure 40). A blue tarpaulin, similar to the one used for the blue tent[16], is also observed on the front of the plot of land. However, its shape did not seem to match the profile of the blue tent [16] from ImageC20.
\n
\n
Figure 40: April 2008. View of the Hotel Angkor Sayana from Pokambor Avenue (north).The construction site is shown still enclosed by a fence of galvanised steel sheets. A blue/green tarpaulin is seen at the front of the adjacent plot of land (marked in pink) just to one side of the tree fork and electricity post. Although in the same position next to the hotel, this does not seem to be the tent seen on ImageC20-C and is likely a more improvised version
Figure 41: August 2008. View of the Hotel Angkor Sayana from across the Wat Bo Bridge. All floors seem to be fully fitted with wooden windows including the ground floor
\n
A picture dated October 2008 is shown in Figure 42. The long distance shot was taken from approximately this point (13°21’32.20″N 103°51’32.41″E) by the river on Pokambor Avenue and pointing towards the corner of the Angkor Sayana Hotel. Although at a long distance, many details can be observed. One footing and a pole belonging to a traffic sign post is visible. This sign is shown in clear detail on another image from 2009. Beyond the traffic sign, a red parasol largely matches the tone, shape and lettering of the one seen on ImageG20-C. Close to the parasol there is a blue area in similar position and height to the tent [16] seen on ImageC20-C.
\n
In October 2009, flooding affected the riverside area. Several images were found with different angles of the spot right behind the tree fork next to the hotel. A wooden structure similar to the one of the tent [16] on ImageC20-C is shown dismantled or collapsed.
\n
As no blue tent is seen on the corner of the plot of land, the blue tarpaulin and wooden structure seen in Figure 43 are believed to be the decommissioned tent [16] from ImageC20, partially seen in Figure 42 from October 2008.
\n
A faded red parasol is also observed in the image, with the parasol’s lettering on it looking very similar to the one featured on ImageC20-C. The faded colour could be the result of the sun striking on the same original bright red parasol observed in the Image from 2008 in Figure 42.
\n
An air compressor was identified close to the parasol. The people in charge of the shack to the right of the image may have had a small stall selling drinks and air service for tuk-tuks, motorbikes and bicycles as seen in other stalls in Siem Reap. The stall might have been operating from the blue tent [16] before its decommissioning.
Figure 42: October 2008. This image found on Flickr was shot from the approximate coordinates (13°21’32.20″N 103°51’32.41″E) on Pokambor Avenue, pointing towards the corner of the Angkor Sayana Hotel. One footing and a pole which belonged to a traffic sign post in front of the hotel is visible. Behind the traffic sign, a red parasol largely matching the tone, shape and lettering of the one seen on ImageG20-C. Close to the parasol a blue area in similar position and height to the tent seen on ImageC20-C
Figure 43: October 2009. Flooding affected the riverside area on October 1st 2009. Different views of the corner of the adjacent plot of land next to the hotel construction site. A wooden structure similar to the one featured on ImageC20-C is shown collapsed. A damaged blue/green tarpaulin is seen next to a faded red parasol. The lettering on this parasol looks very similar to the one featured on ImageC20-C. An air compressor seen close to the parasol. Image Source/ Flickr: SarahDepper – “Siem Reap in flood”
\n
Young leaves in mango trees are of purple or copper colour. This leaf type is observed on an image from October 2009 in approximately the same location of the mango tree from ImageC20-C. However, the tree observed is higher than sheet metal fencing line from the construction site. Due to perspective, the young and less dense tree from ImageC20-C is estimated to be approximately on or below this line.
\n
In ideal conditions, young mango trees can grow up to 1.5m/year. If ImageC20-C is assumed was produced in 2008, this growth rate is consistent with the approximate difference in height observed between the mango tree in 2009 and ImageC20’s on Figure 44.
\n
A blue tent [13] is also observed within the construction site. This element is featured on ImageC20 and has been highlighted in blue.
\n
Figure 44: A mango tree is observed above the fence line. A tree was identified having leaves of copper colour which is typical for young leaves. The same type of leaves are observed on ImageC20-C. An example is offered as a reference. Within the premises of the hotel construction site, a blue tent [13] is also visible. this tent is featured on ImageC20-C and has been highlighted in blue. Image Source/ Flickr: Ignacio Izquierdo – “Siem Reap Flooded”
On Figure 45, an image from November 2011 features a ChildSafe network certified tuk-tuk passing by the plot of land where children were abused. The ChildSafe Movement is a global child protection system established and powered by Friends-International. It was designed to train local citizens and tourists on how to detect and report abuse, and to teach children how to seek protection and create awareness campaigns targeting the international community. During the present investigation, other posts from other institutions were found advertising similar efforts around Siem Reap as early as 2005 (see Figure 46). Although not the objective of this report, it is important to highlight these initiatives as they might be an indicator of the fight against high levels of child abuse seen in this area of Siem Reap during the period in question. Moreover, they might have information of a child abuse network operating in the area around that time.\n
The same air compressor observed in October 2009 was identified on Figure 45 in 2011, indicating that the people in charge of the shack/stall in 2009 might have been still working/living on this site.
\n
The galvanised metal fencing was replaced by a solid wall with spikes. Also, a young mango tree is seen in the same relative position as the smaller one observed on ImageC20. The shack observed in October 2009 during the flooding was relocated to approximately the same spot where the blue tent [16] was located on ImageC20. Satellite imagery from 2010 indicates this change was already implemented in March 2010. These observations suggest ImageC20 was more likely produced prior to 2010.
Figure 45: November 2011. A ChildSafe network-certified tuk-tuk passing by the plot of land where children were abused. A similar air compressor observed in October 2009 was identified. A young taller mango tree is seen in the same relative position as the smaller one observed on ImageC20. The shack observed in October 2009 during the flooding was relocated to approximately the same spot where the blue tent [16] was located on ImageC20. Satellite imagery from 2010 indicates this change was already implemented in March 2010. The galvanised metal fencing was replaced by a solid wall with spikes
\n
Figure 46: Another example of other initiatives against child abuse seen in Siem reap back in 2005
\n
\n
Table 2: Approximate timeline of the construction of the hotel building and observations on the adjacent property. The information is used to establish a likely period of time within which Imagec20 could have been produced
\n
Using satellite and street view imagery, simplified models of the hotel building and nearby houses were prepared. Orientation and relative sizes of buildings were considered. Based on solar data for the site’s coordinates and neglecting atmospheric and surface factors, an approximation of the shadows cast by the hotel building on adjacent houses was computed.
\n
Although the shape of the resulting shadows might not be exactly the same due to the simplification of the building features and atmospheric factors, the analysis is aimed at providing an intuitive approach that could tell when it is more likely for the shadows of the building to reach the nearby building as seen on ImageC20.
\n
The results are shown in Figure 47. An approximate date was estimated.
\n
There are two options possible: early May or early August at approximately 3pm.
\n
\n
Figure 47: A simplified model of the hotel building helped to calculate an approximate date for ImageC20
\n
\n
Conclusion – Part II ImageC20/Siem Reap, Cambodia
\n
The title of the image posted by Europol helped to focus the search on Asia. After editing ImageC20-C, more features of an electricity post seen in the background of the image were revealed. This information allowed us to narrow down the search to the main roads of Siem Reap in Cambodia, where exactly the same posts were found.
\n
All the trees in the background of the image had long straight trunks and dark leaves. This type of trees were similar to those observed along the Siem Reap River road. The posts featured on ImaceC20 was next to a tree fork with the characteristics described above. A preliminary sketch of the site was prepared to assist in the search of this tree along the river road.
\n
ImageC20 was geolocated in the Wat Bo village of Siem Reap on coordinates13°21’24.90″N 103°51’31.60″E next to a hotel formerly known as the Angkor Sayana Hotel and within the perimeters of an empty plot of land.
\n
Although no specific images were found showing a clear angle of the blue tent [16] observed on ImageC20-C, the information gathered allowed to establish a timeline for the construction of the hotel to identify its progress year by year until the features observed on ImageC20 were all identified. A red parasol largely matching the one featuring on ImageC20 was also identified from an image dated October 2008.
\n
The mango tree featured on ImageC20-C was identified as a taller tree on an image from 2009. Assuming the tree followed a growth rate lower than the maximum theoretical value of 1.5m/year, It is likely the ImageC20-C was taken in 2008 in order to account for the difference in height observed in the trees until 2009.
\n
After modelling the shadows the hotel building casts on nearby houses, the results were compared against the original image to estimate the time of the year when the images were produced.
\n
Based on the evidence discussed two scenarios are possible:
\n
1)ImageC20 was likely produced between 2007 and 2009; either in early May or August
\n
2)ImageC20 was most likely produced in August 2008 at around 3 pm.
\n
There is evidence the town of Siem Reap was host of many initiatives against child abuse during the analysed period.
\n
Based on the information discussed, there seemed to be a continuous presence of people living or working within the property adjacent to the hotel between 2008 and 2011. The tenants of the shack and stall might have witnessed some suspicious activity on the property. Also, as the operation of the stall remained ongoing until 2012, the construction company, the staff working for the hotel, as well as tuk-tuk drivers regularly stopping by may still be able to provide relevant information about some of the activities that have taken place on the property.
\n
According to experts consulted, the images seem to have been taken with a phone camera. For the period of time in question, this type of technology was just emerging. Perhaps it might be possible for the authorities and experts to obtain data or other information that helps to identify the phone to then trace it back to the owner who used it in Cambodia on the dates and time proposed.
\n
\n
Research by : Carlos Gonzales, Daniel Romein, Timmi Allen and “Bo”
\n
",
+ "content_text": "In the\nprevious part of this investigation\n, we showed how a batch of images\npublished by Europol\n, Group ImageG13, was geolocated to a small coastal village in southwest Madagascar. Based on the evidence discussed, it was determined that ImageG13 was most likely produced between January 2002 and March 2004. Given the number of hotels and bars in the area during this period and the reported levels of exploitation of minors, it is also very likely the image was related to sex tourism.\nPart II of the investigation analyzes another batch of images, ImageC20, which have been geolocated to Siem Reap, Cambodia. Similarly to the previous case, the location was found close to a hotel building and is most probably related to sex tourism too.\nOn this occasion, the city district investigated was found to be visited by thousands of local and international tourists during a yearly festival. The amount of images recorded by visitors allowed us to build a timeline to chronolocate ImageC20 by observing a nearby construction site as a reference. The preparation of a 3D model in conjunction with solar data and shadow analysis allowed us to estimate the approximate year, month and time of the day when ImageC20 was taken.\nSiem Reap, Cambodia\n1.\nThe Location:\nGroup ImageC20 was geolocated to coordinates (\n13°21’24.90″N 103°51’31.60″E\n) in Siem Reap, Cambodia at the level of the\nWat Bo\nvillage in the city centre.\nFigure 21: Group ImageC20 A,B & C as published by Europol on their website #StopChildAbuse #TraceAnObject. The images were geolocated to Siem Reap, Cambodia near the Wat Bo village (bottom). The red mark indicates approximately where the pictures were taken: an empty plot of land now occupied by a shopping centre and situated next to a hotel formerly known as the “Angkor Sayana Hotel & Spa”.\n2. Initial Observations\n–\nImageC20-A,B & C from\nFigure 21\nwere assumed to be taken on the same plot of land. In order to determine the site layout, the main elements on each image were identified and numbered. An analysis based on these elements and their respective shadows allowed us to establish a connection between the three images.\n–\nThe wall featured on ImageC20-C was believed to be a continuation of the wall [7] featured on ImageC20-B. By assuming this, different sections of a building are also visible behind the fences and bushes. A sketch of a hypothetical building [1] behind the fence, served as the connection between the three pieces as shown in\nFigure 22\n. All of the puzzle pieces were put together indicating that the offender did the three shots pointing towards the same wall.\nFigure 22: Group ImageC20 rearranged to define the land plot orientation based on sunlight direction, shadows and other elements seen in the images. A building was believed to be located behind the fence and some of its parts appeared on both ImageC20 A & C\n–\nOn the right side of ImageC20-C, just behind the fence, a window [2] located just between a tree [6*] and the bushes was visible. On the top left corner of ImageC20-A, a building overhang [3] was noticed. A shadow seemed to be projected on another nearby building surface [19], indicating that the building [1] or another object that was not visible was probably taller than the adjacent houses.\n–\nImageC20-A & B feature a building with pink roofing sheets [4]. They also show a pile of long, straight wooden poles [5] lying on the ground between trees [6] and a fencing wall [7].\n–\nThe borders of the land plot suggested an improvised construction. The bottom wall section [7], made of bricks, seems to have lost its render and painting. Above and along the same wall, a top section made from galvanised-type roofing sheets [9,12] was noticed. Some of the sections of the lower brick wall featured metal spikes [11].\n–\nPower line cables and a post [14] were noticed on the top left corner of ImageC20-C. This indicated the plot of land was likely to be next to a road. Based on the multiple number of cables hanging, the property was likely to be in a large village. More details of the electricity post [14] were found by zooming in and reducing the brightness of the original ImageC20-C.\nThe post seemed to be made of concrete and of trapezoidal shape. It also featured several reinforcing concrete ribs as shown in Figure 25. This element was of extreme importance when carrying out the screening process detailed in the following section.\n–\nThe\noriginal title\nposted by Europol on ImageC20-C was the following: “This is most likely a village in Asia, do you recognise where is it located?”. This statement was used to centre the initial search efforts in Asia. Further analysis on vegetation led the investigation team to believe it could be in southeast Asia.\n–\nThick and high tree trunks [18] elevate in a straight line behind the posts and cables. There are also several trees of the same type within the land plot [6,6*].\n– Also, on the left side of ImageC20-C, an improvised tent was visible. The tent seemed to be made of a blue-green tarpaulin held together by a wooden structure [16]. On one side of the tent, there were clothes hanging as if they were drying after having been washed. On the other side of the tent, to the left, a red parasol was visible [17]. The whole configuration indicated the possibility of people living or working under this tent.\n–\nBehind the steel roofing sheet fence [12], the tip of a similar blue tent structure is visible[13].\n–\nBased on all this information, a preliminary sketch of the site was produced to assist the geolocation process. Using this sketch, general urban layouts from satellite imagery could be quickly assessed as potential leads for a geolocation.\nFigure 23: Preliminary sketch of the site produced to assist the geolocation process. The layout is based on the initial observations described and the connection of ImageC20-A,B&C as described previously.\n3. The Screening Process\nInitial efforts focused on finding urban centres in southeast Asia that could lead to a village featuring the following elements:\n–\nHouses with pink roofing sheets type [4].\n–\nConcrete electricity posts type [14].\n–\nTall trees type [18] along the roads.\nThe following countries were checked for potential geolocation matches: Thailand, Singapore, Vietnam, Indonesia, Laos and Cambodia. After a full day of navigating via street view in Google Earth through the principal cities of these countries, only Siem Reap was shortlisted as a potential candidate. This city presented the largest number of similarities to the source photograph, including vegetation, buildings, fencing and, most importantly, electricity posts type [14]. Some of these similarities are listed in\nFigure 24\n–\nFigure 27\nin the exact order they were found.\nFigure 24: Top: A section of ImageC20-B. Bottom: Pink roofing sheets was the main material observed from satellite on many of the houses in Siem Reap\nFigure 25: Left: edited image of the electricity post featured on ImageC20-C. Right: matching sample found on the main roads of Siem Reap. The posts are of trapezoidal shape with two partially hollow faces. The post is reinforced with ribs up to approximately 3/5 of its total height\nFigure 26: Middle: A section of ImageC20-B showing a fence. Top and Bottom: Fencing samples observed in houses across Siem Reap. They feature similar brickwork, height and coping to those featuring in ImageC20-B. In some low-income residential areas and construction sites, galvanised steel sheets are a popular solution to increase the height of walls for security reasons. Some of them feature zig-zag spikes made by cutting the sheets as noticed on ImageC20. The type of young palm trees were very similar.\nFigure 27: Similar trees to those featuring in the background of ImageC20 were mainly found along the Siem Reap River which crosses the city from north to south. Two adjacent main roads run along its path: “River Road” (northbound) and “Pokambor Avenue” (southbound).\n4. Geolocating Group ImageC20\nAll the findings listed in the screening process strongly indicated that group ImageC20 was likely to be geolocated in Siem Reap. A satellite image of the city revealed a large concentration of trees type [18] along the Siem Reap river which has been marked in yellow in\nFigure 28\n. Parallel to the river, there are two main roads called “River Road” (northbound) and “Pokambor Avenue” (southbound). Both have been fully fitted with concrete posts type [14] as seen in 2013 Google maps street view.\nFigure 28: A dense belt of trees type [18] was identified along the Siem Reap River. Two main roads run along the river: “River Road” (northbound) and “Pokambor Avenue” (southbound). Both became the main search area.\nWe took a section of the original ImageC20-C featuring a post [14] and a “tree fork” [15] next to it. This image was zoomed and flipped horizontally to recreate the relative view a person standing on the road would see as shown in\nFigure 29\n. The result was used as a visual pattern to check trees next to posts along 5 km, up and down the river.\nUsing Google street view, a match was found on River Road at the level of the “Wat Bo” village (coordinates\n13°21’24.90″N 103°51’31.60″E\n); just between a hotel formerly known as the “\nAngkor Sayana Hotel & Spa\n” and an empty plot of land which is now occupied by a shopping centre as shown in\nFigure 21\n. The empty plot of land, and its surroundings, largely matched the sketch drawn in\nFigure 23\nin the previous section.\nFigure 29: A visual pattern was used to check every pair of tree-posts located along 10 km of road along the river\nNow that a match for the tree fork and post was found, further verifications of the site were needed. By exploring Road 22 adjacent to the plot of land, more matching elements were identified. The triangular roof elevation covered in pink roofing sheets [4], the fencing wall [7] and the overhang [3] are visible on street view from 2013 (See\nFigure 30\n).\nFigure 30: Verification of additional matching elements from ImageC20. The pink roof sheets building, the original fencing wall, as well as the building overhang were identified using Google street view\nObserving the hotel from Samdech Tep Vong Street and Road 22, it was determined that the overhang [3] seen on ImageC20-A is an awning attached around the building’s perimeter above the third floor (See\nFigure 31\n). Doing further visual checks, the middle brown window on the ground floor of the hotel matched the relative position, style and colour of window [2] seen on ImageC20-C (See\nFigure 32\n).\nHaving identified all these elements,\nthe location was considered verified\n.\nFigure 31: Google street view of the back of the hotel from Road 22 and Samdech Tep Vong Street in 2013. The awning of the hotel above the third floor was already fitted to the hotel when ImageC20 was taken\nFigure 32: Verification of additional matching elements from ImageC20. One of the side elevation windows on the ground floor match the relative position, height and style of the one featuring on ImageC20-C\n5. When Was ImageC20 Likely Taken?\nFrom the previous section, we know ImageC20-C was taken in a plot of land featuring several trees and a section of a building which belonged to the\nAngkor Sayana Hotel.\nThe earliest satellite image available for the location dates back to April 2004. According to this image, the hotel did not exist by then. The surrounding land was occupied by houses and trees.\nAnother satellite image from December 2015 shows the hotel; but this time, the adjacent property to the south was fully occupied by a concrete structure. Hence, ImageC20 could not have been taken at this point in time since trees are missing on the image.\nThe investigation was then limited to the period 2004-2015. However, more information was needed to build a more detailed timeline based on the evolution of the hotel construction and its surroundings.\nFigure 33: Satellite images of corner of River Road & Samdech Tep Vong in different years. In 2004 the hotel building (marked in red) did not exist. In 2015 the adjacent property south of the hotel was fully occupied with a concrete structure and all trees were missing. Therefore, ImageC20 could have only be taken during the period 2004-2015\nThe original name of the Hotel was\nAngkor Sayana Hotel & Spa\nbut it was recently rebranded as\nPierre Hotel\n.\nThe former managing company was called\nINUNISSION Group Co., Ltd\nand acquired the hotel around 2011. In January 2013, the\nPhnom Pehn Post\nreported that Innussion had the lease of the hotel for at least 10 years. The hotel opened its doors in\nJanuary 2012\nand this is confirmed by multiple users on\nTripAdvisor\n.\nAn\nimage of the Hotel from 2013\nfeatures the pool and a wall with spikes in the background (See\nFigure 34\n). This wall is located exactly where the galvanised roofing sheets [12] were on ImageC20-C. For obvious aesthetic reasons, this means that ImageC20 was likely taken before the hotel was fully finished and operating in January 2012.\nFigure 34: Google street view of the hotel from Road 22 in 2013 (left). A TripAdvisor picture of the pool in 2013 (right). Both were compared with ImageC20-C. The brown wooden windows were already fitted to the building when ImageC20 was taken. Also the missing galvanised roofing sheets on the wall in 2013 suggests ImageC20-C was taken during the construction of the hotel before it started to operate.\nFrom\nFigure 31\nand\nFigure 32\n, there is evidence indicating that the hotel was built at least up to the third floor and fitted with awnings [3] when ImageC20 was taken. Windows around the ground level were likely already in place.\nHence, the estimation of the date for ImageC20 was done based on answers to the following questions:\nat which point,\nbetween April 2004 and January 2012, that is,\nduring the construction of the\nAngkor Sayana Hotel,\nwas the following criteria more likely to be met?\nCriteria for the hotel:\nBuilt up to the third floor or above with awnings [3] attached.\nWindows [2] fitted in the ground floor.\nPerimeter fence [12] constructed with galvanised sheets.\nCriteria for the adjacent property\n:\nA blue tent [16] on the front corner adjacent to the hotel.\nA red parasol [17] in front of the blue tent on the street side.\nA young mango tree close to the tent.\nTall trees along the boundary with the hotel.\nEstablishing A Timeline\nIn Siem Reap, the\nrainy season\nnormally runs from June until November. Flooding of the Siem Reap River is a\nfrequent problem\nin the city. The town centre was reported to be severely affected by flooding in 2009, 2010 and 2011. Also,\nBon Om Touk\n, a\nCambodian Water Festival,\nis celebrated in November at the level of Wat Bo Bridge close to the Hotel Angkor Sayana.\nUsing Flickr, Google maps, YouTube and news articles, images mainly related to these events were investigated. Those showing the hotel were used to build a chronological sequence of changes in the construction of the building between 2005 and 2011 (See\nFigure 35\nto\nFigure 46\n). A timeline with the main observations is\nsummarised in\nTable 2\n.\nThe majority of the structure and façade of the hotel was finished between 2004 and 2007, with many of the outer accessories fitted by the end of this period (see\nFigure 35\nto\nFigure 38\n).\nIn the second half of 2006 the building was surrounded by scaffolding and green netting (See\nFigure 36\nand\n37\n). This green netting is not observed on ImageC20. Furthermore, imagery of different\nconstruction sites\nas found in Siem Reap revealed that wooden poles are commonly used for scaffolding. The wooden poles [5] laying on the ground on ImageC20-B (see\nFigure 22\n), were likely part of the scaffold structure used in the hotel construction. The rear awnings [3] were fitted most probably whilst the scaffolding was in place.\nTherefore, ImageC20 was likely taken after the scaffolding was decommissioned.\nApril 2005\nFigure 35: April 2005. View of the construction of the Hotel Angkor Sayana from across the Wat Bo Bridge. The construction of the third floor just started\nNovember 2006\nFigure 36: November 2006. View of the construction of the Hotel Angkor Sayana from across the Wat Bo Bridge. The Hotel was already fitted with a roof . A section of the awning is visible. However, wooden scaffolding and green netting are still in place. No windows are fitted yet to the building\nDecember 2006\nFigure 37: December 2006. View from the corner River Road with Road 22. The hotel was still surrounded with wooden scaffolding and green netting. The latter seems to fall onto the front corner of neighbouring plot of land. This is not featured on ImageC20. A row of tall trees is seen along the fencing dividing the two properties\nOn February 2007, after scaffolds were decommissioned, the building featured no windows (see\nFigure 38\n). By April 2008, the majority of the floors were fitted with wooden frames. However, the ground floor on the north side did not have all of the windows fitted until\nAugust 2008 (\nsee\nFigure 39\nand\n41).\nFebruary 2007\nFigure 38: February 2007. View of the construction of the Hotel Angkor Sayana from across the Wat Bo Bridge. The hotel no longer features scaffolding and green netting around it. No fitted wooden windows are yet visible. The awning above the third floor is clearly visible\nApril 2008\nFigure 39: April 2008. View of the Hotel Angkor Sayana from Pokambor Avenue (bottom) compared to a view from Samdech Tep Vong Street in 2013 (top). The ground floor of the hotel is not finished. The north wall and its windows are missing. The opposite side feature on ImageC20 could have been in the same conditions. The first floor seems to have the windows frames fitted though\nOn April 2008, the fencing of the hotel appeared to be made of galvanised steel sheets (see\nFigure 40\n). A blue tarpaulin, similar to the one used for the blue tent[16], is also observed on the front of the plot of land. However, its shape did not seem to match the profile of the blue tent [16] from ImageC20.\nFigure 40: April 2008. View of the Hotel Angkor Sayana from Pokambor Avenue (north).The construction site is shown still enclosed by a fence of galvanised steel sheets. A blue/green tarpaulin is seen at the front of the adjacent plot of land (marked in pink) just to one side of the tree fork and electricity post. Although in the same position next to the hotel, this does not seem to be the tent seen on ImageC20-C and is likely a more improvised version\nAugust 2008\nFigure 41: August 2008. View of the Hotel Angkor Sayana from across the Wat Bo Bridge. All floors seem to be fully fitted with wooden windows including the ground floor\nA\npicture dated October 2008\nis shown in\nFigure 42\n. The long distance shot was taken from approximately this point (\n13°21’32.20″N 103°51’32.41″E\n) by the river on Pokambor Avenue and pointing towards the corner of the Angkor Sayana Hotel. Although at a long distance, many details can be observed. One footing and a pole belonging to a traffic sign post is visible. This sign is shown in clear detail on another image from 2009. Beyond the traffic sign, a red parasol largely matches the tone, shape and lettering of the one seen on ImageG20-C. Close to the parasol there is a blue area in similar position and height to the tent [16] seen on ImageC20-C.\nIn\nOctober 2009\n, flooding affected the riverside area.\nSeveral images\nwere found with different angles of the spot right behind the tree fork next to the hotel. A wooden structure similar to the one of the tent [16] on ImageC20-C is shown dismantled or collapsed.\nAs no blue tent is seen on the corner of the plot of land, the blue tarpaulin and wooden structure seen in\nFigure 43\nare believed to be the decommissioned tent [16] from ImageC20, partially seen in\nFigure 42\nfrom October 2008.\nA faded red parasol is also observed in the image, with the parasol’s lettering on it looking very similar to the one featured on ImageC20-C. The faded colour could be the result of the sun striking on the same original bright red parasol observed in the Image from 2008 in\nFigure 42\n.\nAn air compressor was identified close to the parasol. The people in charge of the shack to the right of the image may have had a small stall selling drinks and air service for tuk-tuks, motorbikes and bicycles as seen in other stalls in Siem Reap. The stall might have been operating from the blue tent [16] before its decommissioning.\nOctober 2008\nFigure 42: October 2008. This image found on Flickr was shot from the approximate coordinates (13°21’32.20″N 103°51’32.41″E) on Pokambor Avenue, pointing towards the corner of the Angkor Sayana Hotel. One footing and a pole which belonged to a traffic sign post in front of the hotel is visible. Behind the traffic sign, a red parasol largely matching the tone, shape and lettering of the one seen on ImageG20-C. Close to the parasol a blue area in similar position and height to the tent seen on ImageC20-C\nOctober 2009\nFigure 43: October 2009. Flooding affected the riverside area on October 1st 2009. Different views of the corner of the adjacent plot of land next to the hotel construction site. A wooden structure similar to the one featured on ImageC20-C is shown collapsed. A damaged blue/green tarpaulin is seen next to a faded red parasol. The lettering on this parasol looks very similar to the one featured on ImageC20-C. An air compressor seen close to the parasol. Image Source/ Flickr: SarahDepper – “Siem Reap in flood”\nYoung leaves in mango trees\nare of purple or copper colour. This leaf type is observed on an image from\nOctober 2009\nin approximately the same location of the mango tree from ImageC20-C. However, the tree observed is higher than sheet metal fencing line from the construction site. Due to perspective, the young and less dense tree from ImageC20-C is estimated to be approximately on or below this line.\nIn ideal conditions, young mango trees can grow up to\n1.5m/year\n. If ImageC20-C is assumed was produced in 2008, this growth rate is consistent with the approximate difference in height observed between the mango tree in 2009 and ImageC20’s on\nFigure 44\n.\nA blue tent [13] is also observed within the construction site. This element is featured on ImageC20 and has been highlighted in blue.\nFigure 44: A mango tree is observed above the fence line. A tree was identified having leaves of copper colour which is typical for young leaves. The same type of leaves are observed on ImageC20-C. An example is offered as a reference. Within the premises of the hotel construction site, a blue tent [13] is also visible. this tent is featured on ImageC20-C and has been highlighted in blue. Image Source/ Flickr: Ignacio Izquierdo – “Siem Reap Flooded”\nOn\nFigure 45\n,\nan image from November 2011\nfeatures a\nChildSafe network\ncertified tuk-tuk passing by the plot of land where children were abused. The\nChildSafe\nMovement is a global child protection system established and powered by Friends-International. It was designed to train local citizens and tourists on how to detect and report abuse, and to teach children how to seek protection and create awareness campaigns targeting the international community. During the present investigation, other posts from other institutions were found advertising similar efforts around Siem Reap as early as 2005 (see\nFigure 46\n). Although not the objective of this report, it is important to highlight these initiatives as they might be an indicator of the fight against high levels of child abuse seen in this area of Siem Reap during the period in question. Moreover, they might have information of a child abuse network operating in the area around that time.\nThe same air compressor observed in October 2009 was identified on\nFigure 45\nin 2011, indicating that the people in charge of the shack/stall in 2009 might have been still working/living on this site.\nThe galvanised metal fencing was replaced by a solid wall with spikes. Also, a young mango tree is seen in the same relative position as the smaller one observed on ImageC20. The shack observed in October 2009 during the flooding was relocated to approximately the same spot where the blue tent [16] was located on ImageC20. Satellite imagery from 2010 indicates this change was already implemented in March 2010. These observations suggest ImageC20 was more likely produced prior to 2010.\nNovember 2011\nFigure 45: November 2011. A ChildSafe network-certified tuk-tuk passing by the plot of land where children were abused. A similar air compressor observed in October 2009 was identified. A young taller mango tree is seen in the same relative position as the smaller one observed on ImageC20. The shack observed in October 2009 during the flooding was relocated to approximately the same spot where the blue tent [16] was located on ImageC20. Satellite imagery from 2010 indicates this change was already implemented in March 2010. The galvanised metal fencing was replaced by a solid wall with spikes\nFigure 46: Another example of other initiatives against child abuse seen in Siem reap back in 2005\nTable 2: Approximate timeline of the construction of the hotel building and observations on the adjacent property. The information is used to establish a likely period of time within which Imagec20 could have been produced\nUsing satellite and street view imagery, simplified models of the hotel building and nearby houses were prepared. Orientation and relative sizes of buildings were considered. Based on solar data for the site’s coordinates and neglecting atmospheric and surface factors, an approximation of the shadows cast by the hotel building on adjacent houses was computed.\nAlthough the shape of the resulting shadows might not be exactly the same due to the simplification of the building features and atmospheric factors, the analysis is aimed at providing an intuitive approach that could tell when it is more likely for the shadows of the building to reach the nearby building as seen on ImageC20.\nThe results are shown in\nFigure 47\n. An approximate date was estimated.\nThere are\ntwo options possible: early May or early August at approximately 3pm.\nFigure 47: A simplified model of the hotel building helped to calculate an approximate date for ImageC20\nConclusion – Part II ImageC20/Siem Reap, Cambodia\nThe title of the image posted by Europol helped to focus the search on Asia. After editing ImageC20-C, more features of an electricity post seen in the background of the image were revealed. This information allowed us to narrow down the search to the main roads of Siem Reap in Cambodia, where exactly the same posts were found.\nAll the trees in the background of the image had long straight trunks and dark leaves. This type of trees were similar to those observed along the Siem Reap River road. The posts featured on ImaceC20 was next to a tree fork with the characteristics described above. A preliminary sketch of the site was prepared to assist in the search of this tree along the river road.\nI\nmageC20 was geolocated in the Wat Bo village of Siem Reap on coordinates\n13°21’24.90″N 103°51’31.60″E\nnext to a hotel formerly known as the Angkor Sayana Hotel and within the perimeters of an empty plot of land.\nAlthough no specific images were found showing a clear angle of the blue tent [16] observed on ImageC20-C, the information gathered allowed to establish a timeline for the construction of the hotel to identify its progress year by year until the features observed on ImageC20 were all identified.\nA red parasol largely matching the one featuring on ImageC20 was also identified from an image dated October 2008\n.\nThe mango tree featured on ImageC20-C was identified as a taller tree on an image from 2009. Assuming the tree followed a growth rate lower than the maximum theoretical value of 1.5m/year, It is likely the ImageC20-C was taken in 2008 in order to account for the difference in height observed in the trees until 2009.\nAfter modelling the shadows the hotel building casts on nearby houses, the results were compared against the original image to estimate the time of the year when the images were produced.\nBased on the evidence discussed two scenarios are possible:\n1)\nImageC20 was\nlikely\nproduced between 2007 and 2009; either in early May or August\n2)\nImageC20 was\nmost likely\nproduced in August 2008 at around 3 pm.\nThere is evidence the town of Siem Reap was host of many initiatives against child abuse during the analysed period.\nBased on the information discussed, there seemed to be a continuous presence of people living or working within the property adjacent to the hotel between 2008 and 2011. The tenants of the shack and stall might have witnessed some suspicious activity on the property. Also, as the operation of the stall remained ongoing until 2012, the construction company, the staff working for the hotel, as well as tuk-tuk drivers regularly stopping by may still be able to provide relevant information about some of the activities that have taken place on the property.\nAccording to experts consulted, the images seem to have been taken with a phone camera. For the period of time in question, this type of technology was just emerging. Perhaps it might be possible for the authorities and experts to obtain data or other information that helps to identify the phone to then trace it back to the owner who used it in Cambodia on the dates and time proposed.\nResearch by : Carlos Gonzales, Daniel Romein, Timmi Allen and “Bo”",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/21.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/22.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/23.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/24.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/25.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/26.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/27.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/28.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/29-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/30-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/31-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/32-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/33-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/34-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/35-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/36-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/37-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/38-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/39-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/40-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/41-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/42-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/43.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/44.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/45-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/46-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/Table-2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2019/12/47-1.png",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2020-01-09_video-apparently-showing-flight-ps572-missile-strike-geolocated-to-iranian-suburb.json b/bellingcat_geolocation_json/2020-01-09_video-apparently-showing-flight-ps572-missile-strike-geolocated-to-iranian-suburb.json
new file mode 100644
index 0000000000000000000000000000000000000000..187d988e88839ed47ad38aedac155300c9435b89
--- /dev/null
+++ b/bellingcat_geolocation_json/2020-01-09_video-apparently-showing-flight-ps572-missile-strike-geolocated-to-iranian-suburb.json
@@ -0,0 +1,42 @@
+{
+ "slug": "video-apparently-showing-flight-ps572-missile-strike-geolocated-to-iranian-suburb",
+ "title": "Video Apparently Showing Flight PS752 Missile Strike Geolocated to Iranian Suburb",
+ "url": "https://www.bellingcat.com/news/middle-east/2020/01/09/video-apparently-showing-flight-ps572-missile-strike-geolocated-to-iranian-suburb/",
+ "date": "2020-01-09T21:05:51+00:00",
+ "modified": "2020-07-16T16:18:05+00:00",
+ "author": "Bellingcat Investigation Team",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)Русский (Россия)On January 9th, a video spread online after it was posted onto a public Telegram channel showing what was apparently a mid-air explosion. The New York Times has contacted the person who filmed the video, received it in high resolution, and confirmed its authenticity. Below, an annotated version of the video created by […]",
+ "content_html": "
On January 9th, a video spread online after it was posted onto a public Telegram channel showing what was apparently a mid-air explosion. The New York Times has contacted the person who filmed the video, received it in high resolution, and confirmed its authenticity. Below, an annotated version of the video created by Jake Godin (Newsy) highlights the events of the brief clip. (Uploaded directly here)
\n\n
We have geolocated this video to a residential area in Parand (coordinates 35.489414, 50.906917), a suburb to the west of Imam Khomeini International Airport, from which Ukraine International Airlines Flight 752 (PS752) departed to Kyiv.
\n
\n
The camera here is facing northeast, towards the flight trajectory of PS752. A number of elements in the source video can be found in satellite imagery, including a series of apartment blocks, a number of light fixtures, and a construction area to the immediate left of the camera position. In new satellite imagery posted on Terraserver dated 15 November 2019 (not included in this article due to copyright issues), the buildings visible in the video are constructed and in the same position.
\n
\n
By measuring the time that it took for the camera to hear the explosion (~10.7 seconds), we estimated the straight-line distance of the event from the camera at approximately 3.6 kilometers. By using the Pythagoras theorem, we calculated the land distance from the camera to the event to be approximately 3.3 kilometers. We then cross-referenced this distance with our tentative geolocation of the video, and also plotted the flight trajectory of PS752 (taken from from FlightRadar24, found here). The land location of the event aligned with the trajectory of the plane as extended based on FlightRadar24’s data.
\n
[Note: We made a slight adjustment to the vertical side of the triangle, resulting in an extension of the radius by 500 meters (2.8 to 3.3 kilometers). The graphic below is adjusted accordingly. Thank you to ourreaders who helped notice a necessary change in the elevation.]
\n
\n
\n\n
\n
\n
\n
It is unclear why the person holding the camera was filming at the time, but it is possible that there were two missiles, prompting the decision to start filming for the second strike. The New York Times reported that the person filming started doing so after hearing “some sort of shot fired“.
\n
Unresolved questions around missile fragment
\n
While the footage showing an apparent missile strike has been geolocated, two photographs apparently showing part of a Tor M-1 missile have yet to be verified, despite claims from a number of sources. The warhead is located midway on the missile, meaning that its nose may not be destroyed in an explosion. A number of similar photographs of Tor missile fragments have been taken in eastern Ukraine, but none have been discovered to be the same as the ones attributed to the recent incident.
\n
\n
\n
The origin of these photographs is still being determined, as the people who snapped the images have not come forward publicly. This object is likely located in a residential area near Parand, not near the crash site. Both of the images show the same location and object, as seen with many of the same rocks and damage patterns on the curb.
\n
If you make any progress in geolocating these two photographs of the missile fragment, please leave your ideas in the comments, or tweet at us.
\n
\n
Update: New photographs and videos from this location in Parand confirm the geolocation.
\n
\n
Here is a photograph from the exact location that we geolocated last night. Photo taken today. Angle is only about 5-6 m to the left. pic.twitter.com/HZbnx9FijD
And here's a video that shows pretty much a panoramic view from the perspective of the videographer of the #PS572 missile impact video pic.twitter.com/Koagl31LQg
Edit: A previous version of this story had a typo in the title for “PS572” instead of “PS752”. This has been fixed, but the typo persists in the URL.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nOn January 9th, a video spread online after it was posted onto a\npublic Telegram channel\nshowing what was apparently a mid-air explosion. The New York Times has contacted the person who filmed the video,\nreceived it in high resolution\n, and confirmed its authenticity. Below, an annotated version of the video created by\nJake Godin\n(Newsy) highlights the events of the brief clip. (Uploaded directly\nhere\n)\nWe have geolocated this video to a residential area in Parand (coordinates\n35.489414, 50.906917\n), a suburb to the west of Imam Khomeini International Airport, from which Ukraine International Airlines Flight 752 (PS752) departed to Kyiv.\nThe camera here is facing northeast, towards the flight trajectory of PS752. A number of elements in the source video can be found in satellite imagery, including a series of apartment blocks, a number of light fixtures, and a construction area to the immediate left of the camera position. In new satellite imagery posted on Terraserver dated 15 November 2019 (not included in this article due to copyright issues), the buildings visible in the video are constructed and in the same position.\nBy measuring the time that it took for the camera to hear the explosion (~10.7 seconds), we estimated the straight-line distance of the event from the camera at approximately 3.6 kilometers. By using the Pythagoras theorem, we calculated the land distance from the camera to the event to be approximately 3.3 kilometers. We then cross-referenced this distance with our tentative geolocation of the video, and also plotted the flight trajectory of PS752 (taken from from FlightRadar24, found\nhere\n). The land location of the event aligned with the trajectory of the plane as extended based on FlightRadar24’s data.\n[\nNote: We made a slight adjustment to the vertical side of the triangle, resulting in an extension of the radius by 500 meters (2.8 to 3.3 kilometers). The graphic below is adjusted accordingly. Thank you to\nour\nreaders\nwho helped notice a necessary change in the elevation.]\nIt is unclear why the person holding the camera was filming at the time, but it is possible that there were\ntwo missiles\n, prompting the decision to start filming for the second strike. The New York Times reported that the person filming started doing so after hearing “\nsome sort of shot fired\n“.\nUnresolved questions around missile fragment\nWhile the footage showing an apparent missile strike has been geolocated, two photographs apparently showing part of a Tor M-1 missile have yet to be verified, despite\nclaims from a number of sources\n. The warhead is located\nmidway on the missile,\nmeaning that its nose may not be destroyed in an explosion. A number of\nsimilar photographs\nof Tor missile fragments have been taken in eastern Ukraine, but none have been discovered to be the same as the ones attributed to the recent incident.\nThe origin of these photographs is still being determined, as the people who snapped the images have not come forward publicly. This object is likely located in a residential area near Parand, not near the crash site. Both of the images show the same location and object, as seen with many of the same rocks and damage patterns on the curb.\nIf you make any progress in geolocating these two photographs of the missile fragment, please leave your ideas in the comments, or\ntweet at us\n.\nUpdate\n: New photographs and videos from this location in Parand confirm the geolocation.\nHere is a photograph from the exact location that we geolocated last night. Photo taken today. Angle is only about 5-6 m to the left.\npic.twitter.com/HZbnx9FijD\n— Christo Grozev (@christogrozev)\nJanuary 10, 2020\nAnd here's a video that shows pretty much a panoramic view from the perspective of the videographer of the\n#PS572\nmissile impact video\npic.twitter.com/Koagl31LQg\n— Christo Grozev (@christogrozev)\nJanuary 10, 2020\nEdit: A previous version of this story had a typo in the title for “PS572” instead of “PS752”. This has been fixed, but the typo persists in the URL.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/2020-01-09_13-20-59.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/image-38.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/image-11.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/ENxkCymUUAAPtXB.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/EN2FJHnW4AAYrAZ.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2020-01-21_geolocating-venezuelan-lawmakers-in-europe.json b/bellingcat_geolocation_json/2020-01-21_geolocating-venezuelan-lawmakers-in-europe.json
new file mode 100644
index 0000000000000000000000000000000000000000..c3ff5ee5158a12b64892661566b7706569076ffc
--- /dev/null
+++ b/bellingcat_geolocation_json/2020-01-21_geolocating-venezuelan-lawmakers-in-europe.json
@@ -0,0 +1,87 @@
+{
+ "slug": "geolocating-venezuelan-lawmakers-in-europe",
+ "title": "Geolocating Venezuelan Lawmakers In Europe",
+ "url": "https://www.bellingcat.com/news/2020/01/21/geolocating-venezuelan-lawmakers-in-europe/",
+ "date": "2020-01-21T19:12:03+00:00",
+ "modified": "2020-01-21T19:13:58+00:00",
+ "author": "Giancarlo Fiorella",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Translations:English (UK)EspañolOne of the newest chapters in the Venezuelan political saga involves a group of National Assembly lawmakers who were named in a report by the investigative news website Armando.Info alongside allegations that said lawmakers were involved in a corruption network. According to an investigation published by the website in December 2019, this group of […]",
+ "content_html": "
One of the newest chapters in the Venezuelan political saga involves a group of National Assembly lawmakers who were named in a report by the investigative news website Armando.Info alongside allegations that said lawmakers were involved in a corruption network. According to an investigation published by the website in December 2019, this group of lawmakers lobbied international institutions like the U.S. Department of the Treasury and the Colombian Office of the Attorney General in order to seek “indulgences” for individuals connected to the government of President Nicolas Maduro and who were themselves accused of corruption.
\n
On January 19, 2020, Armando.Info published its newest findings on this case. The latest investigation revealed how these lawmakers traveled through Europe in the spring of 2019 in order to lobby on behalf of a group of individuals headed by Alex Saab, a central figure in a massive corruption case involving the Venezuelan government’s subsidized food distribution network.
\n
While the legislators had admitted to traveling to Spain and Italy last year, the investigation by Armando.Info showed that they also secretly traveled to Portugal, Bulgaria and Liechtenstein, allegedly on errands on behalf of Saab and his companies. Part of the evidence of the lawmakers’ trip came from the geolocation of two images to Bulgaria and Switzerland. These images had been publiclyavailable for months, but they had not been geolocated. When Armando.Info contacted Bellingcat for assistance with this task, their investigation into the case had already yielded some clues about where the images might have been captured.
\n
The first of these images shows deputies Conrado Pérez and Richard Arteaga exiting a vehicle in an urban centre somewhere. The image was rumoured to have been taken somewhere in Bulgaria in front of a bank:
\n
Deputies Conrado Pérez and Richard Arteaga (Source: Armando.Info)
\n
The second image shows six deputies having their picture taken together alongside a journalist. As with the previous image, there were rumours that the image had been taken somewhere in Switzerland while the deputies were on their way to Liechtenstein:
\n
Journalist Carlos Herrera and deputies Luis Parra, Conrado Pérez, José Brito, Adolfo Superlano, José Luis Pirela, and Chaim Bucarán (Source: El Pitazo)
\n
Geolocating each of these two images would add detail to the investigation being conducted by Armando.Info, as they would serve as evidence of the deputies’ movements in Europe.
\n
This article shows the process, step-by-step, that resulted in the geolocation of these two images: the first to a street corner in front of a bank in Sofia, Bulgaria, and the second to a highway rest stop on the shores of Lake Walen in Switzerland.
\n
The Bank In Sofia
\n
The first image shows deputies Conrado Pérez and Richard Arteaga exiting a vehicle on a street, allegedly somewhere in Bulgaria in front of a bank. With this information, we can begin to geolocate the image. If the building behind the deputies is really a bank, then we should be able to find its logo — roughly two black sticks on a yellow background — somewhere online.
\n
First, we conduct a Google search for “list of banks in Bulgaria”. The search provides many results, including this list that includes the names of several banks that operate in Bulgaria. By searching the names of each bank on Google Image Search to see their respective logos, we eventually reach “Raiffeisen Bank”, which yields the following results:
\n
The results of the Google Images search for “Raiffeisen Bank” (Source: Google)
\n
The Raiffeisen Bank logo appears to be a good candidate for the one that we can see behind the deputies in the picture.
\n
If the deputies were in Bulgaria, then it was likely that they would have gone to Sofia, the country’s capital. Starting on this assumption, we can search for “Raiffeisen Bank” on Google Maps in Sofia to determine if 1) there are Raiffeisen Bank branches in Sofia, and 2) if there are, whether the one in the picture of the deputies is among them.
\n
Google Maps gives us 20 results for “Raiffeisen Bank” in Sofia:
\n
The 20 results for the search “Raiffeisen Bank” in Sofia, Bulgaria (Source: Google)
\n
From this point, the next step would be to check each of these locations on the map on Google Street View to see if any one of them is the one that is in the background of the image of the deputies. There is a way to make this task a little easier and faster than it might be ordinarily.
\n
We can draw a sketch of what the scene behind the lawmakers might look like from above: that is to say, how might it look like on Google Maps? This sketch would have to include big details, like road layouts and buildings, and should omit small and/or mobile details like the positions of vehicles and people since the exact same ones seen in the picture are unlikely to appear in a satellite image.
\n
Below, my sketch of what the scene in the picture of the deputies might look like on Google Maps:
\n
\n
The “X” marks the spot where the person who took the picture would have been standing to take the picture of the deputies. The spotted lines include the presence of roads which, while not directly visible in the image, can be inferred
\n
Note that the bank’s facade is the only one between two streets, and that the street that runs next to the bank on the right side of the image appears to curve into the background given the position of the vehicles on it.
\n
With this sketch, we can quickly check each of the bank locations on Google Maps to see if any of them could be a candidate. Below is a compilation of three possibilities. Which one most closely resembles the sketch?
\n
Three of the 20 results for the bank in Sofia given by Google Maps. By comparing our sketch to these options, the first one on the left should call our attention, given the narrow facade of the branch and the curved street on one of its sides (Source: Google)
\n
Without having to check each branch one-by-one on Google Street View, we can eliminate the ones that are obviously different from our sketch. Of the three possibilities above, the first option is a good candidate since it shows a bank between two streets with a street that curves along one side of it.
\n
If we go to the Google Street View image of that location, we can see that it is the same one that is in the image of the deputies:
\n
Google Street View image of the location as it appeared on April 2012 (Source: Google)
\n
There is also a Google Street View image from this location dated June 2015, which was used to do the comparison below:
\n
Note the bank’s facade (red box), the position of the traffic lights (purple and blue box), and the arcs on the building in the background (green box) (Source: Google)
\n
There are also some small but notable differences between the two images above. The letters that spell out the bank’s name in the picture of the deputies take up less space than on the Google Street View imagery, and there is what appears to be a yellow awning to the right of the door of the bank. Sometimes, it is necessary to go to other sources for more recent reference images from a given location.
\n
Mapillary is a website that allows us to access dashcam images taken by users around the world. If we go to this same street corner in Sofia on Mapillary, we can see that a user drove by this exact spot on January 26 2019 and captured the features in the bank’s storefront that we see in the picture of the deputies, as well as the trash can and the tree that are visible in the foreground of the image:
\n
In these two images from Mapillary, we can see that the letters that spell out the name of the bank (red box) take up the same space as on the picture of the deputies, and that the yellow awning is also visible (green box). We can also see the garbage can (yellow box) and the tree (purple box) in the foreground of the picture, as well as the place where the person who took the picture was standing (black circle) (Source: Mapillary)
\n
Having completed this geolocation, we can determine that these deputies were standing on this spot in Sofia, Bulgaria, when they had their picture taken.
\n
The Rest Stop By Lake Walen
\n
The second picture shows deputies, Parra, Pérez, Brito, Superlano, Pirela, and Bucarán somewhere picturesque alongside a journalist named Carlos Herrera. In the background we can see a body of water, like a river or a lake, along with a mountain. The information that accompanied the image suggested that it had been taken in the spring of 2019 in Switzerland, somewhere along a road that the deputies took to drive to Liechtenstein.
\n
In order to geolocate this image, we begin with the allegation that it was taken somewhere along a road that connects Switzerland and Liechtenstein. Google Maps shows us that there are in fact several major roads that connect these two countries:
\n
A map of the area border between Switzerland (on the left of the image) and Liechtenstein, which is highlighted in red (Source: Google)
\n
Switzerland is home to many lakes, and we can begin to search those that are near main roads and that are flanked on at least one side by a mountain. Using Google Maps, we can navigate to the lakes in this area to see if any of them has these characteristics. Of the three options below, one stands out as a possible location:
\n
There are three lakes in Switzerland that are near Liechtenstein that we can quickly check to see if they have the characteristics that we see in the picture. Above: Lakes Zurich (left), Constance (middle), and Walen (right) (Source: Google/Airbus, Maxar Technologies)
\n
The lake on the right, Walen, is flanked by mountains. This makes it a good place to conduct a closer search to see if the deputies took their picture on its shores.
\n
Although the resolution of the deputies’ picture is not very high, it still contains important details that we can use for the geolocation. In particular, there is a rock formation on the face of the mountain that could act as a unique identifier:
\n
While the resolution of the image is low, this rock stratification (red box) could be visible on satellite imagery, which makes it a possible point of reference. Note the horizontal layering of the rocks
\n
With this rock formation as reference, we can use Google Earth Pro to navigate to the lake and glide across the top of the road that runs across its southern shore to see if we can spot the formation. This is what that process looks like, resulting in a point of interest that we can check with other sources:
\n
\n
\n
The rock formation on the face of the mountain that is visible at the end of the clip above is located here. As this rock formation called our attention, the next step is to look for reference images that show that same spot on the mountain. By going to the opposite shore of the lake on Google Maps, we can see that there is a logical spot where the image might have been taken: by a rest stop on the highway. It is not difficult to imagine that, on their journey on this highway, many people have stopped at this rest stop, used the washroom, stretched their legs, and taken a picture of the vista.
\n
Some of those people have shared their images from this spot on Google Maps. Thanks to them, we can see what this location looks like from their point of view. The image below was taken at that location, showing the mountain, the lake, and our rock formation of interest in the background:
\n
This picture shows the view from the rest stop on the highway (Source: Google)
\n
There are severalpictures of the same location which contain elements that are visible in the picture of the deputies: not only the mountain and the lake behind them, but also the street, the guardrail, the tree to the right of the men, the grass and trees behind them, and the cabins in the distance. There are also more features on the mountain behind them that we can match to reference images:
\n
A comparison of the picture of the deputies and an image. Note in particular what appears to be a sewer/drainage cover (green box), marking the spot where the deputies posed to take their picture (Source: Google)
\n
The tree with flowers behind the men is probably a variety of cherry blossom tree, which are famous for flowering in the spring—the season when Armando.Info suspected that the deputies had visited Switzerland and Liechtenstein.
\n
Conclusion
\n
Image geolocation is an important skill for open source investigators, given that it could contribute clues to investigations including confirmation of the movements of individuals. Image geolocation is rarely done “cold”: that is, we often have some clues regarding where an image was taken, even if those clues are rumours at first. In this case, the rumours that swirled around these images were enough to allow for a geolocation of where these lawmakers were photographed during their tour of Europe.
\n
If you would like to practice your geolocation skills, follow Quiztime on Twitter, an account which regularly posts geolocation challenges.
\n
With special thanks to Klement Anders for his help with this article.
\n
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nEspañol\nOne of the newest chapters in the Venezuelan political saga involves a group of National Assembly lawmakers who were named in a report by the investigative news website\nArmando.Info\nalongside allegations that said lawmakers were involved in a corruption network. According to an investigation published by the website in December 2019, this group of lawmakers lobbied international institutions like the U.S. Department of the Treasury and the Colombian Office of the Attorney General in order to seek\n“indulgences”\nfor individuals connected to the government of President Nicolas Maduro and who were themselves accused of corruption.\nOn January 19, 2020,\nArmando.Info\npublished its newest findings on this case. The\nlatest investigation\nrevealed how these lawmakers traveled through Europe in the spring of 2019 in order to lobby on behalf of a group of individuals headed by Alex Saab, a central figure in a massive corruption case involving the Venezuelan government’s subsidized food distribution network.\nWhile the legislators had admitted to traveling to Spain and Italy last year, the investigation by\nArmando.Info\nshowed that they also secretly traveled to Portugal, Bulgaria and Liechtenstein, allegedly on errands on behalf of Saab and his companies.\nPart of the evidence of the lawmakers’ trip came from the geolocation of two images to Bulgaria and Switzerland. These images had been\npublicly\navailable\nfor months, but they had not been geolocated. When Armando.Info contacted Bellingcat for assistance with this task, their investigation into the case had already yielded some clues about where the images might have been captured.\nThe first of these images shows deputies Conrado Pérez and Richard Arteaga exiting a vehicle in an urban centre somewhere. The image was rumoured to have been taken somewhere in Bulgaria in front of a bank:\nDeputies Conrado Pérez and Richard Arteaga (Source: Armando.Info)\nThe\nsecond\nimage shows six deputies having their picture taken together alongside a journalist. As with the previous image, there were rumours that the image had been taken somewhere in Switzerland while the deputies were on their way to Liechtenstein:\nJournalist Carlos Herrera and deputies Luis Parra, Conrado Pérez, José Brito, Adolfo Superlano, José Luis Pirela, and Chaim Bucarán (Source: El Pitazo)\nGeolocating each of these two images would add detail to the investigation being conducted by\nArmando.Info\n, as they would serve as evidence of the deputies’ movements in Europe.\nThis article shows the process, step-by-step, that resulted in the geolocation of these two images: the first to a street corner in front of a bank in Sofia, Bulgaria, and the second to a highway rest stop on the shores of Lake Walen in Switzerland.\nThe Bank In Sofia\nThe first image shows deputies Conrado Pérez and Richard Arteaga exiting a vehicle on a street, allegedly somewhere in Bulgaria in front of a bank. With this information, we can begin to geolocate the image. If the building behind the deputies is really a bank, then we should be able to find its logo — roughly two black sticks on a yellow background — somewhere online.\nFirst, we conduct a Google search for “list of banks in Bulgaria”. The search provides many results, including\nthis list\nthat includes the names of several banks that operate in Bulgaria. By searching the names of each bank on Google Image Search to see their respective logos, we eventually reach\n“Raiffeisen Bank”\n, which yields the following results:\nThe results of the Google Images search for “Raiffeisen Bank” (Source: Google)\nThe Raiffeisen Bank logo appears to be a good candidate for the one that we can see behind the deputies in the picture.\nIf the deputies were in Bulgaria, then it was likely that they would have gone to Sofia, the country’s capital. Starting on this assumption, we can search for “Raiffeisen Bank” on Google Maps in Sofia to determine if 1) there are Raiffeisen Bank branches in Sofia, and 2) if there are, whether the one in the picture of the deputies is among them.\nGoogle Maps gives us 20 results for “Raiffeisen Bank” in Sofia:\nThe 20 results for the search “Raiffeisen Bank” in Sofia, Bulgaria (Source: Google)\nFrom this point, the next step would be to check each of these locations on the map on Google Street View to see if any one of them is the one that is in the background of the image of the deputies. There is a way to make this task a little easier and faster than it might be ordinarily.\nWe can draw a sketch of what the scene behind the lawmakers might look like from above: that is to say, how might it look like on Google Maps? This sketch would have to include big details, like road layouts and buildings, and should omit small and/or mobile details like the positions of vehicles and people since the exact same ones seen in the picture are unlikely to appear in a satellite image.\nBelow, my sketch of what the scene in the picture of the deputies might look like on Google Maps:\nThe “X” marks the spot where the person who took the picture would have been standing to take the picture of the deputies. The spotted lines include the presence of roads which, while not directly visible in the image, can be inferred\nNote that the bank’s facade is the only one between two streets, and that the street that runs next to the bank on the right side of the image appears to curve into the background given the position of the vehicles on it.\nWith this sketch, we can quickly check each of the bank locations on Google Maps to see if any of them could be a candidate. Below is a compilation of three possibilities. Which one most closely resembles the sketch?\nThree of the 20 results for the bank in Sofia given by Google Maps. By comparing our sketch to these options, the first one on the left should call our attention, given the narrow facade of the branch and the curved street on one of its sides (Source: Google)\nWithout having to check each branch one-by-one on Google Street View, we can eliminate the ones that are obviously different from our sketch. Of the three possibilities above, the first option is a good candidate since it shows a bank between two streets with a street that curves along one side of it.\nIf we go to the Google Street View\nimage\nof that\nlocation\n, we can see that it is the same one that is in the image of the deputies:\nGoogle Street View image of the location as it appeared on April 2012 (Source: Google)\nThere is also a Google Street View\nimage\nfrom this location dated June 2015, which was used to do the comparison below:\nNote the bank’s facade (red box), the position of the traffic lights (purple and blue box), and the arcs on the building in the background (green box) (Source: Google)\nThere are also some small but notable differences between the two images above. The letters that spell out the bank’s name in the picture of the deputies take up less space than on the Google Street View imagery, and there is what appears to be a yellow awning to the right of the door of the bank. Sometimes, it is necessary to go to other sources for more recent reference images from a given location.\nMapillary\nis a website that allows us to access dashcam images taken by users around the world. If we go to this same street corner in Sofia on Mapillary, we can see that a user drove by this exact spot on January 26 2019 and\ncaptured\nthe features in the bank’s storefront that we see in the picture of the deputies,\nas well as\nthe trash can and the tree that are visible in the foreground of the image:\nIn these two images from Mapillary, we can see that the letters that spell out the name of the bank (red box) take up the same space as on the picture of the deputies, and that the yellow awning is also visible (green box). We can also see the garbage can (yellow box) and the tree (purple box) in the foreground of the picture, as well as the place where the person who took the picture was standing (black circle) (Source: Mapillary)\nHaving completed this geolocation, we can determine that these deputies were standing on this\nspot\nin Sofia, Bulgaria, when they had their picture taken.\nThe Rest Stop By Lake Walen\nThe second picture shows deputies, Parra, Pérez, Brito, Superlano, Pirela, and Bucarán somewhere picturesque alongside a journalist named Carlos Herrera. In the background we can see a body of water, like a river or a lake, along with a mountain. The information that accompanied the image suggested that it had been taken in the spring of 2019 in Switzerland, somewhere along a road that the deputies took to drive to Liechtenstein.\nIn order to geolocate this image, we begin with the allegation that it was taken somewhere along a road that connects Switzerland and Liechtenstein. Google Maps shows us that there are in fact several major roads that connect these two countries:\nA map of the area border between Switzerland (on the left of the image) and Liechtenstein, which is highlighted in red (Source: Google)\nSwitzerland is home to many lakes, and we can begin to search those that are near main roads and that are flanked on at least one side by a mountain. Using Google Maps, we can navigate to the lakes in this area to see if any of them has these characteristics. Of the three options below, one stands out as a possible location:\nThere are three lakes in Switzerland that are near Liechtenstein that we can quickly check to see if they have the characteristics that we see in the picture. Above: Lakes Zurich (left), Constance (middle), and Walen (right) (Source: Google/Airbus, Maxar Technologies)\nThe lake on the right, Walen, is flanked by mountains. This makes it a good place to conduct a closer search to see if the deputies took their picture on its shores.\nAlthough the resolution of the deputies’ picture is not very high, it still contains important details that we can use for the geolocation. In particular, there is a rock formation on the face of the mountain that could act as a unique identifier:\nWhile the resolution of the image is low, this rock stratification (red box) could be visible on satellite imagery, which makes it a possible point of reference. Note the horizontal layering of the rocks\nWith this rock formation as reference, we can use Google Earth Pro to navigate to the lake and glide across the top of the road that runs across its southern shore to see if we can spot the formation. This is what that process looks like, resulting in a point of interest that we can check with other sources:\nhttps://www.bellingcat.com/app/uploads/2020/01/Mountain-Check-1.mp4\nThe rock formation on the face of the mountain that is visible at the end of the clip above is located\nhere\n. As this rock formation called our attention, the next step is to look for reference images that show that same spot on the mountain. By going to the opposite shore of the lake on Google Maps, we can see that there is a logical spot where the image might have been taken: by a\nrest stop\non the highway. It is not difficult to imagine that, on their journey on this highway, many people have stopped at this rest stop, used the washroom, stretched their legs, and taken a picture of the vista.\nSome of those people have shared their images from this spot on Google Maps. Thanks to them, we can see what this location looks like from their point of view. The\nimage\nbelow was taken at that\nlocation\n, showing the mountain, the lake, and our rock formation of interest in the background:\nThis picture shows the view from the rest stop on the highway (Source: Google)\nThere are\nseveral\npictures\nof the same\nlocation\nwhich contain elements that are visible in the picture of the deputies: not only the mountain and the lake behind them, but also the street, the guardrail, the tree to the right of the men, the grass and trees behind them, and the cabins in the distance. There are also more features on the mountain behind them that we can match to reference images:\nA comparison of the picture of the deputies and an image. Note in particular what appears to be a sewer/drainage cover (green box), marking the spot where the deputies posed to take their picture (Source: Google)\nThe tree with flowers behind the men is probably a variety of\ncherry blossom tree\n, which are famous for flowering in the spring\n—t\nhe season when\nArmando.Info\nsuspected that the deputies had visited Switzerland and Liechtenstein.\nConclusion\nImage geolocation is an important skill for open source investigators, given that it could contribute clues to investigations including confirmation of the movements of individuals. Image geolocation is rarely done “cold”: that is, we often have some clues regarding where an image was taken, even if those clues are rumours at first. In this case, the rumours that swirled around these images\nwere enough to allow for a geolocation of where these lawmakers were photographed during their tour of Europe.\nIf you would like to practice your geolocation skills, follow\nQuiztime\non Twitter, an account which regularly posts geolocation challenges.\nWith special thanks to Klement Anders for his help with this article.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/Men.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/Diputados-Lago.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/1-Bank-Name-Image-Search-1200x345.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/2-Bank-Sofia-Results-1200x586.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/Bank01-1200x900.jpeg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/4.-ThreeOptions-1200x305.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/5-Google-Street-View-Bank-1200x586.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/6-Geolocation-1-1200x1125.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/7.-Mapillary-Bank-839x1200.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/8-Roads-1200x678.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/9-Lake-Comparison-1200x195.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/10-Hombres-1.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/MountainView1-1200x586.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/01/13-Men-Geolocation-830x1200.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2020-12-03_using-the-sun-and-the-shadows-for-geolocation.json b/bellingcat_geolocation_json/2020-12-03_using-the-sun-and-the-shadows-for-geolocation.json
new file mode 100644
index 0000000000000000000000000000000000000000..b9874681118904e815718842fd66eb874ed7ade6
--- /dev/null
+++ b/bellingcat_geolocation_json/2020-12-03_using-the-sun-and-the-shadows-for-geolocation.json
@@ -0,0 +1,47 @@
+{
+ "slug": "using-the-sun-and-the-shadows-for-geolocation",
+ "title": "Using the Sun and the Shadows for Geolocation",
+ "url": "https://www.bellingcat.com/resources/2020/12/03/using-the-sun-and-the-shadows-for-geolocation/",
+ "date": "2020-12-03T15:02:44+00:00",
+ "modified": "2021-07-21T17:53:29+00:00",
+ "author": "Youri van der Weide",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "OSINT researchers are familiar with the SunCalc tool for chronolocation. But did you know it can also be used for geolocation? Here's how.",
+ "content_html": "
\n
Most people who have been verifying photos and videos for some time will already be aware of a tool called SunCalc, which chronolocates a video or image. What is less well known is that the position of shadows and the sun can also be useful for geolocation.
\n
For those not in the know, SunCalc lets users analyse the position of shadows and the sun at any given time and date, at any given location. By choosing a date and zooming in on a location such as, say, Puerta del Sol in Madrid, users can click the drag-the-sun icon at the top of the page to see the position of the sun at a certain time. The screenshot below from SunCalc shows the position of the sun over Puerta del Sol at noon on November 22, 2020. Furthermore, when users select the height of any object, the algorithm indicates the corresponding length of its shadow.
\n
Screenshot from the SunCalc tool, showing the Puerta del Sol in Madrid.
\n
Chronolocation is an art in itself. This tutorial by Twitter user and OSINT expert @Sector035 offers some useful tips. In another example, Bellingcat’s Nick Waters recently used the technique in this investigation into the execution of two Armenian men in Nagorno-Karabakh by the Azerbaijani armed forces.
\n
Chronolocation as geolocation
\n
But there is another, often overlooked, way to use SunCalc. The technique can be highly effective in narrowing down a search area, by excluding false positives and establishing in which direction the camera is pointed in a certain frame.
\n
Let’s take a look at how to use it.
\n
The following video was uploaded by an Instagram user on November 22, 2020.
\n\n
It was uploaded as a story, meaning that it disappeared from this person’s profile after 24 hours (its author has given permission for Bellingcat to publish the footage for educational purposes).
\n
The video is timestamped and clearly displays the sun and shadows. This is the first hint. The second is that the user is a Portuguese speaker.
\n
In other words, it was likely filmed in a Lusophone country — perhaps Brazil, perhaps Portugal, or perhaps a Portuguese-speaking country in Africa. Although the accompanying song (Surrade cama by Xand Avião) is by a Brazilian singer, this video was actually filmed in Portugal.
\n
How was this established?
\n
The video’s timestamp indicates that the video was taken at 4:31 pm (UTC). Moreover, when opening a story on Instagram, the reader can discover when the story was uploaded (e.g. “25 min ago”). Both these timestamps corresponded with the actual time in Portugal, meaning that the video was also uploaded at around 4:31 pm (UTC). However, even without the timestamp the casual observer could see that the footage was shot in the late afternoon.
\n
It is plausible to assume that the video was shot in Lisbon, as it is the biggest metropolitan area of Portugal. The sea is visible; this looks like it was either filmed in a medium-size city or bigger.
\n
All this can be checked with SunCalc. In this case, by opening the page and selecting November 22, 2020 at 4:31 pm. This link opens SunCalc with the exact same variables.
\n
\n
As the image above shows, the sun was about to set at that time and to the South-West of the city. In the video, the sun is above the sea to the right of the person filming. It is also clear that the footage was filmed from an apartment next to a four-lane avenue with a line of palm trees down the middle, a parking lot, and an elongated roundabout that appears to have been recently built:
\n
Panorama image based on screenshots from Instagram video. Fair use
\n
So far, we know that the video was taken at 4:31 pm and that the sun was in a southwest direction at that time. For this reason, we can conclude that the person filming was facing South and that the avenue’s direction runs roughly East to West. We also know that the sea or the bay lies due South of the location. This allowed me to put together this rough sketch:
\n
\n
If the sea or other body of water must be to the South, that excludes several locations — not only further inland, but also any areas of Greater Lisbon which are South of the Tagus River. Furthermore, given that the avenue runs not only East to West but also parallel to the coastline, we can also exclude most places on the Tagus estuary which are North-East of Lisbon.
\n
\n
This narrows down the possible locations to places due West of central Lisbon, along the coast all the way towards Cascais. Importantly, the other side of the Tagus estuary cannot be seen in the footage. This suggests that the video was shot further to the West where the Tagus meets the ocean or along the Atlantic coast.
\n
A search from East to West along the shoreline, looking for wide avenues divided by a line of palm trees near an elongated roundabout, yields only a handful of potential matches. If you are using Google Maps or Google Earth’s pre-selected satellite imagery you might miss the elongated roundabout altogether, but even then it is hard to miss the distinctive Avenida da República in Oeiras — with four lanes and a row of palm trees. There are no other streets like it in the immediate area.
\n
\n
A useful shortcut
\n
As mentioned earlier, this technique is often overlooked. Nevertheless, it is a fairly simple one that can save plenty of time once learnt. It took me just five to ten minutes in total to find the location, starting only with a hunch that this footage might have been taken in Portugal – and more specifically, in Lisbon. Without a general orientation, i.e. where the cardinal directions are in relation to some landmarks (in this case the ocean), researchers can waste a lot of time checking false positives they could otherwise have excluded immediately – in this case, coastal areas South or northeast of Lisbon. \n
\n
I hope this is a useful tip in your research. If anything, maybe this will make you think twice before posting a video or picture of that amazing view from your apartment window.
\n
\n
\n
(Вы можете читать эту статью на русском языке на сайте GIJN)
\n
",
+ "content_text": "Most people who have been verifying photos and videos for some time will already be aware of a tool called SunCalc, which\nchronolocates\na video or image. What is less well known is that the position of shadows and the sun can also be useful for\ngeolocation\n.\nFor those not in the know,\nSunCalc\nlets users analyse the position of shadows and the sun at any given time and date, at any given location. By choosing a date and zooming in on a location such as, say, Puerta del Sol in Madrid, users can click the drag-the-sun icon at the top of the page to see the position of the sun at a certain time. The screenshot below from SunCalc shows the position of the sun over Puerta del Sol at noon on November 22, 2020. Furthermore, when users select the height of any object, the algorithm indicates the corresponding length of its shadow.\nScreenshot from the SunCalc tool, showing the Puerta del Sol in Madrid.\nChronolocation is an art in itself. This\ntutorial\nby Twitter user and OSINT expert\n@Sector035\noffers some useful tips. In another example, Bellingcat’s Nick Waters recently used the technique in this\ninvestigation\ninto the execution of two Armenian men in Nagorno-Karabakh by the Azerbaijani armed forces.\nChronolocation as geolocation\nBut there is another, often overlooked, way to use SunCalc. The technique can be highly effective in narrowing down a search area, by excluding false positives and establishing in which direction the camera is pointed in a certain frame.\nLet’s take a look at how to use it.\nThe\nfollowing video\nwas uploaded by an Instagram user on November 22, 2020.\nIt was uploaded as a story, meaning that it disappeared from this person’s profile after 24 hours (its author has given permission for Bellingcat to publish the footage for educational purposes).\nThe video is timestamped and clearly displays the sun and shadows. This is the first hint. The second is that the user is a Portuguese speaker.\nIn other words, it was likely filmed in a Lusophone country — perhaps Brazil, perhaps Portugal, or perhaps a Portuguese-speaking country in Africa. Although the accompanying song (\nSurra\nde cama\nby Xand Avião) is by a Brazilian singer, this video was actually filmed in Portugal.\nHow was this established?\nThe video’s timestamp indicates that the video was taken at 4:31 pm (UTC). Moreover, when opening a story on Instagram, the reader can discover when the story was uploaded (e.g. “25 min ago”). Both these timestamps corresponded with the actual time in Portugal, meaning that the video was also uploaded at around 4:31 pm (UTC). However, even without the timestamp the casual observer could see that the footage was shot in the late afternoon.\nIt is plausible to assume that the video was shot in Lisbon, as it is the biggest metropolitan area of Portugal. The sea is visible; this looks like it was either filmed in a medium-size city or bigger.\nAll this can be checked with SunCalc. In this case, by opening the page and selecting November 22, 2020 at 4:31 pm. This\nlink\nopens SunCalc with the exact same variables.\nAs the image above shows, the sun was about to set at that time and to the South-West of the city. In the video, the sun is above the sea to the right of the person filming. It is also clear that the footage was filmed from an apartment next to a four-lane avenue with a line of palm trees down the middle, a parking lot, and an elongated roundabout that appears to have been recently built:\nPanorama image based on screenshots from Instagram video. Fair use\nSo far, we know that the video was taken at 4:31 pm and that the sun was in a southwest direction at that time. For this reason, we can conclude that the person filming was facing South and that the avenue’s direction runs roughly East to West. We also know that the sea or the bay lies due South of the location. This allowed me to put together this rough sketch:\nIf the sea or other body of water must be to the South, that excludes several locations — not only further inland, but also any areas of Greater Lisbon which are South of the Tagus River. Furthermore, given that the avenue runs not only East to West but also parallel to the coastline, we can also exclude most places on the Tagus estuary which are North-East of Lisbon.\nThis narrows down the possible locations to places due West of central Lisbon, along the coast all the way towards Cascais. Importantly, the other side of the Tagus estuary cannot be seen in the footage. This suggests that the video was shot further to the West where the Tagus meets the ocean or along the Atlantic coast.\nA search from East to West along the shoreline, looking for wide avenues divided by a line of palm trees near an elongated roundabout, yields only a handful of potential matches. If you are using Google Maps or Google Earth’s pre-selected satellite imagery you might miss the elongated roundabout altogether, but even then it is hard to miss the distinctive Avenida da República in Oeiras — with four lanes and a row of palm trees. There are no other streets like it in the immediate area.\nA useful shortcut\nAs mentioned earlier, this technique is often overlooked. Nevertheless, it is a fairly simple one that can save plenty of time once learnt. It took me just five to ten minutes in total to find the location, starting only with a hunch that this footage might have been taken in Portugal – and more specifically, in Lisbon. Without a general orientation, i.e. where the cardinal directions are in relation to some landmarks (in this case the ocean), researchers can waste a lot of time checking false positives they could otherwise have excluded immediately – in this case, coastal areas South or northeast of Lisbon.\nI hope this is a useful tip in your research. If anything, maybe this will make you think twice before posting a video or picture of that amazing view from your apartment window.\n(Вы можете читать эту статью на русском языке\nна сайте GIJN\n)",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/12/suncalc1-1200x729.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/11/image7-1-1200x865.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/11/image10.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/11/image6-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/11/image8.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2020/11/image1.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2021-04-30_john-doe-29-image-from-fbi-child-exploitation-case-geolocated-in-turkey.json b/bellingcat_geolocation_json/2021-04-30_john-doe-29-image-from-fbi-child-exploitation-case-geolocated-in-turkey.json
new file mode 100644
index 0000000000000000000000000000000000000000..d58a8f9968a84cfa73ee6ceffae09b3b60f305e0
--- /dev/null
+++ b/bellingcat_geolocation_json/2021-04-30_john-doe-29-image-from-fbi-child-exploitation-case-geolocated-in-turkey.json
@@ -0,0 +1,112 @@
+{
+ "slug": "john-doe-29-image-from-fbi-child-exploitation-case-geolocated-in-turkey",
+ "title": "John Doe 29: Image From FBI Child Exploitation Case Geolocated to Turkey",
+ "url": "https://www.bellingcat.com/news/2021/04/30/john-doe-29-image-from-fbi-child-exploitation-case-geolocated-in-turkey/",
+ "date": "2021-04-30T11:40:46+00:00",
+ "modified": "2022-02-16T18:09:03+00:00",
+ "author": "Carlos Gonzales",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Two images posted by law enforcement officials looking for leads in a child exploitation case have been geolocated to a tourist town in Turkey.",
+ "content_html": "
\n
The following report contains reference to an active child sexual exploitation case. None of the people or commercial establishments, directly or indirectly referenced in this article, are suspected of being involved in the abuse of children. For the avoidance of doubt, it must be noted that all the information used for this investigation was publicly available. Bellingcat has decided to publish some of the findings to create awareness of the case and to show how open source techniques can help in instances such as these.
\n
According to the FBI’s Endangered Child Alert Program (ECAP), law enforcement officials are seeking information which will lead to the identification of an unknown individual known asJohn Doe 29.
\n
Images containing John Doe 29 first came to the attention of the National Center for Missing and Exploited Children (NCMEC) in January 2008.
\n
In a YouTubevideo released by the FBI in 2015, they stated clearly that John Doe 29 is NOT accused of committing any abuse but may have critical information pertaining to the identity of a child victim in an ongoing sexual exploitation investigation.
\n
Source: FBI ECAP
\n
The FBI also released the following images where John Doe 29 was seen photographed next to the victim. To reiterate, John Doe 29 is not a suspect but is sought to help identify the victim.
In 2020, a new set of photos linked to the case were published in an FBI poster. Images from this set labelled BB001 and BB002 are shown below.
\n
Source: FBI ECAP
\n
After careful analysis, we were able to geolocate the poolside seating area in one of the images to a hotel near to the town of Side in Turkey (coordinates: 36.810011, 31.346188). Again, and to be clear, no person orcommercial establishments, directly or indirectly referenced in this article, is suspected of being involved in the abuse of children.
Source of top images: Holidaycheck.de Source of bottom image: FBI ECAP
\n
Forensics
\n
We started breaking down images BB001 and BB002 in smaller sections to analyse relevant details featured on them.
\n
We observed a particular shape and transition of green tones in a green leaf behind John Doe 29. Selecting some examples we determined the folded leaf in question was most likely a banana leaf against bright light. Although possible to grow in colder climates,banana plants thrive in tropical-subtropical regions.
We took this information and compared it against a global banana production map as a way to explore in which countries would it be more likely to find this type of tree. Apart from the expected tropical areas, the map includes countries along the Mediterranean coastline.
\n
Although initially too general to carry out specific searches, we kept this information as a reference to help validate what we were to discover next.
A street with double “T” paver blocks is featured in image BB002. In the background, there is also a shop window. We enhanced the image to reveal a small red object seen through the glazing. It was an element that seemed to contain a key piece of information: a Turkish flag.
\n
In order to test the probability for this kind of display to happen, we googled the term “shops in turkey” to find that, very frequently, shops of different types display the national Turkish flag in their windows, walls, products, etc; as shown in the image below.
\n
Source of images left and top: FBI ECAP. Image Bottom centre: REUTERS/Umit Bektas. Image Bottom left: REUTERS/Osman Orsal
\n
Looking at image BB001 in detail, we spotted what seemed to be a metal litter bin reflected in the pool water. This particular design, with a hoop at the top and in different colour variants, was found to be very popular in parks and hotels across different towns in Turkey. Some examples of local manufacturers/distributors were found here.
The sun lounger was also verified to belong to the Grosfillex brand. The model series is called “Miami”. The curved feet and arm support as well as the seven stripes on the backrest match the loungers featured in the poolside seating image BB001.
\n
Source top left and right images: FBI ECAP. Image centre and bottom right: Tabularasa
\n
We found a distributor for this brand of loungers in Turkey. After checking around 280 references listed in theirwebsite, we couldn’t find a matching pool setting. However this search did help us to understand the type of facility we were looking for.
\n
In an attempt to determine the layout and size of the pool area, we roughly recreated the scene observed in BB001 by constructing a 3D model with SketchUp, a computer program that is more generally used by interior designers, engineers, architects and in video game design. In this case, however, it helped us further visualise and gather more detail about the area around where the image was taken.
\n
Source image on right: FBI ECAP. Image on left: A screengrab of a 3D model created in SketchUp
\n
In order to achieve a similar configuration and perspective; we positioned scaled models of typical sun loungers and parasols at different distributions. We set the point of view height at approximately 1.5 meters and at a short distance from a representative small child figure sitting on a raised surface.
\n
The results suggest the victim was most likely seated at a height of approximately one meter. Considering that on average a typical seat or bench is raised just 50 centimetres from the floor and given the proximity to the pool; we concluded the surface in question was most likely a round pool bar.
\n
The length of the pool in the direction of the shot was roughly estimated to be above 25m and the distance between the camera and the first row of loungers at least 9m.
\n
Source: A screengrab of a 3D model created in SketchUp
In the small village of Side, on the country’s Mediterranean coast, we found several examples of buildings, shops and streets sharing similarities with features observed in image BB002. In particular: the stone façade, brown window frames, marble curbs and double “T” street paver blocks.
\n
Source of image on right: Google Street View. Images in centre and on right: FBI ECAP
\n
Relatively close to the town of Side, we discovered a hotel facility using the same type of stainless steel litter bins next to a pool and creating the same type of reflection in the water. The hotel in question belongs to a group called Primasol.
Although the location was not a match, this piece of information allowed us to determine the hotel was related to another facility: The Primasol Hane Family Hotel.
\n
Analysing imagery for the early 2000’s, the pool area we found in this new hotel matched all of the elements seen in BB001: a round pool bar, trees, metal bins with top hoop, cream parasols and green Grosfillex chairs.
The pool area distribution was also very close to what we predicted with our 3D model. The victim was indeed seated inside a pool bar when image BB001 was taken. The location was verified.
\n
Source for image on left: Holidaycheck.de Source for image on right: FBI: Graphic taken as screen grab from SketchUp
\n
Having geolocated the pool, other details allowed us to determine when approximately image BB001 was taken via imagery compilation and comparison.
\n
The round bar, the green loungers and the cream parasols were already on site in October 2005. In May 2006, the litter bin at the end of the pool was brown (most likely wood) and was not changed to stainless steel until June 2006.
\n
By May 2007 the round pool bar was not on site anymore and the cream parasols were changed to a green-white stripes pattern. This means BB001 must have been taken before this date.
\n
Source of images dated 2005, 2006 and 2007: holidaycheck.de. Image bottom centre: FBI ECAP. Source of matching chair image: holidaycheck.de
\n
Analysing image BB001, there is a turquoise object on a small side table next to one of the green loungers. Surprisingly, when surveying different hotel pictures, we noticed a turquoise bag on the floor very close to the pool area and matching the object seen in BB001. The date of the picture with the bag was detailed as July 2006.
\n
Source image left: FBI. Image right and centre: holidaycheck.de
\n
After surveying thousands of guests’photos andvideos taken in the hotel, it appeared that German had been the main language spoken by guests. We found examples of Turkish staff engaging in conversations and written messages in German. Entertainment shows were also put on in German.
\n
On Tripadvisor, German appeared to be the main language in which reviews were left prior to 2010. However, there also appeared to be a significant number of reviews left in Turkish too.
\n
A screengrab detailing the languages in which reviews were left on TripAdvisor prior to 2010.
\n
While that is noteworthy, it does not confirm that John Doe 29 or the victim from any of these countries.
\n
Conclusion
\n
Analysing the evidence presented above, we conclude the FBI’s picture BB001 was taken in Turkey during summer 2006 and most likely in July of the same year.
\n
Although the victim now could be anywhere in the world, the hotel appeared to mainly cater for German and local Turkish guests during the early 2000’s. However, this does not offer proof of the potential nationality of John Doe 29 or the victim.
\n
Again, to be clear, the FBI state that the man known as John Doe 29 is not accused of any exploitation or abuse but may have useful information pertaining to this case.
\n
The exact location and name of the hotel has been voluntarily passed to the relevant authorities before the publication of this article.
\n
If you have any information concerning this case, please contact your local law enforcement authorities or alternatively a Crimes Against Children FBI Investigator at the nearest American Embassy or Consulate, or submit a tip online at https://tips.fbi.gov/.
\n
",
+ "content_text": "The following report contains reference to an active child sexual exploitation case. None of the people or commercial establishments, directly or indirectly referenced in this article, are suspected of being involved in the abuse of children. For the avoidance of doubt, it must be noted that all the information used for this investigation was publicly available. Bellingcat has decided to publish some of the findings to create awareness of the case and to show how open source techniques can help in instances such as these.\nAccording to the FBI’s Endangered Child Alert Program (ECAP), law enforcement officials are seeking information which will lead to the identification of an unknown individual known as\nJohn Doe 29\n.\nImages containing John Doe 29 first came to the attention of the National Center for Missing and Exploited Children (\nNCMEC\n) in January 2008.\nIn a YouTube\nvideo\nreleased by the FBI in 2015, they stated clearly that\nJohn Doe 29 is\nNOT\naccused of committing any abuse\nbut may have critical information pertaining to the identity of a child victim in an ongoing sexual exploitation investigation.\nSource: FBI ECAP\nThe FBI also released the following images where John Doe 29 was seen photographed next to the victim. To reiterate, John Doe 29 is not a suspect but is sought to help identify the victim.\nSource:\nFBI ECAP\nIn 2020, a new set of photos linked to the case were published in an FBI\nposter\n. Images from this set labelled BB001 and BB002 are shown below.\nSource: FBI ECAP\nAfter careful analysis, we were able to geolocate the poolside seating area in one of the images to a hotel\nnear to the town of\nSide in Turkey\n(c\noordinates:\n36.810011, 31.346188\n)\n. Again, and to be clear,\nno person or\ncommercial establishments, directly or indirectly referenced in this article, is suspected of being involved in the abuse of children.\nSource Left:\nGoogle Maps\n. Source Right: FBI ECAP\nSource of top images:\nHolidaycheck.de\nSource of bottom image: FBI ECAP\nForensics\nWe started breaking down images BB001 and BB002 in smaller sections to analyse relevant details featured on them.\nWe observed a particular shape and transition of green tones in a green leaf behind John Doe 29. Selecting some examples we determined the folded leaf in question was most likely a\nbanana leaf against bright light\n. Although possible to grow in colder climates,\nbanana plants\nthrive in tropical-subtropical regions.\nImage left: FBI ECAP. Image Right:\nCreative Commons\nWe took this information and compared it against a global banana production map as a way to explore in which countries would it be more likely to find this type of tree. Apart from the expected tropical areas, the map includes countries along the Mediterranean coastline.\nAlthough initially too general to carry out specific searches, we kept this information as a reference to help validate what we were to discover next.\nImage: Screengrab from\nOurWorldInData\nA street with double “T” paver blocks is featured in image BB002. In the background, there is also a shop window. We enhanced the image to reveal a small red object seen through the glazing. It was an element that seemed to contain a key piece of information: a Turkish flag.\nIn order to test the probability for this kind of display to happen, we googled the term “shops in turkey” to find that, very frequently, shops of different types display the national Turkish flag in their windows, walls, products, etc; as shown in the image below.\nSource of images left and top: FBI ECAP. Image Bottom centre: REUTERS/Umit Bektas. Image Bottom left: REUTERS/Osman Orsal\nLooking at image BB001 in detail, we spotted what seemed to be a metal litter bin reflected in the pool water. This particular\ndesign\n, with a hoop at the top and in different colour variants, was found to be very popular in parks and hotels across different towns in Turkey. Some examples of local manufacturers/distributors were found\nhere\n.\nSource of image left: FBI ECAP. Image centre:\nbooking.com\nImage right:\nUzmanKirala\nThe sun lounger was also verified to belong to the\nGrosfillex\nbrand. The model series is called “Miami”. The curved feet and arm support as well as the seven stripes on the backrest match the loungers featured in the poolside seating image BB001.\nSource top left and right images: FBI ECAP. Image centre and bottom right:\nTabularasa\nWe found a distributor for this brand of loungers in Turkey. After checking around 280 references listed in their\nwebsite\n, we couldn’t find a matching pool setting. However this search did help us to understand the type of facility we were looking for.\nIn an attempt to determine the layout and size of the pool area, we roughly recreated the scene observed in BB001 by constructing a 3D model with\nSketchUp\n, a computer program that is more generally used by interior designers, engineers, architects and in video game design. In this case, however, it helped us further visualise and gather more detail about the area around where the image was taken\n.\nSource image on right: FBI ECAP. Image on left: A screengrab of a 3D model created in SketchUp\nIn order to achieve a similar configuration and perspective; we positioned scaled models of typical sun loungers and parasols at different distributions. We set the point of view height at approximately 1.5 meters and at a short distance from a representative small child figure sitting on a raised surface.\nThe results suggest the victim was most likely seated at a height of approximately one meter. Considering that on average a typical seat or bench is raised just 50 centimetres from the floor and given the proximity to the pool; we concluded the surface in question was most likely a round pool bar.\nThe length of the pool in the direction of the shot was roughly estimated to be above 25m and the distance between the camera and the first row of loungers at least 9m.\nSource: A screengrab of a 3D model created in SketchUp\nWe explored the\ntop 10 tourist destinations and surrounding villages\nin Turkey. Starting from Istanbul, we moved north to south along the west coast.\nIn the small village of Side,\non the country’s Mediterranean coast,\nwe found several examples of buildings, shops and streets sharing similarities with features observed in image BB002. In particular: the stone façade, brown window frames, marble curbs and double “T” street paver blocks.\nSource of image on right: Google Street View. Images in centre and on right: FBI ECAP\nRelatively close to the town of Side, we discovered a\nhotel facility\nusing the same type of stainless steel litter bins next to a pool and creating the same type of reflection in the water. The hotel in question belongs to a group called\nPrimasol\n.\nSource image left:\nbooking.com\n. Image right: FBI\nAlthough the location was not a match, this piece of information allowed us to determine the hotel was related to another facility:\nThe Primasol Hane Family Hotel\n.\nAnalysing imagery for the early 2000’s, the pool area we found in this new hotel matched all of the elements seen in BB001: a round pool bar, trees, metal bins with top hoop, cream parasols and green Grosfillex chairs.\nSource image on left FBI. Image on right:\nholidaycheck.de\nThe pool area distribution was also very close to what we predicted with our 3D model. The victim was indeed seated inside a pool bar when image BB001 was taken. The location was verified.\nSource for image on left:\nHolidaycheck.de\nSource for image on right: FBI: Graphic taken as screen grab from SketchUp\nHaving geolocated the pool, other details allowed us to determine when approximately image BB001 was taken via imagery compilation and comparison.\nThe round bar, the green loungers and the cream parasols were already on site in\nOctober 2005\n. In\nMay 2006\n, the litter bin at the end of the pool was brown (most likely wood) and was not changed to stainless steel until\nJune 2006\n.\nBy\nMay 2007\nthe round pool bar was not on site anymore and the cream parasols were changed to a\ngreen-white stripes pattern\n. This means BB001 must have been taken before this date.\nSource of images dated\n2005\n,\n2006\nand\n2007\n: holidaycheck.de. Image bottom centre: FBI ECAP. Source of matching chair image:\nholidaycheck.de\nAnalysing image BB001, there is a turquoise object on a small side table next to one of the green loungers. Surprisingly, when surveying different hotel pictures, we noticed a turquoise bag on the floor very close to the pool area and matching the object seen in BB001. The date of the picture with the bag was detailed as\nJuly 2006\n.\nSource image left: FBI. Image right and centre:\nholidaycheck.de\nAfter surveying thousands of guests’\nphotos\nand\nvideos\ntaken in the hotel,\nit appeared\nthat German had been the main language spoken\nby\nguests. We found examples of Turkish staff engaging in conversations\nand\nwritten messages in German. Entertainment shows were also put on in German.\nOn Tripadvisor, German appeared to be the main language in which reviews were left prior to 2010. However, there also appeared to be a significant number of reviews left in Turkish too.\nA screengrab detailing the languages in which reviews were left on TripAdvisor prior to 2010.\nWhile that is noteworthy, it does not confirm that John Doe 29 or the victim from any of these countries.\nConclusion\nAnalysing the evidence presented above, we conclude the FBI’s picture BB001 was taken in Turkey during summer 2006 and most likely in July of the same year.\nAlthough the victim now could be anywhere in the world, the hotel appeared to mainly cater for German and local Turkish guests during the early 2000’s. However, this does not offer proof of the potential nationality of John Doe 29 or the victim.\nAgain, to be clear, the FBI state that the man known as John Doe 29 is\nnot accused of any exploitation or abuse\nbut may have useful information pertaining to this case.\nThe exact location and name of the hotel has been voluntarily passed to the relevant authorities before the publication of this article.\nIf you have any information concerning this case, please contact your local law enforcement authorities or alternatively a Crimes Against Children FBI Investigator at the nearest American Embassy or Consulate, or submit a tip online at\nhttps://tips.fbi.gov/\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/1.-John-Doe-29-Warning.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/Screenshot-2021-04-30-at-10.19.34.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/3.-FBI-BB001-and-BB002.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/4.-google-maps_2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/5.-Verification-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/6.-Banana-Leaf.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/7.-Banana-map-1200x830.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/8.-The-turkish-flag_2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/9.-Metal-trash-can.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/10.-The-sun-lounger-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/11.-3D-model_2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/12.-3D-model.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/13.-Side-Turkey.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/14.-Pool-reflection.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/15.-Verification-2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/16.-Verification-3_2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/17.-Verification-4.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/Screenshot-2021-04-30-at-10.53.29.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/04/Screenshot-2021-04-30-at-10.55.18.png",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2021-05-10_where-in-the-world-is-q-clues-from-image-metadata.json b/bellingcat_geolocation_json/2021-05-10_where-in-the-world-is-q-clues-from-image-metadata.json
new file mode 100644
index 0000000000000000000000000000000000000000..817ced73d6c4af49141c859baad8226faa5c9a35
--- /dev/null
+++ b/bellingcat_geolocation_json/2021-05-10_where-in-the-world-is-q-clues-from-image-metadata.json
@@ -0,0 +1,117 @@
+{
+ "slug": "where-in-the-world-is-q-clues-from-image-metadata",
+ "title": "Where in the World is Q? Clues from Image Metadata",
+ "url": "https://www.bellingcat.com/news/rest-of-world/2021/05/10/where-in-the-world-is-q-clues-from-image-metadata/",
+ "date": "2021-05-10T18:47:26+00:00",
+ "modified": "2022-09-19T18:06:18+00:00",
+ "author": "Abigail W. Xavier, Robert Amour, The Q Origins Project",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Image metadata could help narrow down the search for Q's real identity. Here’s how.",
+ "content_html": "
The hunt to identify the person or group behind the QAnon conspiracy theory continues. Known only as “Q,” this individual or collective is responsible for the many posts, or “drops,” that appeared on internet imageboards between October 2017 and December 2020.
\n
It has not been an easy search. But a fresh approach has yielded more clues. Image metadata comprises many different kinds of information, some of which can speak to the origin of the image itself and, by extension, who created it. Since many of Q’s posts come with an image attached, examining image metadata can provide clues as to where Q has been — and when.
\n
We examined metadata found on images associated with 27 Q drops. We infer from the metadata that the images are screenshots taken in two time zones: UTC+8 and the Pacific time zone. UTC+8 is a time zone including the Philippines, China, Mongolia, Taiwan, Malaysia, Brunei, and parts of Indonesia, Russia and Australia. Pacific time includes the West coasts of Canada and the USA, Nevada, part of Idaho, and part of the northwest coast of Mexico.
\n
The vast majority of these images, ranging in date from November 7, 2017 to September 23, 2020, appear to originate from the Pacific time zone, while a handful of remaining images from February 2018 and 2019 appear to originate from UTC+8.
\n
Timezone map from Wikimedia Commons, with UTC+8 highlighted in yellow by Robert Amour
\n
If accurate, the UTC+8 cluster of images could suggest that Q is a single individual who lives in the Pacific time zone but has traveled extensively in Asia (with stops at least in Hong Kong, Chongqing, Shanghai, and Bangkok, based on the contents of Q’s photos).
\n\nWhat is QAnon? \n\n
QAnon is a baseless conspiracy theory that flourished on the internet during the Trump era. Members of the online movement followed an anonymous poster, Q, who outlined Trump’s supposed plan to take down a global cabal of child-sacrificing elites. Claiming a high-level security clearance in the Trump administration, Q promised a bloody revenge on enemies like Hillary Clinton, Barack Obama, and George Soros. QAnon has drawn in new adherents during the COVID-19 pandemic, and radicalized members of the movement participated in the storming of the Capitol on January 6, 2021.
\n
Those outside the movement dismissed Q’s declared identity as fantastical, and over time journalists assembled a list of figures who might be behind the account in reality. Vice News has argued that early Q drops were authored by a hacker collective and perhaps followed up by a right-wing operative. The new HBO Documentary, “Q: Into the Storm,” addresses suspicions about the previouslyspeculatedinvolvement of Ron Watkins and his father Jim, who were respectively the manager and owner of the website 8chan (and its successor 8kun), which was the exclusive platform of Q for all but the first month of Q’s existence.
\n\n
The Dataset
\n
We have made our code and results available in a Github repository, and hope other researchers will build on our findings.
The collection of image files includes both images Q themselves posted and images in posts that Q replied to. Using the information in posts.json, we removed all images posted by anonymous users (“anons”) who were not Q. This results in a dataset of 1,087 images.
\n
We then ran the “exiftool” program to extract image metadata from every image and collated the information into a spreadsheet. This CSV spreadsheet file contains one row per image and 275 columns. Not all columns are populated for each row, since individual images do not typically include every possible kind of metadata.
\n
We spot-checked these photos in their original settings, i.e. on 8chan/8kun as well as 4plebs (a popular 4chan archive) to ensure that the image metadata was present in the original and not an artifact of qalerts.net’s collection process. We found that the metadata was indeed present in the original files. We also uploaded a few sample images to the existing 8kun website, and downloaded them again after they were posted. This allowed us to ascertain that the version of the 8kun website now running neither strips nor adds metadata to images stored on it.
\n
Returning to our 1,087 images, we first checked to see which metadata fields were populated with interesting information, and followed up by reverse image-searching those which contained interesting or identifying information. This could have linked Q’s images to other individuals, broadening the scope of possible candidates.
\n
It also weeded out images which Q had downloaded from elsewhere which were then uploaded alongside their drops.
\n
Indeed, some images that looked promising turned out to have been taken from other websites with metadata intact. For instance, the image of a field that appears on drop 3985 was taken from the website of Brownfield News, a media network which covers agriculture in the US Midwest.
\n
Image from Brownfield News, taken from a drop posted by Q.
\n
By examining the metadata on the image taken from the Brownfield News website and matching it to the image downloaded from qalerts, we ascertained that all image metadata was inherited from its origin. That is to say, all of the metadata on the image posted to 8kun was also present on the image that was part of the earlier article. It seems like Q downloaded a copy of the image from the Brownfield News website and uploaded that copy to 8kun without altering any of the metadata.
\n
But what about images from Q drops which appeared nowhere else online?
\n
These seemingly original images proved to be the most interesting subset of all. Several of these images not only could not be matched easily to an online origin (so may be hypothesized to come directly from Q) but also contained enough metadata to analyze.
\n
In particular, we soon zeroed in on a group consisting of 29 images whose UserComment fields all contain the word “Screenshot.” Not only are they marked as screenshots, all of these photos appear to be screenshots taken from within a mobile device’s photo gallery. Black bars can be seen above and below most of the photos, as well as elements such as the horizontal, elliptical bar at the bottom of the photo in drop 3642—clearly a user interface element.
\n
Q drop 3,642
\n
However, the strongest proof that these are screenshots comes from one of the images themselves: drop 4752 contains a screenshot, not merely of a picture Q has selected from within an iPhone’s image gallery and then photographed, but of the gallery itself. This is immediately recognizable to any iPhone user.
\n
Q drop 4,752
\n
Of the 29 photos labeled as screenshots, 27 also contain timestamp information in the field DateCreated. These 27 timestamped screenshots appeared in 29 Q drops (two images each appeared twice).
\n
We calculated the time difference between the timestamp of the image metadata and the timestamp in UTC of the Q drop on the imageboard itself, and used this figure to make inferences about the time zone of Q.
Local time of screenshot + time to post + UTC offset of screenshot maker = UTC time of post
\n
or
\n
Time to post + UTC offset of screenshot maker = UTC time of post − Local time of screenshot
\n
or
\n
Time to post + UTC offset of screenshot maker = calculated time diff
\n
\n\n\n
\n
We know the right side of the equation and need the variables on the left. We assume that Q took a screenshot of a photo and then, minutes rather than hours later, a drop was made containing that image. For instance, the image in drop 125 was created at 15:39 local time (recorded by the device taking the screenshot) on July 7, 2017, and posted on 4chan at 15:43 Pacific time (UTC time 23:43). It’s not certain that only four minutes elapsed between the time the picture was processed and posted (rather than some number of hours plus four minutes), but it seems reasonable to assume.
\n
Furthermore, we know that UTC offsets are constrained; the vast majority of them occur in hourly increments. For instance there is no such thing as a time zone that is one hour and 27 minutes past UTC. Therefore, the assumption that time to post was under an hour fixes the UTC offset for all 27 images.
\n
Infographic by Robert Amour and Abigail W. Xavier
\n
Examining the clusters near UTC−7 and UTC−8, we find that they correlate with changes enforced by Daylight Savings in the Pacific time zone. Pacific time changes from UTC−8 to UTC−7 and back over the course of a year. We believe these visually evident clusters provide strong evidence that images in Q drops are sourced from two time zones, UTC+8 and the Pacific time zone.
\n
Further Evidence for Time to Post
\n
The data indicates that Q either waited for a handful of minutes after taking each screenshot before posting it, or that Q routinely waited for an unknown number of hours plus only a handful of minutes—which would be much more unusual behavior than simply posting right away.
\n
In the case of drop 4752, we can verify that DateCreated is the local device time. Drop 4752 is in our Pacific time cluster, and the image in it includes the iPhone’s clock (which shows 12:07 at the time of the screenshot); our metadata also shows a creation time of 12:07. Q posted it at 12:13 Pacific time.
\n
In general we see the same pattern for all Q drops with an image in the UTC−7 / UTC−8 cluster: for instance, drop 1206 has a screenshot taken at 18:16 local time on April 20, 2018 and posted at what corresponds to 18:18 Pacific time, while drop 1752 includes a screenshot made at 11:09 local time, which Q posted at 11:13 Pacific time.
\n
A similar pattern holds for the UTC+8 cluster. Drop 682 contains a screenshot whose DateCreated metadata indicates that it was created at 8:38 local time. It was posted at 8:40 UTC+8 time. Similarly, drops 683–686 include screenshots containing a DateCreated value of 9:25 and 9:26 on August 2, 2018. They were posted between 9:35:08 and 9:36:59 UTC+8 time, a difference of 10 minutes. Finally, drop 2708 contains two photos that were taken at 18:27 and 18:28 local time, and posted at 18:31 UTC+8 time.
\n
To bolster our confidence that time-to-post was within the hour and the remaining hour difference represents the device time zone’s UTC offset, we can observe the distribution of the time differences between the image’s creation date and the drop timestamp by minute within the hour. If the time to post were an hour or more on average, we’d expect to see a somewhat even distribution of minutes within the hour difference.
\n
Infographic by Robert Amour and Abigail W. Xavier
\n
However, the overwhelming majority of images appear to have been posted within 10 minutes, and none appear to have waited more than half an hour before being posted. This allows us to reasonably establish a pattern of behavior and to conclude that the time to post was within the hour, and that remaining hours of the difference represent the UTC offset of the device’s timezone.
\n
Views from a Hotel Window
\n
We know that Q, or an associate close enough to provide them with these photos, traveled to several locations in Asia — a fact which can be inferred not only from the context of these images, but their content. Several are photos of city skylines, others are close-up views of side streets. They generally do not include obvious tourist sites or any close-up individuals whatsoever — par for the course for Q, whose followers pore over images for arcane clues and metaphors.
\n
For example, in drop 682, posted on February 8, 2018 (by UTC time), Q provided the following image of a modern skyscraper with a distinctive spire. (The black bars above and below the photo have been cropped in all images in this section.)
\n
Q later posted one close-up photo of the awning over the building’s main entrance and another of the car park in front of it. These offered some small clues; cars bear blue license plates and a construction site to the right of the building features several bright red placards featuring Chinese characters. The words “Bank of China” can be seen on white backing on the building to the right.
\n
Q drop 682
\n
All of these details suggest that the tower is in the People’s Republic of China — but where? By cropping the distinctive tip of the building and uploading the image into Yandex’s reverse image search, we soon find a match: the King Tower in Pudong, Shanghai.
\n
It is fair to assume that the photo was taken from a tall building across the New Jinqiao Road to the South-West. According to the popular Chinese mapping website and search engine Baidu, this building is the Shanghai Marriott Hotel Pudong East (上海金桥红枫万豪酒店).
\n
Screenshot from Baidu Maps
\n
Two days later, on 10 February 2018, Q posted a close-up image of a large letter “E” hanging from a ceiling with distinctive alternating wooden slats.
\n
Q drop 707
\n
In the bottom of the image, another item of a similar size and color can be seen hanging from the same ceiling. This leads us to assume that the letter “E” is one of a series in a zoning area — such as those found in airports. A simple Google search of airport arrival hallways in large cities in UTC+8 yields a result. An image from Shutterstock shows the same “E” hanging from the same ceiling at Shanghai’s Pudong airport:
\n
Screenshot of Shutterstock’s website
\n
Q’s next drop contained a photo from central Bangkok — Q may have meant to suggest that members of the “Q Team” had recently flown between Shanghai Pudong airport and the Thai capital.
\n
The image below, posted in drop 708 on February 10, 2018, shows a city skyline. A building with the words “Assumption College” can be seen in the foreground.
\n
Q drop 708
\n
The distinctive “pixelated” skyscraper in the background is the Maha Nakhon tower in the central business district of Bangkok. When completed in 2016, it was the tallest building in the Thai capital.
\n
This photo appears to have been taken from a floor about midway up in a hotel called The Peninsula Bangkok or from a building on the same block with a similar vantage point, such as the River Condominiums or Light House Plaza.
\n
The riverside campus of Assumption College is almost perfectly in line with The Peninsula Bangkok and Maha Nakhon. A user-uploaded photo on Google, taken from The Peninsula in March 2021, shows Assumption College and the Maha Nakhon tower from a similar, though higher, perspective to Q’s.
On March 7, 2018 in drop 869, Q posted an image of a bustling market street in an East Asian country with a reference to Edward Snowden in the accompanying text. In the bottom right-hand corner above a meat stall, and on several banners in the center of the image, Chinese characters can be seen but are not completely legible. Bearing in mind that Snowden first fled to Hong Kong, it’s as good a place to start as any.
\n
Q drop 869
\n
The architecture and commercial profile of the area matches the narrow streets of Mong Kok, a popular shopping district in Hong Kong. A digital trawl through Mong Kok using Baidu’s streetview system shows a match at the crossroads of Nelson Street and Canton Street.
\n
Image from Baidu street view
\n
Importantly, Q’s image shows this intersection from high above, though not from a bird’s eye view. This suggests that the photographer was standing in a tall building to the east. The most obvious candidate is the nearby Cordis Hotel.
\n
Images posted by guests of the hotel and tagged at the location can confirm this match further. In December 2020, one visitor to the Cordis posted a panorama photo of the Mong Kok skyline; the bottom left shows this same intersection in detail.
Nevertheless, it should be noted that the screenshot of Q’s image from the Cordis Hotel was created by a user in the Pacific time zone, from which it can be inferred either that Q did not post all images “live” (and might have selected it from their own old photos), or that Q was posting images they had received from a contact.
\n
On February 14, 2019, Q posted Drop 2706, which included a photo of a shopping street somewhere in a Chinese-speaking country. Unlike the previous two images, which each contain a famous skyscraper, this image was unremarkable. However, as a close-up image it did show several names of businesses such as 朋泉聚，周师兄 (Old Brother Zhou)，中国民生银行 (China Minsheng Bank), as well as 诗丽堂美容SPA (Shilitang Beauty SPA, also known by its English name Silktone SPA). Two of these are business names concentrated in Chongqing, a city in South-West China.
\n
Q drop 2,706
\n
As it offers a highly detailed business registry for China, Baidu Maps is the ideal source to start searching. It also offers an extensive streetview mode which allows users to compare imagery taken in different years. After checking street view imagery for various addresses for Silktone Beauty SPA branches in Chongqing, we have a match on the city’s Beijeng Tiangje Street (北城天街). It shows the same shopfront, the same names of all businesses, the same covered bus shelter, and the same distinctive streetlights:
\n
Image: Baidu Maps
\n
Q’s photo of the entrance was taken from above, from a building across the road to the North-East. Baidu’s map shows a highly likely candidate: the Chongqing Fuli Hyatt Regency Hotel.
\n
Further corroborating this is another image Q posted in the same drop, showing the lights of buildings and a distant cityscape at night. Blurring suggests that this was taken at speed, presumably from an aircraft.
\n
Q drop 2,706
\n
The layout of the roads and buildings matches those to the right of Chongqing Airport’s eastern runway, approaching from the South.
\n
Image: Google Maps
\n
Importantly, drop 2503 from November 2018 includes a photo from the hotel, including a Christmas tree. That same month, in drop 2516, Q congratulated “anons” for successfully geolocating the “last pic” to Chongqing, urging them to “watch the news”.
\n
If these images are, as they appear to be, original, then we know the location of the photographer, and can infer the location of the person who took the screenshot, but not necessarily that of the person who posted the respective drops: it is at least conceivable that whoever took the screenshots might have texted or emailed these photos to a third party, who posted them shortly afterwards.
\n
Nevertheless, the images do teach us a few new things about Q.
\n
We know that Q, or a very close associate, flew to Shanghai Pudong Airport prior to February 8, 2018; that they stayed at The Peninsula hotel in Bangkok prior to or on February 10, 2018; that they stayed at the Cordis Hotel in Hong Kong prior to or on March 7, 2018; that they stayed at the Chongqing Fuli Hyatt prior to or on February 14, 2019 and flew into Chongqing prior to that same date.
\n
We know that Q, or an associate who regularly provides images for Q to post, has a particular taste in hotels and has spent time traveling, but likely not living, in four cities in Thailand and China. The fact that some of the images in China can also be geolocated to the same timezone makes the assumption about Q’s location, or the location of somebody working with Q, more plausible.
\n
One Step Closer?
\n
The overwhelming majority of the images in our data set — over 68 percent — appear to have been created by a user, or users who were in the Pacific time zone. Q’s content appears to have been posted from the Pacific time zone at least as early as November 7, 2017 and at least as late as September 23, 2020; this covers almost the entire period that Q was active.
\n
By contrast, images from UTC+8 exist on only three dates in our data: February 8, 2018, February 10, 2018, and February 14, 2019.
\n
This has several potential implications for Q’s identity.
\n
First, these findings are consistent with the hypothesis that the same person has been posting as Q all along. In that case, Q would live or spend the vast majority of their time in the Pacific time zone, with occasional international travel. The hypothesis of a Pacific Q coheres with other quantitative work on the overall hourly patterns of the Q drops, while the hypothesis of a single Q matches the observedfact that Q had access to the same set of photos in November 2017 and May 2018.
\n
Second, the most popular hypothesis among Q-watchers — that authorship of the drops changed hands between January 4 and January 6 of 2018 — cannot be excluded on the basis of our findings. Under this hypothesis, our evidence requires new Q and original Q both to post from Pacific Time.
\n
That’s not impossible, of course. There are well over 50 million people living in the Pacific time zone; one of them could easily have stepped into the original Q’s place.
\n
We reached out to Jim Watkins, the owner of 8chan/8kun who has lived in the Philippines, to comment on the implications of narrowing down Q’s timezone. In an emailed response crowdsourced via his Gab account, Watkins cast doubt on the methodology of the research and its usefulness, adding that more than a billion people live in the UTC+8 time zone. Watkins has previously denied in comments to Bellingcat that he is Q.
\n
Bellingcat was unable to reach Ron Watkins, who has lived in Japan, for comment. Ron Watkins has also denied that he is Q in the course of the recent HBO documentary series.
\n
It is important to stress our evidence is not by itself sufficient to identify any particular individual as Q or an associate of Q.
\n
However, our evidence should provide a valuable datapoint for other investigators who seek to do so. The hunt for Q goes on — and these findings suggest that it must largely continue in the Pacific time zone.
\n
\n
Gavin Fox (pseudonym) contributed to this research. The authors would like to thank Bellingcat’s Giancarlo Forella and Maxim Edwards for their assistance in geolocating the images from East Asia.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nThe hunt to identify the person or group behind the QAnon conspiracy theory continues. Known only as “Q,” this individual or collective is responsible for the many posts, or “drops,” that appeared on internet imageboards between October 2017 and December 2020.\nIt has not been an easy search. But a fresh approach has yielded more clues. Image metadata comprises many different kinds of information, some of which can speak to the origin of the image itself and, by extension, who created it. Since many of Q’s posts come with an image attached, examining image metadata can provide clues as to where Q has been — and when.\nWe examined metadata found on images associated with 27 Q drops. We infer from the metadata that the images are screenshots taken in two time zones:\nUTC+8\nand the\nPacific time zone\n. UTC+8 is a time zone including the Philippines, China, Mongolia, Taiwan, Malaysia, Brunei, and parts of Indonesia, Russia and Australia. Pacific time includes the West coasts of Canada and the USA, Nevada, part of Idaho, and part of the northwest coast of Mexico.\nThe vast majority of these images, ranging in date from November 7, 2017 to September 23, 2020, appear to originate from the Pacific time zone, while a handful of remaining images from February 2018 and 2019 appear to originate from UTC+8.\nTimezone map from Wikimedia Commons, with UTC+8 highlighted in yellow by Robert Amour\nIf accurate, the UTC+8 cluster of images could suggest that Q is a single individual who lives in the Pacific time zone but has traveled extensively in Asia (with stops at least in Hong Kong, Chongqing, Shanghai, and Bangkok, based on the contents of Q’s photos).\nWhat is QAnon?\nQAnon is a baseless conspiracy theory that flourished on the internet during the Trump era. Members of the online movement followed an anonymous poster, Q, who outlined Trump’s supposed plan to take down a global cabal of child-sacrificing elites. Claiming a high-level security clearance in the Trump administration, Q promised a bloody revenge on enemies like Hillary Clinton, Barack Obama, and George Soros. QAnon has drawn in new adherents during the COVID-19 pandemic, and radicalized members of the movement participated in the storming of the Capitol on January 6, 2021.\nThose outside the movement dismissed Q’s declared identity as fantastical, and over time journalists assembled a list of figures who might be behind the account in reality. Vice News\nhas argued\nthat early Q drops were authored by a hacker collective and perhaps followed up by a right-wing operative. The new HBO Documentary, “Q: Into the Storm,”\naddresses suspicions\nabout the\npreviously\nspeculated\ninvolvement\nof Ron Watkins and his father Jim, who were respectively the manager and owner of the website 8chan (and its successor 8kun), which was the exclusive platform of Q for all but the first month of Q’s existence.\nThe Dataset\nWe have made our code and results\navailable\nin a Github repository, and hope other researchers will build on our findings.\nWe downloaded 1,720 image files from\nhttps://qalerts.net/media/\n. We also downloaded the\nJSON file\nof the text of the Q drops and their basic non-image metadata from\nhttps://qalerts.app/data/json/posts.json\n.\nThe collection of image files includes both images Q themselves posted and images in posts that Q replied to. Using the information in posts.json, we removed all images posted by anonymous users (“anons”) who were not Q. This results in a dataset of 1,087 images.\nWe then ran the\n“exiftool” program\nto extract image metadata from every image and collated the information into a spreadsheet. This CSV spreadsheet file contains one row per image and 275 columns. Not all columns are populated for each row, since individual images do not typically include every possible kind of metadata.\nWe spot-checked these photos in their original settings, i.e. on 8chan/8kun as well as 4plebs (a popular 4chan archive) to ensure that the image metadata was present in the original and not an artifact of qalerts.net’s collection process. We found that the metadata was indeed present in the original files. We also uploaded a few sample images to the existing 8kun website, and downloaded them again after they were posted. This allowed us to ascertain that the version of the 8kun website now running neither strips nor adds metadata to images stored on it.\nReturning to our 1,087 images, we first checked to see which metadata fields were populated with interesting information, and followed up by reverse image-searching those which contained interesting or identifying information. This could have linked Q’s images to other individuals, broadening the scope of possible candidates.\nIt also weeded out images which Q had downloaded from elsewhere which were then uploaded alongside their drops.\nIndeed, some images that looked promising turned out to have been taken from other websites with metadata intact. For instance, the image of a field that appears on drop 3985 was taken from\nthe website of Brownfield News\n, a media network which covers agriculture in the US Midwest.\nImage from Brownfield News, taken from a drop posted by Q.\nBy examining the metadata on the image taken from the Brownfield News website and matching it to the image downloaded from qalerts, we ascertained that all image metadata was inherited from its origin. That is to say, all of the metadata on the image posted to 8kun was also present on the image that was part of the earlier article. It seems like Q downloaded a copy of the image from the Brownfield News website and uploaded that copy to 8kun without altering any of the metadata.\nBut what about images from Q drops which appeared nowhere else online?\nThese seemingly original images proved to be the most interesting subset of all. Several of these images not only could not be matched easily to an online origin (so may be hypothesized to come directly from Q) but also contained enough metadata to analyze.\nIn particular, we soon zeroed in on a group consisting of 29 images whose UserComment fields all contain the word “Screenshot.” Not only are they marked as screenshots, all of these photos appear to be screenshots taken from within a mobile device’s photo gallery. Black bars can be seen above and below most of the photos, as well as elements such as the horizontal, elliptical bar at the bottom of the photo in drop 3642—clearly a user interface element.\nQ drop 3,642\nHowever, the strongest proof that these are screenshots comes from one of the images themselves: drop 4752 contains a screenshot, not merely of a picture Q has selected from within an iPhone’s image gallery and then photographed, but of the gallery itself. This is immediately recognizable to any iPhone user.\nQ drop 4,752\nOf the 29 photos labeled as screenshots, 27 also contain timestamp information in the field\nDateCreated\n. These 27 timestamped screenshots appeared in 29 Q drops (two images each appeared twice).\nWe calculated the time difference between the timestamp of the image metadata and the timestamp in UTC of the Q drop on the imageboard itself, and used this figure to make inferences about the time zone of Q.\nInferring Q’s Time Zone\nThe timestamp metadata associated with these images\ndoes not include a time zone. However, the time zone of the poster, expressed as a\nUTC offset\n, may still be inferred. To infer the time zone, we use the equation:\nLocal time of screenshot + time to post + UTC offset of screenshot maker = UTC time of post\nor\nTime to post + UTC offset of screenshot maker = UTC time of post − Local time of screenshot\nor\nTime to post + UTC offset of screenshot maker = calculated time diff\nWe know the right side of the equation and need the variables on the left. We assume that Q took a screenshot of a photo and then, minutes rather than hours later, a drop was made containing that image. For instance, the image in drop 125 was created at 15:39 local time (recorded by the device taking the screenshot) on July 7, 2017, and posted on 4chan at 15:43 Pacific time (UTC time 23:43). It’s not certain that only four minutes elapsed between the time the picture was processed and posted (rather than some number of hours plus four minutes), but it seems reasonable to assume.\nFurthermore, we know that UTC offsets are constrained; the vast majority of them occur in hourly increments. For instance there is no such thing as a time zone that is one hour and 27 minutes past UTC. Therefore, the assumption that time to post was under an hour fixes the UTC offset for all 27 images.\nInfographic by Robert Amour and Abigail W. Xavier\nExamining the clusters near UTC−7 and UTC−8, we find that they correlate with changes enforced by Daylight Savings in the Pacific time zone. Pacific time changes from UTC−8 to UTC−7 and back over the course of a year. We believe these visually evident clusters provide strong evidence that images in Q drops are sourced from two time zones, UTC+8 and the Pacific time zone.\nFurther Evidence for Time to Post\nThe data indicates that Q either waited for a handful of minutes after taking each screenshot before posting it, or that Q routinely waited for an unknown number of hours plus only a handful of minutes—which would be much more unusual behavior than simply posting right away.\nIn the case of drop 4752, we can verify that\nDateCreated\nis the local device time. Drop 4752 is in our Pacific time cluster, and the image in it includes the iPhone’s clock (which shows 12:07 at the time of the screenshot); our metadata also shows a creation time of 12:07. Q posted it at 12:13 Pacific time.\nIn general we see the same pattern for all Q drops with an image in the UTC−7 / UTC−8 cluster: for instance, drop 1206 has a screenshot taken at 18:16 local time on April 20, 2018 and posted at what corresponds to 18:18 Pacific time, while drop 1752 includes a screenshot made at 11:09 local time, which Q posted at 11:13 Pacific time.\nA similar pattern holds for the UTC+8 cluster. Drop 682 contains a screenshot whose\nDateCreated\nmetadata indicates that it was created at 8:38 local time. It was posted at 8:40 UTC+8 time. Similarly, drops 683–686 include screenshots containing a\nDateCreated\nvalue of 9:25 and 9:26 on August 2, 2018. They were posted between 9:35:08 and 9:36:59 UTC+8 time, a difference of 10 minutes. Finally, drop 2708 contains two photos that were taken at 18:27 and 18:28 local time, and posted at 18:31 UTC+8 time.\nTo bolster our confidence that time-to-post was within the hour and the remaining hour difference represents the device time zone’s UTC offset, we can observe the distribution of the time differences between the image’s creation date and the drop timestamp by minute within the hour. If the time to post were an hour or more on average, we’d expect to see a somewhat even distribution of minutes within the hour difference.\nInfographic by Robert Amour and Abigail W. Xavier\nHowever, the overwhelming majority of images appear to have been posted within 10 minutes, and none appear to have waited more than half an hour before being posted. This allows us to reasonably establish a pattern of behavior and to conclude that the time to post was within the hour, and that remaining hours of the difference represent the UTC offset of the device’s timezone.\nViews from a Hotel Window\nWe know that Q, or an associate close enough to provide them with these photos, traveled to several locations in Asia — a fact which can be inferred not only from the context of these images, but their content. Several are photos of city skylines, others are close-up views of side streets. They generally do not include obvious tourist sites or any close-up individuals whatsoever — par for the course for Q, whose followers pore over images for arcane clues and metaphors.\nFor example, in drop 682, posted on February 8, 2018 (by UTC time), Q provided the following image of a modern skyscraper with a distinctive spire. (The black bars above and below the photo have been cropped in all images in this section.)\nQ later posted one close-up photo of the awning over the building’s main entrance and another of the car park in front of it. These offered some small clues; cars bear blue license plates and a construction site to the right of the building features several bright red placards featuring Chinese characters. The words “Bank of China” can be seen on white backing on the building to the right.\nQ drop 682\nAll of these details suggest that the tower is in the People’s Republic of China — but where? By cropping the distinctive tip of the building and uploading the image into Yandex’s reverse image search, we\nsoon find a match\n: the\nKing Tower\nin Pudong, Shanghai.\nIt is fair to assume that the\nphoto was taken\nfrom a tall building across the New Jinqiao Road to the South-West.\nAccording\nto the popular Chinese mapping website and search engine Baidu, this building is the Shanghai Marriott Hotel Pudong East (上海金桥红枫万豪酒店).\nScreenshot from Baidu Maps\nTwo days later, on 10 February 2018, Q posted a close-up image of a large letter “E” hanging from a ceiling with distinctive alternating wooden slats.\nQ drop 707\nIn the bottom of the image, another item of a similar size and color can be seen hanging from the same ceiling. This leads us to assume that the letter “E” is one of a series in a zoning area — such as those found in airports. A simple Google search of airport arrival hallways in large cities in UTC+8 yields a result. An\nimage from Shutterstock\nshows the same “E” hanging from the same ceiling at Shanghai’s Pudong airport:\nScreenshot of Shutterstock’s website\nQ’s next drop contained a photo from central Bangkok — Q may have meant to suggest that members of the “Q Team” had recently flown between Shanghai Pudong airport and the Thai capital.\nThe image below, posted in drop 708 on February 10, 2018, shows a city skyline. A building with the words “Assumption College” can be seen in the foreground.\nQ drop 708\nThe distinctive “pixelated” skyscraper in the background is the\nMaha Nakhon\ntower in the central business district of Bangkok. When completed in 2016, it was the tallest building in the Thai capital.\nThis photo appears to have been taken from a floor about midway up in a hotel called The Peninsula Bangkok or from a building on the same block with a similar vantage point, such as the River Condominiums or Light House Plaza.\nThe riverside campus of Assumption College is almost perfectly in line with The Peninsula Bangkok and Maha Nakhon. A user-uploaded photo on Google, taken from The Peninsula in March 2021, shows Assumption College and the Maha Nakhon tower from a similar, though higher, perspective to Q’s.\nImage from\nGoogle Maps / The Peninsula Bangkok\nOn March 7, 2018 in drop 869, Q posted an image of a bustling market street in an East Asian country with a reference to Edward Snowden in the accompanying text. In the bottom right-hand corner above a meat stall, and on several banners in the center of the image, Chinese characters can be seen but are not completely legible. Bearing in mind that Snowden first fled to Hong Kong, it’s as good a place to start as any.\nQ drop 869\nThe architecture and commercial profile of the area matches the narrow streets of Mong Kok, a popular shopping district in Hong Kong. A digital trawl through Mong Kok using Baidu’s streetview system shows a\nmatch at the crossroads\nof Nelson Street and Canton Street.\nImage from Baidu street view\nImportantly, Q’s image shows this intersection from high above, though not from a bird’s eye view. This suggests that the photographer was standing in a tall building to the east. The most obvious candidate is the nearby Cordis Hotel.\nImages posted by guests of the hotel and tagged at the location can confirm this match further. In December 2020, one visitor to the Cordis posted a panorama photo of the Mong Kok skyline; the bottom left shows this same intersection in detail.\nImage:\nGoogle Maps / Cordis Hotel\nNevertheless, it should be noted that the screenshot of Q’s image from the Cordis Hotel was created by a user in the Pacific time zone, from which it can be inferred either that Q did not post all images “live” (and might have selected it from their own old photos), or that Q was posting images they had received from a contact.\nOn February 14, 2019, Q posted Drop 2706, which included a photo of a shopping street somewhere in a Chinese-speaking country. Unlike the previous two images, which each contain a famous skyscraper, this image was unremarkable. However, as a close-up image it did show several names of businesses such as 朋泉聚，周师兄 (Old Brother Zhou)，中国民生银行 (China Minsheng Bank), as well as 诗丽堂美容SPA (Shilitang Beauty SPA, also known by its English name Silktone SPA). Two of these are business names concentrated in Chongqing, a city in South-West China.\nQ drop 2,706\nAs it offers a highly detailed business registry for China, Baidu Maps is the ideal source to start searching. It also offers an extensive streetview mode which allows users to compare imagery taken in different years. After checking street view imagery for various addresses for Silktone Beauty SPA branches in Chongqing,\nwe have a match\non the city’s Beijeng Tiangje Street (北城天街). It shows the same shopfront, the same names of all businesses, the same covered bus shelter, and the same distinctive streetlights:\nImage: Baidu Maps\nQ’s photo of the entrance was taken from above, from a building across the road to the North-East. Baidu’s map shows a highly likely candidate: the Chongqing Fuli Hyatt Regency Hotel.\nFurther corroborating this is another image Q posted in the same drop, showing the lights of buildings and a distant cityscape at night. Blurring suggests that this was taken at speed, presumably from an aircraft.\nQ drop 2,706\nThe layout of the roads and buildings\nmatches\nthose to the right of Chongqing Airport’s eastern runway, approaching from the South.\nImage: Google Maps\nImportantly, drop 2503 from November 2018 includes a photo from the hotel, including a Christmas tree. That same month, in drop 2516, Q congratulated “anons” for successfully geolocating the “last pic” to Chongqing, urging them to “watch the news”.\nIf these images are, as they appear to be, original, then we know the location of the photographer, and can infer the location of the person who took the screenshot, but not necessarily that of the person who posted the respective drops: it is at least conceivable that whoever took the screenshots might have texted or emailed these photos to a third party, who posted them shortly afterwards.\nNevertheless, the images do teach us a few new things about Q.\nWe know that Q, or a very close associate, flew to Shanghai Pudong Airport prior to February 8, 2018; that they stayed at The Peninsula hotel in Bangkok prior to or on February 10, 2018; that they stayed at the Cordis Hotel in Hong Kong prior to or on March 7, 2018; that they stayed at the Chongqing Fuli Hyatt prior to or on February 14, 2019 and flew into Chongqing prior to that same date.\nWe know that Q, or an associate who regularly provides images for Q to post, has a particular taste in hotels and has spent time traveling, but likely not living, in four cities in Thailand and China. The fact that some of the images in China can also be geolocated to the same timezone makes the assumption about Q’s location, or the location of somebody working with Q, more plausible.\nOne Step Closer?\nThe overwhelming majority of the images in our data set — over 68 percent — appear to have been created by a user, or users who were in the Pacific time zone. Q’s content appears to have been posted from the Pacific time zone at least as early as November 7, 2017 and at least as late as September 23, 2020; this covers almost the entire period that Q was active.\nBy contrast, images from UTC+8 exist on only three dates in our data: February 8, 2018, February 10, 2018, and February 14, 2019.\nThis has several potential implications for Q’s identity.\nFirst, these findings are consistent with the hypothesis that the same person has been posting as Q all along. In that case, Q would live or spend the vast majority of their time in the Pacific time zone, with occasional international travel. The hypothesis of a Pacific Q coheres with\nother quantitative work\non the overall hourly patterns of the Q drops, while the hypothesis of a single Q matches the\nobserved\nfact\nthat Q had access to the same set of photos in November 2017 and May 2018.\nSecond, the most\npopular hypothesis\namong Q-watchers — that authorship of the drops changed hands between January 4 and January 6 of 2018 — cannot be excluded on the basis of our findings. Under this hypothesis, our evidence requires new Q and original Q both to post from Pacific Time.\nThat’s not impossible, of course. There are well over 50 million people living in the Pacific time zone; one of them could easily have stepped into the original Q’s place.\nWe reached out to Jim Watkins, the owner of 8chan/8kun who has lived in the Philippines, to comment on the implications of narrowing down Q’s timezone. In an emailed response\ncrowdsourced\nvia his Gab account, Watkins cast doubt on the methodology of the research and its usefulness, adding that more than a billion people live in the UTC+8 time zone. Watkins has\npreviously denied\nin comments to Bellingcat that he is Q.\nBellingcat was unable to reach Ron Watkins, who has lived in Japan, for comment. Ron Watkins has also denied that he is Q in the course of the recent HBO documentary series.\nIt is important to stress our evidence is not by itself sufficient to identify any particular individual as Q or an associate of Q.\nHowever, our evidence should provide a valuable datapoint for other investigators who seek to do so. The hunt for Q goes on — and these findings suggest that it must largely continue in the Pacific time zone.\nGavin Fox (pseudonym) contributed to this research. The authors would like to thank Bellingcat’s Giancarlo Forella and Maxim Edwards for their assistance in geolocating the images from East Asia.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image14-300x160.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image19-300x200.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image13-139x300.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image4-138x300.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/static.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image7-1-300x188.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image8-300x156.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image20-224x300.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image3-300x206.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image11-225x300.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image2-300x263.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image12-225x300.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image15-300x262.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image16-225x300.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image6-300x128.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image10-300x170.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image1-300x223.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image18-300x131.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image17-224x300.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/image9-300x199.png",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2021-05-18_unsure-when-a-video-or-photo-was-taken-how-to-tell-by-measuring-the-length-of-shadows.json b/bellingcat_geolocation_json/2021-05-18_unsure-when-a-video-or-photo-was-taken-how-to-tell-by-measuring-the-length-of-shadows.json
new file mode 100644
index 0000000000000000000000000000000000000000..591eacd3bd4a54df91138aaa940353f34553a785
--- /dev/null
+++ b/bellingcat_geolocation_json/2021-05-18_unsure-when-a-video-or-photo-was-taken-how-to-tell-by-measuring-the-length-of-shadows.json
@@ -0,0 +1,52 @@
+{
+ "slug": "unsure-when-a-video-or-photo-was-taken-how-to-tell-by-measuring-the-length-of-shadows",
+ "title": "Unsure When a Video or Photo was Taken? How to Tell by Measuring the Length of Shadows",
+ "url": "https://www.bellingcat.com/resources/2021/05/18/unsure-when-a-video-or-photo-was-taken-how-to-tell-by-measuring-the-length-of-shadows/",
+ "date": "2021-05-18T11:29:04+00:00",
+ "modified": "2021-05-18T11:35:46+00:00",
+ "author": "Nick Waters",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Establishing the time of day that a video or image was taken can be a difficult and frustrating task for open source researchers. But measuring the length of shadows can help provide a solution.",
+ "content_html": "
Establishing the time within a video or image can be a difficult and frustrating task for open source researchers. If you’re lucky, you’ll get a brief glance at a wall-clock or wristwatch — but in most cases you won’t. Measuring shadows, however, can help provide a solution. \n \nIt’s already well known that it’s possible to use the angle of shadows to assess the time in an image. SunCalc, a tool which simulates the movement of the sun through the sky on different dates and locations across the Earth, can be used to do this relatively easily.
\n
However, sometimes the angle of these shadows might not be clear or there may not be sufficient reference points in a video or picture to accurately judge the angle.
\n
But just as the angle of a shadow can be used to work out the angle of the sun from North (the azimuth), so the relative length of a shadow can be used to work the angle of the sun from the horizon (the altitude).
\n
As the sun rises in the sky, shadows will initially point west and start long before reaching their shortest length at around midday. They will then point east as they gradually lengthen before sunset. This predictable change in length can be used to judge the time. \n \nImportantly, it’s not necessary to know the actual height of the object casting the shadow. What matters is only the proportional length of the shadow compared to the object. This means that the measurements seen in this article bear no relation to the actual heights and lengths of the objects and shadows. But since we’re only interested in their proportions in comparison to each other, this doesn’t matter.
\n
In order for this technique to work well, the image or video needs to fit several conditions:
\n\n
The location and date of the image must be known
\n
The object and its shadow must be roughly at right angles to the camera
\n
The shadow must be cast on a surface level with the horizon (a flat floor)
\n
The image should not be overly distorted by camera lenses, such as fish-eye lenses
\n\n
Finally, when carrying out this technique, it should be noted that shadows will be a particular length twice during a day: once in the morning, once in the afternoon. Therefore it’s important to check the direction of the shadow.
\n
Example One: Controlled Conditions
\n
In this example I have used a photo I took at the Al Aqsa compound in Jerusalem on 14 October 2019. Let’s apply the shadow by proportional length technique and see what time we calculate.
\n
The Al Aqsa compound in Jerusalem. Photo by Nick Waters.
\n
In this photo a member of the security forces can be seen leaning against a fence. We can see that the picture fits the four conditions:
\n\n
The location and date of the image must be known
\n
The object and its shadow must be roughly at right angles to the camera
\n
The shadow must be cast on a surface level with the horizon (a flat floor)
\n
The image should not be overly distorted by camera lenses, such as fish-eye lenses
\n\n
Next, we measure the lengths of the object and shadow within the image. Remember that what is important here is the proportion between these two measurements, rather than the actual measurements themselves.
\n
The Al Aqsa compound in Jerusalem with shadow measurements added. Photo by Nick Waters.
\n
We then go to Suncalc, set the location to this spot in the Al Aqsa Compound and input the height of the object as 2.33 metres, which SunCalc refers to as “at an object level”. We then adjust the time until the shadow length is as close as possible to that seen in the image (3.53 m). The SunCalc input process is detailed in the image below.
\n
Suncalc.org calculation of shadow length. Click or tap to see full size image.
\n
This gives us a time of 09:37. After conducting this analysis I checked the EXIF data from my camera and found it had been taken at 09:43, a mere six minute difference. \n
\n
Example Two: Still From Video
\n
In the example below we see a still from a video showing a group of people watching a Saudi-led coalition airstrike which killed scores of civilians in the town of Zabid, Yemen, on 12 May 2015.
\n
Still image from a video posted by Mohammad al-Hadad.
\n
As we can see, the shadow of the man in the white shirt on the left is being cast on a flat surface at a rough right angle to the camera. We can therefore measure the length of the shadow compared to the length of the object casting it in order to calculate a time.
\n
In this case, the height of the man in the image was measured as 5.67 centimetres and his shadow, which appears to just stretch past the leg of the man to the right, as 9.84 cm.
\n
Still image from a video posted by Mohammad al-Hadad with shadow measurements added.
\n
Using Suncalc, we then set the “object level” (the height of the object casting the shadow) as 5.67m and change the time until the shadow length shows as close as possible to 9.84m. This gives us a time of 16:14.
\n
Suncalc.org calculation of shadow length. Click or tap to see full size image.
\n
In order to test the accuracy of the method in this case we can change the time by 15 minutes either side of 16:14 to see what shadow lengths we would get.
\n
\n
At 16:14 the shadow is 1.74 times the height of the object casting it.
\n
15 minutes earlier, at 15:59 the shadow is 1.51 times the height of the object casting it.
\n
15 minutes later, at 1629 the shadow is 2.02 times the height of the object casting it.
\n
\n
Still image from a video posted by Mohammad al-Hadad with shadow measurements added.
\n
As we can see, although the time established by this shadow calculation should be regarded as relatively rough, it is still certainly within a period of 15 minutes either side of 16:14.
\n
When applying this technique it’s often useful to establish a time window in this way, rather than stating a specific time. This can be done quite easily by adjusting the time within SunCalc until the length of the shadow is clearly no longer consistent with the calculation.
\n
Conclusion
\n
Although this technique is slightly more difficult to apply than judging time by angle of shadow, it can still be useful when there are insufficient reference points to estimate an angle. As long as the correct conditions are met, this technique can be reasonably accurate, sometimes surprisingly so.
\n
If you’re interested in how SunCalc can also be used for geolocation, we’ve written about this technique here.
\n
",
+ "content_text": "Translations:\nEnglish (UK)\nРусский (Россия)\nEstablishing the time within a video or image can be a difficult and frustrating task for open source researchers. If you’re lucky, you’ll get a brief glance at a wall-clock or wristwatch — but in most cases you won’t. Measuring shadows, however, can help provide a solution.\nIt’s already well known that it’s possible to use the angle of shadows to assess the time in an image.\nSunCalc\n, a tool which simulates the movement of the sun through the sky on different dates and locations across the Earth, can be used to do this relatively easily.\nHowever, sometimes the angle of these shadows might not be clear or there may not be sufficient reference points in a video or picture to accurately judge the angle.\nBut just as the angle of a shadow can be used to work out the angle of the sun from North (the azimuth), so the relative length of a shadow can be used to work the angle of the sun from the horizon (the altitude).\nAs the sun rises in the sky, shadows will initially point west and start long before reaching their shortest length at around midday. They will then point east as they gradually lengthen before sunset. This predictable change in length can be used to judge the time.\nImportantly, it’s not necessary to know the actual height of the object casting the shadow. What matters is only the\nproportional\nlength of the shadow compared to the object. This means that the measurements seen in this article bear no relation to the actual heights and lengths of the objects and shadows. But since we’re only interested in their proportions in comparison to each other, this doesn’t matter.\nIn order for this technique to work well, the image or video needs to fit several conditions:\nThe location and date of the image must be known\nThe object and its shadow must be roughly at right angles to the camera\nThe shadow must be cast on a surface level with the horizon (a flat floor)\nThe image should not be overly distorted by camera lenses, such as fish-eye lenses\nFinally, when carrying out this technique, it should be noted that shadows will be a particular length twice during a day: once in the morning, once in the afternoon. Therefore it’s important to check the direction of the shadow.\nExample One: Controlled Conditions\nIn this example I have used a photo I took at the Al Aqsa compound in Jerusalem on 14 October 2019. Let’s apply the shadow by proportional length technique and see what time we calculate.\nThe\nAl Aqsa compound in Jerusalem.\nPhoto by Nick Waters.\nIn this photo a member of the security forces can be seen leaning against a\nfence\n. We can see that the picture fits the four conditions:\nThe location and date of the image must be known\nThe object and its shadow must be roughly at right angles to the camera\nThe shadow must be cast on a surface level with the horizon (a flat floor)\nThe image should not be overly distorted by camera lenses, such as fish-eye lenses\nNext, we measure the lengths of the object and shadow within the image. Remember that what is important here is the proportion between these two measurements, rather than the actual measurements themselves.\nThe\nAl Aqsa compound in Jerusalem with shadow measurements added.\nPhoto by Nick Waters.\nWe then go to Suncalc, set the location to this spot in the Al Aqsa Compound and input the height of the object as 2.33 metres, which SunCalc refers to as “at an object level”. We then adjust the time until the shadow length is as close as possible to that seen in the image (3.53 m). The SunCalc input process is detailed in the image below.\nSuncalc.org\ncalculation of shadow length. Click or tap to see full size image.\nThis gives us a time of 09:37. After conducting this analysis I checked the EXIF data from my camera and found it had been taken at 09:43, a mere six minute difference.\nExample Two: Still From Video\nIn the example below we see a still from a video showing a group of people watching a Saudi-led coalition\nairstrike\nwhich killed scores of civilians in the town of Zabid, Yemen, on 12 May 2015.\nStill image from a\nvideo\nposted by Mohammad al-Hadad.\nAs we can see, the shadow of the man in the white shirt on the left is being cast on a flat surface at a rough right angle to the camera. We can therefore measure the length of the shadow compared to the length of the object casting it in order to calculate a time.\nIn this case, the height of the man in the image was measured as 5.67 centimetres and his shadow, which appears to just stretch past the leg of the man to the right, as 9.84 cm.\nStill image from a\nvideo\nposted by Mohammad al-Hadad with shadow measurements added.\nUsing\nSuncalc\n, we then set the “object level” (the height of the object casting the shadow) as 5.67m and change the time until the shadow length shows as close as possible to 9.84m. This gives us a time of 16:14.\nSuncalc.org\ncalculation of shadow length. Click or tap to see full size image.\nIn order to test the accuracy of the method in this case we can change the time by 15 minutes either side of 16:14 to see what shadow lengths we would get.\nAt 16:14 the shadow is 1.74 times the height of the object casting it.\n15 minutes earlier, at 15:59 the shadow is 1.51 times the height of the object casting it.\n15 minutes later, at 1629 the shadow is 2.02 times the height of the object casting it.\nStill image from a\nvideo\nposted by Mohammad al-Hadad with shadow measurements added.\nAs we can see, although the time established by this shadow calculation should be regarded as relatively rough, it is still certainly within a period of 15 minutes either side of 16:14.\nWhen applying this technique it’s often useful to establish a time window in this way, rather than stating a specific time. This can be done quite easily by adjusting the time within SunCalc until the length of the shadow is clearly no longer consistent with the calculation.\nConclusion\nAlthough this technique is slightly more difficult to apply than judging time by angle of shadow, it can still be useful when there are insufficient reference points to estimate an angle. As long as the correct conditions are met, this technique can be reasonably accurate, sometimes surprisingly so.\nIf you’re interested in how SunCalc can also be used for geolocation, we’ve written about this technique\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/Temple-mount-1st.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/Temple-Mount-2nd-1200x835.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/Temple-Mount-Suncalc-1200x797.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/Zabid-First-picture.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/Zabid-chronolocation.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/Zabid-Suncalc-1200x787.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/05/pasted-image-0.png",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2022-10-28_how-we-geolocated-a-photo-of-a-russian-missile-programming-team.json b/bellingcat_geolocation_json/2022-10-28_how-we-geolocated-a-photo-of-a-russian-missile-programming-team.json
new file mode 100644
index 0000000000000000000000000000000000000000..676f397617574cf9f48da3e3d447d7eae63aceca
--- /dev/null
+++ b/bellingcat_geolocation_json/2022-10-28_how-we-geolocated-a-photo-of-a-russian-missile-programming-team.json
@@ -0,0 +1,92 @@
+{
+ "slug": "how-we-geolocated-a-photo-of-a-russian-missile-programming-team",
+ "title": "How we Geolocated a Photo of a Russian Missile Programming Team",
+ "url": "https://www.bellingcat.com/resources/2022/10/28/how-we-geolocated-a-photo-of-a-russian-missile-programming-team/",
+ "date": "2022-10-28T12:48:06+00:00",
+ "modified": "2022-10-28T13:48:49+00:00",
+ "author": "Aric Toler",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Geolocating a photo of Russian missile programmers was one key part of our investigation into the people behind cruise missiles strikes on Ukraine.",
+ "content_html": "
\n
While conducting research for his latest Bellingcat investigation, Christo Grozev received a group photograph of the missile guidance team that were purported to be behind programming many of the cruise missiles that have hit Ukraine in recent months. As detailed in the article:
\n
“Another team member, whose identity is not known as they contacted reporters via a burner email account that was provided by Bellingcat and The Insider to all contacted members, shared two group photos of the GVC team and two photos of their commander, Lt. Col. Bagnyuk, wearing his many medals.”
\n
One of the photographs sent to Bellingcat anonymously by a GVC team member.
\n
The team member who provided this 2013 photograph said it was taken at the Main Computation Centre of the General Staff (GVC) in Moscow, but we needed to independently verify this ourselves. During the course of our investigation, we had already discovered the GVC was located at Znamenka 19 – the headquarters of the Russian Ministry of Defence (MoD). This fact was gleaned by looking for the names of engineers we had identified as working for the GVC in leaked Russian databases (including those that contain information about addresses used to register cars or sign up to online delivery apps). Some had registered addresses at Znamenka 19.
\n
As such, the next challenge was to use geolocation techniques to confirm conclusively where the image was taken so that we could include it in our findings about the missile team.
\n
This brief article provides a walkthrough on how we geolocated this photograph, eventually verifying that it was indeed taken at the Russian MoD’s Znamenka 19 facility, one of the locations that hosts the GVC.
\n
Details Within the Photograph
\n
As with all geolocation, the first step is to identify details in the photograph that could be useful. There are broadly two types of details that we look for when performing this type of task: big-picture details, and micro-details. The big-picture details are ones that inform us where to look in the first place – in this case, what does the building look like from the street or from satellite images taken at an angle?
\n
The micro-details help us make a positive identification of the site once we find a potential match, such as cracks in a floor, paint patterns on building columns, the design of a no-smoking sign and so on.
\n
Some of the big-picture details that we identified for this photo include:
\n
\n
The red-brick wall pattern and the presence of a roof at the entranceway that is propped up by white columns.
\n
\n
Red brick wall pattern detailed in picture provided to Bellingcat.
\n
\n
The presence of a white wall or columns with windows in between, opposite of the brick building- as seen in a reflection in the glass.
\n
\n
A close-up of the photo background shows the reflection of a white wall or columns in a window behind where the photo was taken.
\n
\n
Three-sided stairs leading up to the entrance, which may be visible in satellite imagery taken at an angle.
\n
\n
The distinctive three-sided steps detailed in the image.
\n
Some of the micro-details that we identified for this photo include:
\n
\n
The intersections of the upper-brick pattern, the lower-brick (with a lighter red colour), and the white path along the wall and a gutter drain alongside that may be visible in imagery taken closer to the area. We could not expect to clearly see these details in satellite imagery, unless it was very high-resolution.
\n
\n
A close up of the brick patterns and gutter drain visible in the photo.
\n
\n
A no-smoking sign near the entrance door, which would only be visible from close by or an image that was taken from a short distance away. The design of the sign may provide extra clues as well.
\n
\n
A no-smoking sign visible in the photo.
\n
\n
The door and window design, as well as the colour of these items.
\n
\n
A zoomed in image of a door and windows visible in the photo.
\n
\n
Reflection showing barred windows with air conditioning units and a cable running above the far window.
\n
\n
A zoomed-in image showing the reflection of air conditioning units and a cable.
\n
\n
Cracks and discolouration in the bottom step and the ground.
\n
\n
A cropped image shows features of the bottom step and ground visible in the photo.
\n
A cropped image shows similar features elsewhere on the steps.
\n
There are plenty of additional micro-details to pick out, but these are the ones we honed in on.
\n
Znamenka or Frunze Naberezhnaya?
\n
Even though the source who provided the photograph noted that it was the Main Computation Centre of the General Staff (GVC) in Moscow, we had to first discover where that was.
\n
As detailed above, we did this by cross-referencing car-registration data and delivery orders of the people we had identified as working at the GVC with leaked Russian databases. Once we had confirmed the GVC staff were based at Znamenka 19, we then had to prove that the photo was indeed taken there. Our initial geolocation efforts gave us the suspicion that the photo may actually have been taken at another primary Ministry of Defence facility (the National Defence Management Centre), located on Frunze Naberezhnaya along the Moskva River.
\n
When looking at the white wall with windows reflected in the group photograph, we thought that there was a resemblance to similar walls at the Frunze Naberezhnaya facility, as seen in the reflection this photograph published by the Russian MoD.
\n
Top: A screengrab from a Russian Ministry of Defense video (may require a VPN to access) shows white walls in the reflective frontage that appeared similar to that detailed in the photo provided to Bellingcat. Bottom: A close up of a reflective frontage in the photo provided to Bellingcat in which white walls are visible.
\n
A Russian state television documentary detailing this MoD facility shows additional perspectives of the building, along with explaining how defence officials guided strikes in the Syrian campaign from this headquarters (at approximately 27:15 of the video). Yet despite these strong contextual clues (the GVC team was also involved in missile programming in Syria), we could not find any images that matched the entranceway in the group photograph around the grounds of the Frunze Naberezhnaya facility, either from 2013 (the apparent date of the photograph) or more recently. It’s worth noting that this facility has undergone extensive renovations over the last decade, further complicating the verification efforts.
\n
Back to Znamenka 19
\n
With the Frunze Naberezhnaya facility ruled a less-likely candidate, despite the red herring of the white walls, we studied satellite imagery of the location more closely. We also looked for user-uploaded photographs of the buildings. This location is not exactly a tourist destination – it is under extremely strict security, and unlike the Frunze Naberezhnaya facility, there are seemingly no ordinary people visiting for ceremonies, such as concerts.
\n
After looking through all of the satellite imagery we could find, we found a potential match for the entrance with the three-sided pattern observed from the steps: the southeast corner of the courtyard within the Znamenka 19 grounds on Yandex (note: map below has been reversed from north).
\n
A potential match for the stairs where the missile program team was seen posing.
\n
We then hit the geolocation jackpot: by searching through geotagged photos on Vkontakte for Znamenka street in Moscow, we found a high-resolution photograph taken by a professional roofer atop the north wall of the facility, facing south.
\n
An image found on Vkontakte appears to show the same entranceway as seen in the GVC team photo.
\n
The VK post tagged to Znamenka.
\n
Clearly visible in the bottom-left corner of the image is the same entrance that we see in the group photograph.
\n
Top: A close up of the entranceway shown in a picture found on Vkontakte. Bottom: The original photo of the GVC team at the same spot.
\n
Going back through our checklist of big-picture and micro-details, almost everything matches, even though this photograph was taken eight years later. The number of stairs and their colour, the colour of the bricks, the presence of the gutter drain, the barred windows, the grey door to the right of the posing group, the white columns and everything else that is visible are a clear match.
\n
We can now confidently say that the photograph sent to us by a source from the missile guidance team was indeed taken at the southeast courtyard entryway into the Znamenka 19 facility.
\n\n
Michael Sheldon and Timmi Allen contributed to this report.
\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter here.
\n
",
+ "content_text": "While conducting research for his latest Bellingcat\ninvestigation\n, Christo Grozev received a group photograph of the missile guidance team that were purported to be behind programming many of the cruise missiles that have hit Ukraine in recent months. As detailed in the article:\n“Another team member, whose identity is not known as they contacted reporters via a burner email account that was provided by Bellingcat and The Insider to all contacted members, shared two group photos of the GVC team and two photos of their commander, Lt. Col. Bagnyuk, wearing his many medals.”\nOne of the photographs sent to Bellingcat anonymously by a GVC team member.\nThe team member who provided this 2013 photograph said it was taken at the Main Computation Centre of the General Staff (GVC) in Moscow, but we needed to independently verify this ourselves. During the course of our investigation, we had already discovered the GVC was located at Znamenka 19 – the headquarters of the Russian Ministry of Defence (MoD). This fact was gleaned by looking for the names of engineers we had identified as working for the GVC in leaked Russian databases (including those that contain information about addresses used to register cars or sign up to online delivery apps). Some had registered addresses at Znamenka 19.\nAs such, the next challenge was to use geolocation techniques to confirm conclusively where the image was taken so that we could include it in our findings about the missile team.\nThis brief article provides a walkthrough on how we geolocated this photograph, eventually verifying that it was indeed taken at the Russian MoD’s Znamenka 19 facility, one of the locations that hosts the GVC.\nDetails Within the Photograph\nAs with all\ngeolocation\n, the first step is to identify details in the photograph that could be useful. There are broadly two types of details that we look for when performing this type of task: big-picture details, and micro-details. The big-picture details are ones that inform us where to look in the first place – in this case, what does the building look like from the street or from satellite images taken at an angle?\nThe micro-details help us make a positive identification of the site once we find a potential match, such as cracks in a floor, paint patterns on building columns, the design of a no-smoking sign and so on.\nSome of the big-picture details that we identified for this photo include:\nThe red-brick wall pattern and the presence of a roof at the entranceway that is propped up by white columns.\nRed brick wall pattern detailed in picture provided to Bellingcat.\nThe presence of a white wall or columns with windows in between, opposite of the brick building- as seen in a reflection in the glass.\nA close-up of the photo background shows the reflection of a white wall or columns in a window behind where the photo was taken.\nThree-sided stairs leading up to the entrance, which may be visible in satellite imagery taken at an angle.\nThe distinctive three-sided steps detailed in the image.\nSome of the micro-details that we identified for this photo include:\nThe intersections of the upper-brick pattern, the lower-brick (with a lighter red colour), and the white path along the wall and a gutter drain alongside that may be visible in imagery taken closer to the area. We could not expect to clearly see these details in satellite imagery, unless it was very high-resolution.\nA close up of the brick patterns and gutter drain visible in the photo.\nA no-smoking sign near the entrance door, which would only be visible from close by or an image that was taken from a short distance away. The design of the sign may provide extra clues as well.\nA no-smoking sign visible in the photo.\nThe door and window design, as well as the colour of these items.\nA zoomed in image of a door and windows visible in the photo.\nReflection showing barred windows with air conditioning units and a cable running above the far window.\nA zoomed-in image showing the reflection of air conditioning units and a cable.\nCracks and discolouration in the bottom step and the ground.\nA cropped image shows features of the bottom step and ground visible in the photo.\nA cropped image shows similar features elsewhere on the steps.\nThere are plenty of additional micro-details to pick out, but these are the ones we honed in on.\nZnamenka or Frunze Naberezhnaya?\nEven though the source who provided the photograph noted that it was the Main Computation Centre of the General Staff (GVC) in Moscow, we had to first discover where that was.\nAs detailed above, we did this by cross-referencing car-registration data and delivery orders of the people we had identified as working at the GVC with leaked Russian databases. Once we had confirmed the GVC staff were based at Znamenka 19, we then had to prove that the photo was indeed taken there. Our initial geolocation efforts gave us the suspicion that the photo may actually have been taken at another primary Ministry of Defence facility (the National Defence Management Centre), located on Frunze Naberezhnaya along the Moskva River.\nWhen looking at the white wall with windows reflected in the group photograph, we thought that there was a resemblance to similar walls at the Frunze Naberezhnaya facility, as seen in the reflection this photograph published by the Russian MoD.\nTop: A screengrab from a Russian Ministry of Defense video (may require a VPN to access) shows white walls in the reflective frontage that appeared similar to that detailed in the photo provided to Bellingcat. Bottom: A close up of a reflective frontage in the photo provided to Bellingcat in which white walls are visible.\nA\nRussian state television documentary\ndetailing this MoD facility shows additional perspectives of the building, along with explaining how defence officials guided strikes in the Syrian campaign from this headquarters (at approximately 27:15 of the video). Yet despite these strong contextual clues (the GVC team was also involved in missile programming in Syria), we could not find any images that matched the entranceway in the group photograph around the grounds of the Frunze Naberezhnaya facility, either from 2013 (the apparent date of the photograph) or more recently. It’s worth noting that this facility has\nundergone extensive renovations\nover the last decade, further complicating the verification efforts.\nBack to Znamenka 19\nWith the Frunze Naberezhnaya facility ruled a less-likely candidate, despite the red herring of the white walls, we studied satellite imagery of the location more closely. We also looked for user-uploaded photographs of the buildings. This location is not exactly a tourist destination – it is under extremely strict security, and unlike the Frunze Naberezhnaya facility, there are seemingly no ordinary people visiting for ceremonies, such as\nconcerts\n.\nAfter looking through all of the satellite imagery we could find, we found a potential match for the entrance with the three-sided pattern observed from the steps: the\nsoutheast corner\nof the courtyard within the Znamenka 19 grounds on Yandex (note: map below has been reversed from north).\nA potential match for the stairs where the missile program team was seen posing.\nWe then hit the geolocation jackpot: by searching through geotagged photos on Vkontakte for Znamenka street in Moscow, we found a high-resolution photograph taken by a professional roofer atop the north wall of the facility, facing south.\nAn image found on Vkontakte appears to show the same entranceway as seen in the GVC team photo.\nThe VK post tagged to Znamenka.\nClearly visible in the bottom-left corner of the image is the same entrance that we see in the group photograph.\nTop: A close up of the entranceway shown in a picture found on Vkontakte. Bottom: The original photo of the GVC team at the same spot.\nGoing back through our checklist of big-picture and micro-details, almost everything matches, even though this photograph was taken eight years later. The number of stairs and their colour, the colour of the bricks, the presence of the gutter drain, the barred windows, the grey door to the right of the posing group, the white columns and everything else that is visible are a clear match.\nWe can now confidently say that the photograph sent to us by a source from the missile guidance team was indeed taken at the southeast courtyard entryway into the Znamenka 19 facility.\nMichael Sheldon and Timmi Allen contributed to this report.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image6-4.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image7-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image9-2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image11-2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image8-2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image3-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image1-3.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image13-2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image15-2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image10-2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image2-2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image4-2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image14-3.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2022/10/image5-2.png",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2023-02-03_wanted-by-interpol-relaxing-in-dubai-geolocating-isabel-dos-santos-life-of-luxury.json b/bellingcat_geolocation_json/2023-02-03_wanted-by-interpol-relaxing-in-dubai-geolocating-isabel-dos-santos-life-of-luxury.json
new file mode 100644
index 0000000000000000000000000000000000000000..51c1ebd8fda868d862e8a36fd1b78b34167d5926
--- /dev/null
+++ b/bellingcat_geolocation_json/2023-02-03_wanted-by-interpol-relaxing-in-dubai-geolocating-isabel-dos-santos-life-of-luxury.json
@@ -0,0 +1,97 @@
+{
+ "slug": "wanted-by-interpol-relaxing-in-dubai-geolocating-isabel-dos-santos-life-of-luxury",
+ "title": "Wanted by Interpol, Relaxing in Dubai: Geolocating Isabel dos Santos’ Life of Luxury",
+ "url": "https://www.bellingcat.com/news/2023/02/03/wanted-by-interpol-relaxing-in-dubai-geolocating-isabel-dos-santos-life-of-luxury/",
+ "date": "2023-02-03T10:41:23+00:00",
+ "modified": "2026-02-11T18:19:05+00:00",
+ "author": "Miguel Ramalho",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Open source information indicates Angolan billionaire has been living in UAE since Red Notice issued for her provisional arrest.",
+ "content_html": "
\n
On 24 November 2022 DW Africa published an interview with Isabel dos Santos – once Africa’s richest woman and the daughter of former Angolan President José dos Santos. She wore a distinctive pink blouse and sat in front of a nondescript, semi-transparent curtain.
\n
Towards the end of the interview dos Santos was asked “Where are you currently at?” to which she replied, after a brief hesitation, “I’m at home”.
\n
She repeated this answer to all further queries about her location.
\n
It was not entirely clear where home was – dos Santos has multiple properties including in London, Lisbon and Dubai – recently, she’s been reported to spend much of her time in Dubai and she told The New York Times in November that she was living in London.
\n
Her whereabouts were of significance at the time of the DW interview as Interpol had reportedly issued a Red Notice requesting global law enforcement locate and provisionally arrest her. Interpol confirmed this a week later.
\n
\n
Collage illustration by Klawe Rzecy. Elements from Pixabay and Wikimedia Commons.
\n
Dos Santos is the central figure in the Luanda Leaks – a trove of leaked documents – that exposed decades of confidential business dealings and resulted in accusations of embezzlement and fraud by Angolan prosecutors, as well as the freezing of more than a billion dollars worth in assets worldwide. Dos Santos maintains she made her money through legitimate business endeavours and denies accusations of fraud and embezzlement. She has described the Luanda Leaks as “without substance” and described the government’s charges as a political “witch hunt”.
\n
At time of publication on 03 February 2023, dos Santos was yet to be provisionally detained.
\n
But she is by no means in hiding. Dos Santos continues to regularly post on her social media channels, detailing visits to high-end restaurants and nights out – something she alluded to in her DW interview when she insisted she was not disguising her whereabouts.
\n
As Bellingcat discovered, those posts reveal very precise information about Isabel dos Santos’ movements over the past three months and show which of her properties she now appears to call home.
\n
Dining and Dancing
\n
When speaking to DW, Isabel Dos Santos denied she was in hiding, stating: “I’m on social media every day. My whereabouts and place of residence are known”.
\n
One look at Dos Santos’ social media accounts confirms that she is indeed online most days. Dos Santos maintains verified Instagram and Facebook accounts. She has not posted on Facebook since August of last year, but posts regularly on Instagram. She also appears to maintain Twitter and TikTok accounts. Although not verified, we observed on several occasions posts appearing on the TikTok account a few minutes before they appeared on her verified Instagram account, indicating likely ownership by dos Santos.
\n
Among other things, what these accounts show is that dos Santos continues to enjoy a life of luxury.
\n
Here’s a post from TikTok dated December 4, geolocated to the pool of Dubai’s Bulgari Yacht Club.
\n
\n
Still from a TikTok featuring Isabel dos Santos, posted on 4 December 2022, that we geolocated to Dubai’s Bulgari Yacht Club, after matching it with this image from Tripadvisor.
\n
Imagery on Tripadvisor shows a 2019 photo with a near identical perspective to the video from 2022. We also checked where the sun would have been at that location on the day the video was posted and found the sun azimuth is consistent with the day and hour of post – within a range of five minutes.
\n
In another instance, dos Santos’ was tagged by a friend in an Instagram post on 27 December 2022, where she can be seen enjoying a meal at Nusr-Et Steakhouse Dubai. She briefly featured in another friend’s post from the restaurant that day and reviews of the restaurant point to Nusret Gökçe – also known as Salt Bae – presence at his Dubai outlet in the final week of December 2022.
\n\n
\n
\nLeft: A screenshot from an Instagram story featuring Isabel dos Santos. Right: A screenshot of Nusret Gökçe also known as Salt Bae, from the same Instagram story.\n
In this picture posted on another friend’s instagram on 8 January, 2023. Dos Santos can be seen at the left of the frame, under a brightly lit roof. This matches the red and purple corridor feature at Dubai restaurant Trove within the Dubai Mall complex which can be seen here.
\n
\n
Isabel dos Santos can be seen in the corner of the frame in an Instagram post she was tagged in. The red and purple corridor behind her matches this image from Trove, a restaurant in the Dubai Mall complex.
\n\n
A Possible Sardinia Visit?
\n
It is worth noting that dos Santos’ online sharing patterns are not without noise, as she often reposts older content alongside more contemporary images.
For example, on 7 December 2022, a Tiktok appeared from a beach where the landscape seemed vastly different from Dubai. It also featured Spanish subtitles. Could she have left the UAE? We used PeakVisor to verify this location and found it to be in Sardinia, Italy. Dos Santos also posted from this location back in September 2022, but has since deleted the older post.
\n
\n
A Mountain silhouette created using PeakVisor laid over a stitch of a TikTok video featuring dos Santos’ at Sardinian Beach Li Itriceddi.
\n
A visit in September when temperatures reached an average high of 27°C would likely make more sense than a visit in December when the average high is 15.4°C, especially as people can be seen bathing in the water behind dos Santos.
\n
But another clue comes from a major global event, and points to one potential international visit that dos Santos apparently made after the Interpol Red Notice was issued.
\n
The World Cup
\n
On 5 December dos Santos’ posted a picture of herself in a Brazil football shirt with a big screen visible behind her. We geolocated this picture to the official FIFA fan zone set up in Dubai for the 2022 World Cup. Dubai’s skyline can be seen in the background. South Korea played Brazil in the round of 16 that evening and the Brazil manager, Tite, can also be seen on the screen behind dos Santos.
Dos Santos was in Dubai on the evening of December 5 for the match which kicked off at 11pm local time. She was also watching football in Dubai on December 10 and posted from Dubai on December 7 – the same day as the Sardinia post. Apart from the single post of the Beach Li Itriceddi there is no other evidence to suggest that dos Santos’ visited Sardinia around this time.
\n
\n
An image of dos Santos at a FIFA fan zone during the South Korea versus Brazil World Cup Match, posted on her Instagram account on December 5, 2022.
\n
This wasn’t the last time dos Santos posted about the 2022 World Cup. She also appeared to post about the tournament from outside Dubai.
\n
On 13 December dos Santos shared a now removed Instagram video from inside the Lusail Stadium in Qatar, ahead of the semi-final match between Argentina and Croatia. She also posted a story showing a ticket for the match. We looked at the seating arrangement of the stadium as well as footage from different angles from inside the stadium. We used this information to confirm the seating where the video was recorded is consistent with the gate number specified in the ticket.
\n
\n
Top: The ticket included in an Instagram story by dos Santos. Bottom: A still from dos Santos’ Instagram video that she posted on December 13.
\n
If she did make the trip, Qatari authorities did not seek to enact the Red Notice.
\n
Dos Santos’ social media content soon appeared to show her back in Dubai. However, this time in a more secluded location.
\n
Hitting the Gym, Again and Again
\n
A common theme in many of dos Santos’ social media posts, before and after the Red Notice was issued, is her gym workouts. On December 14, she posted an instagram story where she can be seen working out at a gym and another story at a restaurant.
\n
We were able to match features of the gym complex she posted from to the Bulgari Residences gym. This fits with her known and widely reported ownership of a residence within the Bulgari Resort and Residences complex in Dubai, as well as previous posts from the gym going back to April, 2021.
\n
The below image was posted on dos Santos Instagram shows distinctive fittings in the gym.
\n
\n
Image from dos Santos’ Instagram. Matching fixtures highlighted.
\n
These matched up perfectly with fittings of the gym in this property advertisement for the Bulgari Residences.
\n
Similarly, the window frames seen in the below image – that was posted by dos Santos’ coach on 9 January 2023- matched up with pictures listed in this property listing for the residences.
\n
\n
Matching patterns in the window frames can be seen in this Instagram post.
\n
A third workout location could be geolocated to what appears to be the Bulgari Resort and Residence squash courts, which are only accessible to residents.
\n
\n
Top: Image of dos Santos shared on her Instagram in June 2021. Bottom Left: A screenshot of a post shared by dos Santos’ coach on 9 January, 2022, showing the same pattern. Bottom Right: A screenshot from a post by dos Santos on her Instagram from the Bulgari Residences’ Squash Court shared on 4 September, 2022.
\n
As can be seen in the montage above, the same patterns visible on the premium property website are visible in the reflection of the squash court walls on dos Santos’ Instagram post.
\n
Since the Red Notice was issued, dos Santos has posted photos from the Bulgari Residences’ gym areas on at least a dozen occasions.
\n
The Pink Shirt
\n
Apart from clearly spending her time in Dubai, was dos Santos, in her own words, “at home”? The Bulgari resort in Dubai offers hotel accommodation as well as permanent residences after all.
\n
The day of the DW interview, November 23, where dos Santos made the claim, she also shared a curious selfie wearing a pink blouse with a blue peter pan collar and wired earbuds.
\n\n
\n
\nLeft: A screenshot of dos Santos’ interview with DW Africa. Source: DW.com Right: A selfie posted by dos Santos on instagram on November 23. \n
In dos Santos’ selfie, a number of buildings are visible in the reflection on a window behind her – as well as some furniture. As this is a reflection within a selfie, there are two mirrorings going on. What we therefore see is actually the non-mirrored view.
\n
Going back to property listings from Bulgari, we found this listing containing matching building shapes, from a slightly offset perspective.
\n
\n
A composite image featuring dos Santos’ Instagram selfie and parts of an image from a Dubai property listing showing the same view and objects in the distance. Property listing image source: Driven Properties.
\n
As can be seen in the montage above, a building across the water matches with what is visible in the reflection. This is the distinctive Bulgari Hotel building, which is situated opposite the site’s residences. Other buildings that can be seen in the distance also line up with this location. It thus seems highly likely dos Santos was indeed speaking from her home in Dubai’s Bulgari Resort and Residences when speaking with DW on November 23.
\n\n
\n
\nLeft: A screenshot from dos Santos’ Instagram posted on 23 May, 2021.Right: A screenshot from DW Africa interview. Source: DW.com\n\n
Further corroboration can be found in the rattan chair that can just be made out behind dos Santos in both the interview with DW and the selfie – these match the chairs visible in images from across the Bulgari Resort and Residences in Dubai. And the semi-transparent curtains match photos posted by dos Santos in previous years.
\n\n
Timeline Since Red Notice Issued
\n
Using open sources, we have systematically geolocated Isabel dos Santos to Dubai since 23 November 2022. Since that time there have been a handful of days we were not able to geolocate her and one period of four-days where she did not post at all.
Dos Santos has been consistently posting content from Dubai since November 3, but it should be noted that we only actively started examining her Instagram stories from 10 December, 2022 onwards; from then on the number of unlocatable days dropped considerably. So it is theoretically possible that dos Santos made other international trips- during the four day period she did not post, or perhaps a shorter one. Open source evidence suggests that since 23 November, 2022, dos Santos’ has been living in the Gulf, and there is reason to believe she has made at least one trip outside Dubai, to visit neighbouring Qatar for a World Cup match.
\n
So, Why is Isabel dos Santos Still Online?
\n
In response to Bellingcat’s questions about dos Santos’ presence in Dubai, the chief of staff for Angola’s Attorney General told us that they have been interacting with both Interpol and authorities in the UAE. Isabel dos Santos was contacted to comment on Bellingcat’s findings but Bellingcat had not received a response at the time of publication. The Dubai Police and UAE Interior Ministry were also contacted and acknowledged our inquiries but had not responded at time of publication.
\n
Dos Santos – who is reported to be a dual Russian-Angolan national – maintains links to several European capitals through business, property and family connections. Last month Portuguese police searched the offices of key former advisors at the request of Angolan authorities following a court-ordered seizure of assets worth around $1 billion. In March this year an Amsterdam Court will hear from parties involved in a civil case brought against dos Santos in the Netherlands. Despite this, dos Santos did not rule out a future presidential bid in her recent DW interview and just one day before Interpol confirmed the Red Notice she signed a contract with a lobbyist whose remit reportedly includes campaigning against US sanctions.
\n
The UAE has been involved in a number of high profile extraditions in recent months and has been negotiating several new extradition treaties. However, no current extradition treaty exists between the UAE and Angola. Angola’s President João Lourenço signed a number of bilateral deals during a visit to Abu Dhabi in January this year, it’s not known if the whereabouts of dos Santos were discussed.
\n
For now, dos Santos’ world appears to have become smaller even though she continues to share posts of her life in the UAE, a member of Interpol and the home country of Interpol’s president until 2025.
Special thanks to Micael Pereira (Expresso) and Luís Garriapa (SIC Notícias) for theircontribution to this report.
\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter here.
\n
",
+ "content_text": "On 24 November 2022 DW Africa published an\ninterview with Isabel dos Santos\n– once Africa’s richest woman and the daughter of former Angolan President José dos Santos. She wore a distinctive pink blouse and sat in front of a nondescript, semi-transparent curtain.\nTowards the end of the interview dos Santos was asked “Where are you currently at?” to which she replied, after a brief hesitation, “I’m at home”.\nShe repeated this answer to all further queries about her location.\nIt was not entirely clear where home was – dos Santos has multiple properties including in\nLondon, Lisbon and Dubai\n– recently, she’s been\nreported\nto spend much of her time in Dubai and she\ntold\nThe New York Times in November that she was living in London.\nHer whereabouts were of significance at the time of the DW interview as Interpol had\nreportedly\nissued a\nRed Notice\nrequesting global law enforcement locate and provisionally arrest her. Interpol\nconfirmed this\na week later.\nCollage illustration by Klawe Rzecy. Elements from Pixabay and Wikimedia Commons.\nDos Santos is the central figure in the Luanda Leaks – a trove of leaked documents – that\nexposed decades of confidential business dealings\nand resulted in\naccusations of embezzlement and fraud\nby Angolan prosecutors, as well as the freezing of more than a billion dollars worth in assets worldwide. Dos Santos maintains she made her money through legitimate business endeavours and denies accusations of fraud and embezzlement. She has described the Luanda Leaks as “\nwithout substance\n” and described the government’s charges as a political “witch hunt”.\nAt time of publication on 03 February 2023, dos Santos was yet to be provisionally detained.\nBut she is by no means in hiding. Dos Santos continues to regularly post on her social media channels, detailing visits to high-end restaurants and nights out – something\nshe alluded to\nin her DW interview when she insisted she was not disguising her whereabouts.\nAs Bellingcat discovered, those posts reveal very precise information about Isabel dos Santos’ movements over the past three months and show which of her properties she now appears to call home.\nDining and Dancing\nWhen speaking to DW, Isabel Dos Santos denied she was in hiding,\nstating\n: “I’m on social media every day. My whereabouts and place of residence are known”.\nOne look at Dos Santos’ social media accounts confirms that she is indeed online most days. Dos Santos maintains verified\nInstagram\nand\nFacebook\naccounts. She has not posted on Facebook since August of last year, but posts regularly on Instagram. She also appears to maintain\nTwitter\nand\nTikTok\naccounts. Although not verified, we observed on several occasions posts appearing on the TikTok account a few minutes before they appeared on her verified Instagram account, indicating likely ownership by dos Santos.\nAmong other things, what these accounts show is that dos Santos continues to enjoy a life of luxury.\nHere’s a post from\nTikTok\ndated December 4, geolocated to the pool of Dubai’s\nBulgari Yacht Club\n.\nStill from a TikTok featuring Isabel dos Santos, posted on 4 December 2022, that we geolocated to Dubai’s Bulgari Yacht Club, after matching it with\nthis image\nfrom Tripadvisor.\nImagery on Tripadvisor shows a\n2019 photo\nwith a near identical perspective to the video from 2022. We also checked\nwhere the sun would have been\nat that location on the day the video was posted and found the sun azimuth is consistent with the day and hour of post – within a range of five minutes.\nIn another instance, dos Santos’ was tagged by a friend in an Instagram post on 27 December 2022, where she can be seen enjoying a meal at Nusr-Et Steakhouse Dubai. She briefly featured in another friend’s post from the restaurant that day and\nreviews\nof the\nrestaurant\npoint to Nusret Gökçe – also known as Salt Bae – presence at his Dubai outlet in the final week of December 2022.\nLeft: A screenshot from an Instagram story featuring Isabel dos Santos. Right: A screenshot of Nusret Gökçe also known as Salt Bae, from the same Instagram story.\nIn this picture posted on another friend’s instagram on 8 January, 2023. Dos Santos can be seen at the left of the frame, under a brightly lit roof. This matches the red and purple corridor feature at\nDubai restaurant Trove\nwithin the Dubai Mall complex which can be seen\nhere\n.\nIsabel dos Santos can be seen in the corner of the frame in an Instagram post she was tagged in. The red and purple corridor behind her matches\nthis imag\ne from Trove, a restaurant in the Dubai Mall complex.\nA Possible Sardinia Visit?\nIt is worth noting that dos Santos’ online sharing patterns are not without noise, as she often reposts older content alongside more contemporary images.\nFor example, on 7 December 2022,\na Tiktok\nappeared from a beach where the landscape seemed vastly different from Dubai. It also featured Spanish subtitles. Could she have left the UAE? We used\nPeakVisor\nto verify this location and found it to be in\nSardinia, Italy\n. Dos Santos also posted from this location back in September 2022, but has since deleted the older post.\nA Mountain silhouette created using PeakVisor laid over a stitch of a TikTok video featuring dos Santos’ at Sardinian Beach Li Itriceddi.\nA visit in September when temperatures reached an average high of 27°C would likely make more sense than a visit in December when the average high is 15.4°C, especially as people can be seen bathing in the water behind dos Santos.\nBut another clue comes from a major global event, and points to one potential international visit that dos Santos apparently made after the Interpol Red Notice was issued.\nThe World Cup\nOn 5 December\ndos Santos’ posted\na picture of herself in a Brazil football shirt with a big screen visible behind her. We\ngeolocated this picture\nto the official FIFA\nfan zone\nset up in Dubai for the 2022 World Cup. Dubai’s skyline can be seen in the background.\nSouth Korea played Brazil\nin the round of 16 that evening and the Brazil manager, Tite, can also be seen on the screen behind dos Santos.\nDos Santos was in Dubai on the evening of December 5 for the match which kicked off at\n11pm local time\n. She was also watching football in Dubai on\nDecember 10\nand posted from Dubai on\nDecember 7\n– the same day as the Sardinia post. Apart from the single post of the Beach Li Itriceddi there is no other evidence to suggest that dos Santos’ visited Sardinia around this time.\nAn image of dos Santos at a FIFA fan zone during the South Korea versus Brazil World Cup Match,\nposted\non her Instagram account on December 5, 2022.\nThis wasn’t the last time dos Santos posted about the 2022 World Cup. She also appeared to post about the tournament from outside Dubai.\nOn 13 December dos Santos shared a now removed Instagram video from inside the Lusail Stadium in Qatar, ahead of the semi-final match between Argentina and Croatia. She also posted a story showing a ticket for the match. We looked at the\nseating arrangement\nof the stadium as well as\nfootage\nfrom different angles from inside the stadium. We used this information to confirm the seating where the video was recorded is consistent with the gate number specified in the ticket.\nTop: The ticket included in an Instagram story by dos Santos. Bottom: A still from dos Santos’ Instagram video that she posted on December 13.\nIf she did make the trip, Qatari authorities did not seek to enact the Red Notice.\nDos Santos’ social media content soon appeared to show her back in Dubai. However, this time in a more secluded location.\nHitting the Gym, Again and Again\nA common theme in many of dos Santos’ social media posts, before and after the Red Notice was issued, is her gym workouts. On December 14, she posted an instagram story where she can be seen working out at a gym and another story at a restaurant.\nWe were able to match features of the gym complex she posted from to the\nBulgari Residences gym\n. This fits with her\nknown and widely reported ownership\nof a residence within the Bulgari Resort and Residences complex in Dubai, as well as previous posts from the gym going back to\nApril, 2021\n.\nThe below image was posted on dos Santos Instagram shows distinctive fittings in the gym.\nImage from dos Santos’ Instagram. Matching fixtures highlighted.\nThese matched up perfectly with fittings of the gym in\nthis\nproperty advertisement for the Bulgari Residences.\nSimilarly, the window frames seen in the below image – that was posted by dos Santos’ coach on 9 January 2023- matched up with pictures listed in\nthis\nproperty listing for the residences.\nMatching patterns in the window frames can be seen in this Instagram post.\nA third workout location could be geolocated to what appears to be the Bulgari Resort and Residence squash courts, which are only accessible to residents.\nTop: Image of dos Santos shared on her Instagram in June 2021. Bottom Left: A screenshot of a post shared by dos Santos’ coach on 9 January, 2022, showing the same pattern. Bottom Right: A screenshot from a post by dos Santos on her Instagram from the Bulgari Residences’ Squash Court shared on 4 September, 2022.\nAs can be seen in the montage above, the same patterns visible on the premium property website are visible in the reflection of the squash court walls on dos Santos’ Instagram post.\nSince the Red Notice was issued, dos Santos has posted photos from the Bulgari Residences’ gym areas on at least a dozen occasions.\nThe Pink Shirt\nApart from clearly spending her time in Dubai, was dos Santos, in her own words, “at home”? The Bulgari resort in Dubai offers hotel accommodation as well as permanent residences after all.\nThe day of the DW interview, November 23, where dos Santos made the claim, she also shared a\ncurious selfie\nwearing a pink blouse with a blue peter pan collar and wired earbuds.\nLeft: A screenshot of dos Santos’ interview with DW Africa. Source:\nDW.com\nRight: A selfie posted by dos Santos on instagram on November 23.\nIn dos Santos’ selfie, a number of buildings are visible in the reflection on a window behind her – as well as some furniture. As this is a reflection within a selfie, there are two mirrorings going on. What we therefore see is actually the non-mirrored view.\nGoing back to property listings from Bulgari, we found\nthis listing\ncontaining matching building shapes, from a slightly offset perspective.\nA composite image featuring dos Santos’ Instagram selfie and parts of an image from a Dubai property listing showing the same view and objects in the distance. Property listing image source: Driven Properties.\nAs can be seen in the montage above, a building across the water matches with what is visible in the reflection. This is the distinctive Bulgari Hotel building, which is situated opposite the site’s residences. Other buildings that can be seen in the distance also line up with this location. It thus seems highly likely dos Santos was indeed speaking from her home in Dubai’s Bulgari Resort and Residences when speaking with DW on November 23.\nLeft: A screenshot from dos Santos’ Instagram posted on 23 May, 2021.Right: A screenshot from DW Africa interview. Source:\nDW.com\nFurther corroboration can be found in the rattan chair that can just be made out behind dos Santos in both the interview with DW and the selfie – these match the chairs visible in images from across the\nBulgari Resort and Residences in Dubai\n. And the semi-transparent curtains match photos posted by\ndos Santos\nin previous years.\nTimeline Since Red Notice Issued\nUsing open sources, we have systematically geolocated Isabel dos Santos to Dubai since 23 November 2022. Since that time there have been a handful of days we were not able to geolocate her and one period of four-days where she did not post at all.\nDos Santos has been consistently posting content from Dubai since November 3, but it should be noted that we only actively started examining her Instagram stories from 10 December, 2022 onwards; from then on the number of unlocatable days dropped considerably. So it is theoretically possible that dos Santos made other international trips- during the four day period she did not post, or perhaps a shorter one. Open source evidence suggests that since 23 November, 2022, dos Santos’ has been living in the Gulf, and there is reason to believe she has made at least one trip outside Dubai, to visit neighbouring Qatar for a World Cup match.\nSo, Why is Isabel dos Santos Still Online?\nIn response to Bellingcat’s questions about dos Santos’ presence in Dubai, the chief of staff for Angola’s Attorney General told us that they have been interacting with both Interpol and authorities in the UAE. Isabel dos Santos was contacted to comment on Bellingcat’s findings but Bellingcat had not received a response at the time of publication. The Dubai Police and UAE Interior Ministry were also contacted and acknowledged our inquiries but had not responded at time of publication.\nDos Santos – who is\nreported\nto be a dual Russian-Angolan national – maintains links to several European capitals through business, property and family connections. Last month\nPortuguese police searched the offices\nof key former advisors at the request of Angolan authorities following a court-ordered seizure of assets\nworth around $1 billion\n. In March this year an Amsterdam Court will hear from parties involved in a civil case brought against dos Santos\nin the Netherlands\n. Despite this, dos Santos did not rule out a future\npresidential bid\nin her recent DW interview and just one day before Interpol confirmed the Red Notice she\nsigned a contract\nwith a lobbyist whose remit\nreportedly includes\ncampaigning against US sanctions.\nThe UAE has been involved in a number of\nhigh profile\nextraditions in recent months and has been negotiating several\nnew extradition treaties\n. However, no current extradition treaty exists between the UAE and Angola. Angola’s President João Lourenço signed a number of\nbilateral deals\nduring a visit to\nAbu Dhabi in January this year, it’s not known if the whereabouts of dos Santos were discussed.\nFor now, dos Santos’ world appears to have become smaller even though she continues to share posts of her life in the UAE, a member of Interpol and the home country of Interpol’s president until 2025.\n_________________________________________________________________________________________\nSpecial thanks to Micael Pereira (Expresso) and Luís Garriapa (SIC Notícias) for their\ncontribution to this report\n.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/LUCY_BELLINGCAT-final-1200x977.jpg",
+ "alt": "Illustration of pink blouse, map of Dubai and images of palm trees and pools.",
+ "caption": "Collage illustration by Klawe Rzecy. Elements from Pixabay and Wikimedia Commons."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/output_0006.png",
+ "alt": "Image of Isabel dos Santos besides a pool at a resort.",
+ "caption": "Still from a TikTok featuring Isabel dos Santos, posted on 4 December 2022, that we geolocated to Dubai’s Bulgari Yacht Club, after matching it with this image from Tripadvisor."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/dos-Santos_SaltBae-150x150.png",
+ "alt": "Image of Isabel dos Santos eating steak.",
+ "caption": "Left: A screenshot from an Instagram story featuring Isabel dos Santos. Right: A screenshot of Nusret Gökçe also known as Salt Bae, from the same Instagram story."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/SaltBae-1-e1675419192317-150x150.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/Trove-Dos-Santos.png",
+ "alt": "Isabel dos Santos seen at left-hand-side of Instagram photo under a red and purple corridor.",
+ "caption": "Isabel dos Santos can be seen in the corner of the frame in an Instagram post she was tagged in. The red and purple corridor behind her matches this imag e from Trove, a restaurant in the Dubai Mall complex."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/Sardinia-dos-Santos-edited.jpeg",
+ "alt": "An image of dos Santos at a beach in Sardinia with PeakVisor image showing mountains behind her.",
+ "caption": "A Mountain silhouette created using PeakVisor laid over a stitch of a TikTok video featuring dos Santos’ at Sardinian Beach Li Itriceddi."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/IMG_5257-738x1200.jpg",
+ "alt": "",
+ "caption": "An image of dos Santos at a FIFA fan zone during the South Korea versus Brazil World Cup Match, posted on her Instagram account on December 5, 2022."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/TicketQatar.png",
+ "alt": "Composite image showing a ticket for a World Cup match and still from footage inside the stadium.",
+ "caption": "Top: The ticket included in an Instagram story by dos Santos. Bottom: A still from dos Santos’ Instagram video that she posted on December 13."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/Gym1.png",
+ "alt": "Image of a gym with light fittings highlighted.",
+ "caption": "Image from dos Santos’ Instagram. Matching fixtures highlighted."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/vis-feaxtures-257x300.jpeg",
+ "alt": "Image of window frame showing matching patterns.",
+ "caption": "Matching patterns in the window frames can be seen in this Instagram post."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/Gym3.png",
+ "alt": "Composite image showing dos Santos outside the gym, with distinctive pattern on bar windows and separate image shows reflection of the pattern in a selfie.",
+ "caption": "Top: Image of dos Santos shared on her Instagram in June 2021. Bottom Left: A screenshot of a post shared by dos Santos’ coach on 9 January, 2022, showing the same pattern. Bottom Right: A screenshot from a post by dos Santos on her Instagram from the Bulgari Residences’ Squash Court shared on 4 September, 2022."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/DW-Interview.png",
+ "alt": "",
+ "caption": "Left: A screenshot of dos Santos’ interview with DW Africa. Source: DW.com Right: A selfie posted by dos Santos on instagram on November 23."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/Pink-Shirt-Selfie.jpeg",
+ "alt": "Two pictures of dos Santos wearing a matching outfit of pink blouse. One from a still from a DW Africa interview the other a Selfie from Instagram.",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/Composite-selfie.png",
+ "alt": "Composite image illustrating how the reflection seen behind dos Santos in her selfie matches a view from a property in the Bulgari residences.",
+ "caption": "A composite image featuring dos Santos’ Instagram selfie and parts of an image from a Dubai property listing showing the same view and objects in the distance. Property listing image source: Driven Properties."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/Oldcurtainpic-1.jpeg",
+ "alt": "",
+ "caption": "Left: A screenshot from dos Santos’ Instagram posted on 23 May, 2021.Right: A screenshot from DW Africa interview. Source: DW.com"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/02/furtinure-dw-1200x741.jpeg",
+ "alt": "Old image of dos Santos in front of the same curtain and image of furniture behind dos Santos.",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2023-05-03_international-far-right-fight-night-comes-to-budapest.json b/bellingcat_geolocation_json/2023-05-03_international-far-right-fight-night-comes-to-budapest.json
new file mode 100644
index 0000000000000000000000000000000000000000..fc196a492ad5dcd29cb8de068e6dbcfe8cebca33
--- /dev/null
+++ b/bellingcat_geolocation_json/2023-05-03_international-far-right-fight-night-comes-to-budapest.json
@@ -0,0 +1,32 @@
+{
+ "slug": "international-far-right-fight-night-comes-to-budapest",
+ "title": "International Far-Right Fight Night Comes to Budapest",
+ "url": "https://www.bellingcat.com/news/2023/05/03/international-far-right-fight-night-comes-to-budapest/",
+ "date": "2023-05-03T13:49:18+00:00",
+ "modified": "2026-02-11T17:49:14+00:00",
+ "author": "Michael Colborne",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Banned in Germany, a far-right combat sports event will take place in Budapest with fighters and supporters expected to attend from across Europe.",
+ "content_html": "
\n
This weekend, Budapest is due to play host to an event that far-right extremists have bragged online will be “the biggest radical nationalist event since coronavirus.”
\n
Scheduled for 6 May, the organisers of ‘European Fight Night’ boast that the event in the Hungarian capital will feature up to 15 combat sports fights with participants from 12 different countries.
\n
According to social media posts promoting the event, it will also play host to concerts and opportunities to buy merchandise from far-right fashion brands. Yet despite the publicity push, secrecy remains a priority. Similar events in Germany have previously been banned. Organisers have also stated attendees won’t be allowed to take photos or videos, and that they’ll kick out anyone who does or who shares the exact location of the event.
\n
Observers have long sounded the alarm on the international far right’s use of combat sports to train up for physical confrontation and prepare for attacks against their perceived foes.
Some experts like Robert Claus, who wrote a book on the far-right martial arts scene, even say that combat sports events like European Fight Night also help facilitate international networking and recruitment.
\n
Past research by anti-fascist groups has linked the organisers of European Fight Night to some of the most secretive far-right extremist groups in Europe — including violent neo-Nazi groups that have been banned in several countries. One of the organisers of European Fight Night has even discussed links to local chapters of these banned groups in previous media interviews.
\n
Meet the Organisers
\n
European Fight Night, at least publicly, presents itself as having three primary organisers: Hungary’s Legio Hungaria, Germany’s Kampf der Nibelungen (“Battle of the Nibelungs”), and Pride France.
\n
In a video released on multiple Telegram channels on February 27, members of each of the three main organising groups appeared side by side to promote the Budapest event and encourage their far-right comrades across Europe to attend.
\n
A screenshot of a February 2023 promotional video for European Fight Night, featuring Legio Hungaria’s Béla Incze (left), Kampf der Nibelungen’s Alexander Deptolla (centre) and Pride France’s Tomasz Szkatulski (right). The video was filmed in Sofia, Bulgaria: a far-right extremist march organised by the Bulgarian National Union took place two days before the video’s release, and this video was filmed near their headquarters.\n
Béla Incze heads up Legio Hungaria, a small far-right extremist group from Hungary. Incze once served as an assistant to a far-right member of Hungary’s parliament before reportedly losing his job for assaulting a police officer. Legio Hungaria, while not a large group, has managed to garner attention for their actions since being founded in 2018, including vandalising a Jewish community centre in 2019, destroying a Black Lives Matter statue in Budapest in 2021 and reportedly assaulting journalists covering a far-right event in 2023. Bellingcat also revealed in a 2021 investigation that a senior Legio Hungaria member played a key role in racist and homophobic displays at Hungarian national team football matches.
\n
Hungary fans display a sign during the EURO 2020 match between Germany and Hungary. Bellingcat later revealed the banner had been in the possession of a member of Legio Hungaria prior to the match. Pool via REUTERS/Kai Pfaffenbach.\n
Alexander Deptolla is the public face of Kampf der Nibelungen. Started in 2013, Kampf der Nibelungen — often referred to by the abbreviation KdN — was once the largest far-right combat sports event in Europe, with its October 2018 version hosting an estimated 850 fighters and spectators. However, authorities in Germany banned the event in 2019, and KdN’s 2020 ‘online’ version saw fewer than half of the planned fights broadcast thanks to a police raid where some were being filmed ahead of time. Deptolla even took to social media afterwards to state that KdN wouldn’t be organising any more events in the immediate future. Nonetheless, German authorities reportedly restricted an abortive attempt at a September 2021 ‘National Fight Night,’ as a September 2022 court ruling upheld the ban on KdN’s events. Despite this, they are still the biggest of the three organisers of European Fight Night. Some German observers have argued that European Fight Night is essentially planned as a “replacement” event for KdN outside German borders.
\n
Alexander Deptolla (left) interviewing French far-right lawyer Pierre-Marie Bonneau (right) at a far-right extremist event in Sofia, Bulgaria on February 25, 2023 (Michael Colborne).\n
Tomasz Szkatulski of Pride France is a Polish-born French neo-Nazi who is reported to have helped organise and even participate in events like European Fight Night for several years. Szkatulski emerged from violent football hooligan scenes in the city of Lille, though now appears to spend much of his time in Bulgaria. Covered in Nazi tattoos, Szkatulski was reportedly convicted and sentenced to a year in prison in 2008 for assaulting a man of African origin with a bicycle chain; he has also reportedly been involved in other acts of violence, including alleged attacks on LGBT+ establishments. Szkatulski is also an associate of American far-right extremist Robert Rundo, who was arrested in Romania in March 2023 and is awaiting extradition to the United States.
\n
Szkatulski did not respond to requests for comment from Bellingcat asking about his involvement in European Fight Night, past reports about his 2008 conviction (or alleged involvement in other apparent acts of violence), as well as his relationship and contacts with Rob Rundo.
\n
When contacted by Bellingcat on Telegram to ask if he would like to respond to the details in this story, Deptolla of KdN said simply: “Hi. No thanks. Bye.” Legio Hungaria similarly declined to answer Bellingcat’s emailed questions about the details contained in this story and about Incze, citing unhappiness with Bellingcat’s reporter’s previous reporting on the group.
\n
Behind the Scenes
\n
There’s also a much less public side to European Fight Night and other far-right combat sports events, where some of the most secretive, publicity-averse extremist networks in Europe operate.
\n
Germany’s Federal Office for the Protection of the Constitution (BfV), the country’s domestic intelligence agency, has publicly stated the members of the Hammerskins, an international white supremacist gang, have been involved in organising Kampf der Nibelungen events since 2013. Journalists and researchers have also extensively documented the Hammerskins’ links with KdN.
\n
As noted in Bellingcat’s 2022 investigation of a planned far-right extremist concert in Belgium, the Hammerskins have a significant presence acrossEurope. They tend to be secretive about their activities but are organised, with chapters, hierarchies and different patches for prospects and full members. Members of the Hammerskins have been extensively involved in criminal activities, including assault and even mass murder.
\n
But it isn’t just Kampf der Nibelungen that works with the Hammerskins. In a February 2023 interview with an American far-right extremist website, Incze stated that his Legio Hungaria organisation had “a strong relationship” with the Hungarian Hammerskins, evidenced by their appearance and central role in previous events Legio Hungaria has hosted.
\n
Incze also said in the same February 2023 interview that Legio Hungaria cooperates with the Hungarian branch of international neo-Nazi organisation Combat 18. Founded in the UK in the early 1990s, Combat 18 is closely associated with the Blood and Honour neo-Nazi group that is banned in several European countries.
\n
In his 2020 book on Europe’s far-right combat sports scenes, German author Robert Claus reported that Tomasz Szkatulski had been affiliated with Blood and Honour structures in France; his Blood and Honour affiliations have also been discussed in otherreports, although Szkatulski did not respond to Bellingcat questions asking about this. Claus also told Bellingcat he had knowledge that Blood and Honour associates would be hosting a concert the evening of 6 May for European Fight Night attendees.
\n
Bellingcat was not able to independently confirm this claim. However, Blood and Honour’s Hungary branch has also promoted European Fight Night on its Telegram channel.
\n
“These networks are generally very much involved in [far-right combat sports] events, either by hosting them themselves or by providing important organisational aspects,” Claus told Bellingcat, referring to the Hammerskins, Combat 18 and Blood and Honour.
\n
A ‘Fundamentally Violent’ Ideology
\n
But barely a week ahead of the event, European Fight Night’s main organiser had already run into trouble. On April 28 Deptolla took to KdN’s Telegram channel to tell fans that he and other German far-right extremists had been banned by authorities from leaving the country to travel to Budapest. Claus told Bellingcat that this German travel ban could seriously impact the event, since KdN and its German comrades are the largest of the three main organisers.
\n
Even if European Fight Night doesn’t go on quite as its organiszers have hoped, Claus says it is still a dangerous event. “Violent key figures and groups of the neo-Nazi scenes from different parts in Europe are expected to come to Budapest,” Claus told Bellingcat. “Their ideology is fundamentally violent, fascist activists [who] use combat sports to train for racist attacks and other kinds of brute force.”
\n
With a concert and other events alongside the fights, Claus warned, “the international neo-Nazi scene gets its combination of violence and sports, political hatred and music, the cultural package which pre-terrorist neo-Nazism has to offer.”
\n
Bellingcat asked Hungarian national police about the European Fight Night event, whether they were aware of it and whether they had taken or planned to take any steps to prevent the event from taking place, or prevent any specific foreign extremists from attending. They did not respond before publication.
\n\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter hereand Mastodon here.
",
+ "content_text": "This weekend, Budapest is due to play host to an event that far-right extremists have bragged online will be “the biggest radical nationalist event since coronavirus.”\nScheduled for 6 May, the organisers of ‘European Fight Night’ boast that the event in the Hungarian capital will feature up to 15 combat sports fights with participants from 12 different countries.\nAccording to social media posts promoting the event, it will also play host to concerts and opportunities to buy merchandise from far-right fashion brands. Yet despite the publicity push, secrecy remains a priority. Similar events in Germany have previously been banned. Organisers have also stated attendees won’t be allowed to take photos or videos, and that they’ll kick out anyone who does or who shares the exact location of the event.\nObservers have long\nsounded\nthe alarm on the international far right’s use of\ncombat sports\nto train up for physical confrontation and prepare for attacks against their perceived foes.\nSome experts like Robert Claus, who wrote a\nbook\non the far-right martial arts scene, even say that combat sports events like European Fight Night also help\nfacilitate\ninternational networking and recruitment.\nPast research by anti-fascist groups has linked the organisers of European Fight Night to some of the most\nsecretive\nfar-right extremist groups in Europe — including\nviolent\nneo-Nazi groups that have been\nbanned\nin several countries. One of the organisers of European Fight Night has even discussed links to local chapters of these banned groups in previous media interviews.\nMeet the Organisers\nEuropean Fight Night, at least publicly, presents itself as having three primary organisers: Hungary’s Legio Hungaria, Germany’s\nKampf der Nibelungen\n(“Battle of the\nNibelungs\n”), and Pride France.\nIn a video released on multiple Telegram channels on February 27, members of each of the three main organising groups appeared side by side to promote the Budapest event and encourage their far-right comrades across Europe to attend.\nA screenshot of a February 2023 promotional video for European Fight Night, featuring Legio Hungaria’s Béla Incze (left), Kampf der Nibelungen’s Alexander Deptolla (centre) and Pride France’s Tomasz Szkatulski (right). The video was filmed in Sofia, Bulgaria: a far-right extremist march organised by the Bulgarian National Union took place two days before the video’s release, and this video was filmed\nnear\ntheir headquarters.\nBéla Incze heads up Legio Hungaria, a small far-right extremist group from Hungary. Incze once served as an assistant to a far-right member of Hungary’s parliament before reportedly losing his job for\nassaulting\na police officer. Legio Hungaria, while not a large group, has managed to garner attention for their actions since being founded in 2018, including\nvandalising\na Jewish community centre in 2019,\ndestroying\na Black Lives Matter statue in Budapest in 2021 and reportedly\nassaulting\njournalists covering a far-right event in 2023. Bellingcat also revealed in a 2021\ninvestigation\nthat a senior Legio Hungaria member played a key role in racist and homophobic displays at Hungarian national team football matches.\nHungary fans display a sign during the EURO 2020 match between Germany and Hungary. Bellingcat later\nrevealed\nthe banner had been in the possession of a member of Legio Hungaria prior to the match. Pool via REUTERS/Kai Pfaffenbach.\nAlexander Deptolla is the public face of Kampf der Nibelungen. Started in 2013, Kampf der Nibelungen — often referred to by the abbreviation KdN — was once the largest far-right combat sports event in Europe, with its October 2018 version hosting an estimated 850 fighters and spectators. However, authorities in Germany banned the event in 2019, and KdN’s 2020\n‘online’\nversion\nsaw\nfewer than half of the planned fights\nbroadcast\nthanks to a police raid where some were being filmed ahead of time. Deptolla even took to social media afterwards to state that KdN wouldn’t be organising any more events in the immediate future. Nonetheless, German\nauthorities\nreportedly restricted an abortive\nattempt\nat a September 2021 ‘National Fight Night,’ as a September 2022 court ruling\nupheld\nthe ban on KdN’s events. Despite this, they are still the biggest of the three organisers of European Fight Night. Some German\nobservers\nhave argued that European Fight Night is essentially planned as a “replacement” event for KdN outside German borders.\nAlexander Deptolla (left) interviewing French far-right lawyer Pierre-Marie Bonneau (right) at a far-right extremist event in Sofia, Bulgaria on February 25, 2023 (Michael Colborne)\n.\nTomasz Szkatulski of Pride France is a Polish-born French neo-Nazi who is reported to have helped\norganise\nand even\nparticipate\nin events like European Fight Night for several years. Szkatulski emerged from violent football hooligan scenes in the city of\nLille\n, though now appears to spend much of his time in\nBulgaria\n.\nCovered in Nazi tattoos\n, Szkatulski was\nreportedly\nconvicted and sentenced to a year in prison in 2008 for assaulting a man of African origin with a bicycle chain; he has also\nreportedly\nbeen involved in other acts of violence, including alleged attacks on LGBT+ establishments. Szkatulski is also an\nassociate\nof American far-right extremist Robert Rundo, who was\narrested\nin Romania in March 2023 and is awaiting extradition to the United States.\nSzkatulski did not respond to requests for comment from Bellingcat asking about his involvement in European Fight Night, past reports about his 2008 conviction (or alleged involvement in other apparent acts of violence), as well as his relationship and contacts with Rob Rundo.\nWhen contacted by Bellingcat on Telegram to ask if he would like to respond to the details in this story, Deptolla of KdN said simply: “Hi. No thanks. Bye.” Legio Hungaria similarly declined to answer Bellingcat’s emailed questions about the details contained in this story and about Incze, citing unhappiness with Bellingcat’s reporter’s previous reporting on the group.\nBehind the Scenes\nThere’s also a much less public side to European Fight Night and other far-right combat sports events, where some of the most secretive, publicity-averse extremist networks in Europe operate.\nGermany’s Federal Office for the Protection of the Constitution (BfV), the country’s domestic intelligence agency, has publicly\nstated\nthe members of the\nHammerskins\n, an international white supremacist gang, have been involved in organising Kampf der Nibelungen events since 2013. Journalists and researchers have also extensively\ndocumented\nthe Hammerskins’ links with KdN.\nAs noted in Bellingcat’s 2022\ninvestigation\nof a planned far-right extremist concert in Belgium, the Hammerskins have a significant presence\nacross\nEurope\n. They tend to be\nsecretive\nabout their activities but are organised, with chapters, hierarchies and different patches for prospects and full members. Members of the Hammerskins have been extensively involved in criminal activities, including\nassault\nand even\nmass murder\n.\nBut it isn’t just Kampf der Nibelungen that works with the Hammerskins. In a February 2023 interview with an American far-right extremist website, Incze stated that his Legio Hungaria organisation had “a strong relationship” with the Hungarian Hammerskins, evidenced by their appearance and central role in\nprevious events\nLegio Hungaria has hosted.\nIncze also said in the same February 2023 interview that Legio Hungaria cooperates with the Hungarian branch of international neo-Nazi organisation\nCombat 18\n. Founded in the UK in the early 1990s, Combat 18 is closely associated with the\nBlood and Honour\nneo-Nazi group that is banned in several European countries.\nIn his 2020\nbook\non Europe’s far-right combat sports scenes, German author Robert Claus reported that Tomasz Szkatulski had been affiliated with Blood and Honour structures in France; his Blood and Honour affiliations have also been discussed in\nother\nreports\n, although Szkatulski did not respond to Bellingcat questions asking about this. Claus also told Bellingcat he had knowledge that Blood and Honour associates would be hosting a concert the evening of 6 May for European Fight Night attendees.\nBellingcat was not able to independently confirm this claim. However, Blood and Honour’s Hungary branch has also promoted European Fight Night on its Telegram channel.\n“These networks are generally very much involved in [far-right combat sports] events, either by hosting them themselves or by providing important organisational aspects,” Claus told Bellingcat, referring to the Hammerskins, Combat 18 and Blood and Honour.\nA ‘Fundamentally Violent’ Ideology\nBut barely a week ahead of the event, European Fight Night’s main organiser had already run into trouble. On April 28 Deptolla took to KdN’s Telegram channel to tell fans that he and other German far-right extremists had been banned by authorities from leaving the country to travel to Budapest. Claus told Bellingcat that this German travel ban could seriously impact the event, since KdN and its German comrades are the largest of the three main organisers.\nEven if European Fight Night doesn’t go on quite as its organiszers have hoped, Claus says it is still a dangerous event. “Violent key figures and groups of the neo-Nazi scenes from different parts in Europe are expected to come to Budapest,” Claus told Bellingcat. “Their ideology is fundamentally violent, fascist activists [who] use combat sports to train for racist attacks and other kinds of brute force.”\nWith a concert and other events alongside the fights, Claus warned, “the international neo-Nazi scene gets its combination of violence and sports, political hatred and music, the cultural package which pre-terrorist neo-Nazism has to offer.”\nBellingcat asked Hungarian national police about the European Fight Night event, whether they were aware of it and whether they had taken or planned to take any steps to prevent the event from taking place, or prevent any specific foreign extremists from attending. They did not respond before publication.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\nand Mastodon\nhere\n.\nItt található a cikk magyar fordítása telexes kollégáinktól / Click here for a Hungarian translation of this article from our colleagues at Telex",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-03-at-11.44.38-1200x673.png",
+ "alt": "",
+ "caption": "A screenshot of a February 2023 promotional video for European Fight Night, featuring Legio Hungaria’s Béla Incze (left), Kampf der Nibelungen’s Alexander Deptolla (centre) and Pride France’s Tomasz Szkatulski (right). The video was filmed in Sofia, Bulgaria: a far-right extremist march organised by the Bulgarian National Union took place two days before the video’s release, and this video was filmed near their headquarters."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2021/09/2021-06-23T195324Z_582482922_UP1EH6N1J8YUH_RTRMADP_3_SOCCER-EURO-GER-HUN-REPORT-1200x795.jpg",
+ "alt": "",
+ "caption": "Hungary fans display a sign during the EURO 2020 match between Germany and Hungary. Bellingcat later revealed the banner had been in the possession of a member of Legio Hungaria prior to the match. Pool via REUTERS/Kai Pfaffenbach."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-03-at-11.49.34-1200x791.png",
+ "alt": "",
+ "caption": "Alexander Deptolla (left) interviewing French far-right lawyer Pierre-Marie Bonneau (right) at a far-right extremist event in Sofia, Bulgaria on February 25, 2023 (Michael Colborne) ."
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2023-05-08_chronolocation-determining-when-a-photo-was-taken-using-facebook-google-street-view-and-assorted-tiny-details.json b/bellingcat_geolocation_json/2023-05-08_chronolocation-determining-when-a-photo-was-taken-using-facebook-google-street-view-and-assorted-tiny-details.json
new file mode 100644
index 0000000000000000000000000000000000000000..80a542969041d849845b2a60da4176273dd0cbb0
--- /dev/null
+++ b/bellingcat_geolocation_json/2023-05-08_chronolocation-determining-when-a-photo-was-taken-using-facebook-google-street-view-and-assorted-tiny-details.json
@@ -0,0 +1,172 @@
+{
+ "slug": "chronolocation-determining-when-a-photo-was-taken-using-facebook-google-street-view-and-assorted-tiny-details",
+ "title": "Chronolocation: Determining When a Photo was Taken Using Facebook, Google Street View and Assorted Tiny Details",
+ "url": "https://www.bellingcat.com/resources/2023/05/08/chronolocation-determining-when-a-photo-was-taken-using-facebook-google-street-view-and-assorted-tiny-details/",
+ "date": "2023-05-08T13:20:03+00:00",
+ "modified": "2023-05-08T18:41:27+00:00",
+ "author": "Youri van der Weide",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Social media posts, mapping tools and a being aware of small but important details can help researchers determine when an undated image was taken.",
+ "content_html": "
\n
How can open source investigators determine when a photograph or video was taken?
But this process of “chronolocation” – determining when a picture was taken rather than just where (which is known as geolocation) – isn’t always feasible. For a start, it has to be daytime and there must be a shadow cast somewhere in the video or image of interest.
\n
If these options aren’t available, looking for other clues in a “source” image that can be cross-referenced with contextual “reference” images or other related information can also help narrow down a date range.
Storefronts (similar to buildings but might be even more useful as this changes more frequently).
\n
Seasonal indicators such as weather, foliage etc.
\n
Natural landmarks
\n
Clock towers
\n
Public transport stops, bus lines etc.
\n
Graffiti and murals
\n
Banners
\n
Advertisements
\n
\n
Essentially, any aspect of a source image could be of use, provided that it has changed over time. Sometimes clues will be so obvious that it’s possible to immediately figure out the rough date of the source image from one detail alone.
\n
This picture taken in front of CitiField, a baseball stadium in New York that serves as the home of the New York Mets, offers an example.
\n
Fans gather outside CitiField in New York (photo: Youri van der Weide).\n
Firstly, there are many features in this picture that can narrow down time and date.
\n
The most obvious is the banners on the lampposts in front of the stadium. The year is detailed as 2017.
\n
\n
A close up image of banners on display outside CitiField in New York (photo: Youri van der Weide).
\n
To determine the exact date, however, we need more information.
\n
There is a wooden structure on the left with a banner that reads, “Home is the Key” while the name of a non-profit, Habitat for Humanity, is also displayed prominently. Googling Habitat for Humanity in combination with the Mets yields only a few results.
\n
A page on Habitat’s website is one of those results. It explains that their Home is the Keycampaign launched on April 3, 2017, on the opening day of the new baseball season for the New York Mets.
\n
Text visible on a page of Habitat’s website details the “Home is the Key” campaign.\n
Another website visible in the same Google search results also reported on the topic and shared a picture of the same wooden frame that is visible in the source image. We can further confirm the date of April 3 by looking for the Mets’ 2017 season schedule, indeed the first game of the 2017 season was on April 3.
\n
It appears that a couple of key details and a quick Google search was all that was required to find a probable correct date for the image. Calling up or emailing Habitat to ask them whether this image was indeed captured on this date can provide further cast iron assurances.
\n
Narrowing Down Dates
\n
Unfortunately, chronolocation isn’t always that simple.
\n
In most cases there will be a few extra steps and it will only be possible to figure out the earliest possible date or latest possible date that an image was taken, rather than the precise date and time.
\n
Researchers will often already have a latest possible date when they start the chronolocation process. This will likely be the date an image was uploaded to the platform they found it on.
\n
Of course, it may be possible to refine the latest upload date down to an earlier time frame with some investigation.
\n
In 2018, Bellingcat published an article looking at claims that GRU officer “Ruslan Boshirov” (the cover identity for one of the men accused of poisoning former Russian spy Sergei Skripal and his daughter Yulia) made a trip to Prague back in 2014.
\n
A Facebook account using the name of Ruslan Boshirov had indeed posted from Prague on 11 October 2014.
\n
\n
A screen grab of a Facebook post by an account bearing the name of Ruslan Boshirov.
\n
Yet while it was not possible to say for sure that this indeed was the Facebook account of the GRU spy, it was possible to confirm that the image it displayed had indeed been taken in Prague between mid to late 2014.
\n
This would become relevant several years later when it became apparent that Boshirov, whose real name is Anatoliy Chepiga, had indeed flown to Prague around this time and was linked with attacks on munitions depots by Czech police.
\n
The picture clearly shows Prague’s popular Old Town square, a site that will be familiar to any tourist who has visited the Czech capital.
\n
Yet other, more subtle details allowed Bellingcat to confirm that the photograph had in fact been taken in Prague around that time.
\n
Firstly, it is not possible that the image could have been taken after 11 October, 2014, given that’s when it was posted to Facebook.
\n
But by zooming in on a cream coloured building in the background, a number of large black letters flanked by two black squares can just about be made out.
\n
\n
A number of letters can be made out on the frontage of a building in the distance of a picture uploaded by the Facebook account of an individual named Ruslan Boshirov.
\n
By looking at other images of this building, which Google Mapsshows is called the Central Gallery, that appear on search engines or that were posted on social media sites during the European summer and autumn of 2014, we can begin to piece together a timeframe of when this photo was likely taken.
\n
An image of the exterior of the Central Gallery in Prague uploaded on 17 June 2014. \n
The above image was uploaded on 17 June 2014 and shows the names of artists (Alphonse) Mucha, (Andy) Warhol and (Salvador) Dali, flanked by two black squares. This makes sense as the building has long hosted exhibits by these artists, as detailed in Trip Advisor reviews stretching back to 2014.
\n
The position of the names and black squares also matched the 11 October Facebook photo and other photos posted online around mid-June.
\n
However, photos that were uploaded in early June show a different signage on the building.
\n
\n
Images of the Central Gallery in Prague taken on 4 June and 14 June 2014.\n
Only the names of two artists appear and there are no black boxes in images uploaded on June 4 and 14.
\n
We may now be able to state that the photograph was dated sometime between mid June and 11 October when it was posted on Facebook. As with the CitiField stadium example earlier, a call to the museum could help confirm these details.
\n
Hanging out in Istanbul
\n
Let’s look at another example, this time a picture taken in Istanbul.
\n
A picture taken on İstiklal Avenue in Istanbul (credit: Youri van der Weide)\n
Geolocating this image isn’t too difficult, but readers can check the exact spot where it was taken here on İstiklal Avenue, a popular street in Istanbul, if they so wish.
\n
There are lots of reference points to look at in this image, including:
\n
\n
The pavement
\n
The mural and the construction site to the right
\n
The banners on lamp posts
\n
Stores and businesses (Lefon, Flormar, English Time, Zara)
\n
Advertisements
\n
Weather
\n
\n
One big hint is contained in the banner that is visible in the top left corner of the image.
\n
\n
A banner on İstiklal Avenue in Istanbul appears to show the dates of an event.
\n
Two dates appear to be written on the banner “28 eylül” (September) to “12 kas…”.
\n
The full word beginning “kas” is not completed, but Googling “months of the year in Turkish” shows that the only month that begins with these three letters is “kasım” (November).
\n
This could suggest that the picture was taken in the run up to or in between these months, although in which year is not clear.
\n
We just need to look for corroborating information to help confirm or reject this hypothesis. As with the Prague example above, reference images can begin to narrow down possible timeframes of when the image was taken
\n
Some handy tools that can provide reference imagery include Google Street View, which is extremely useful in most cities. However, in more rural areas or even in less visited areas of big cities there is less street level imagery. For example, compare the number of pictures that you can find from Times Square in New York City with what is available for a residential neighbourhood in the city, for example.
\n
There are, of course, alternatives to Google Maps or Street View, including Yandex Maps, Bing Maps, Mapillary and Open Street Cam (as detailed in the below thread).
\n
\n
Hallo, here is your unroll: @y_vdw: 🗓️🎄 Day 18 of the @quiztime Advent Calendar: I present you alternatives for Google Street View. As of today… https://t.co/ggsT01gFr9 Have a good day. 🤖
As with the Prague and New York examples above, there are also other options that can help narrow down a timeframe, including:
\n
\n
Instagram, Facebook pictures
\n
Foursquare, TripAdvisor etc.
\n
YouTube, Facebook videos from the area
\n
\n
In short, there is lots to dig into to help us find reference imagery that we can compare to the source image which we are trying to date.
\n
Starting with Google Street View, the latest available image that can be seen (at time of publication) is from August 2019. There are clearly significant differences from the source image.
\n
The mural isn’t there for one, nor is the scaffolding. We can also see that the pavement appears different.
\n
A Google Street View image from İstiklal Avenue in Istanbul captured in 2019.\n
One of the many useful features offered by Google Street View is the ability to look back through older images by clicking on the “see more dates” option in the left upper corner.
\n
\n
Google Street View offers users the opportunity to see other dates images have been captured at the same location.
\n
This brings up images from a range of other dates when Google Street View images have been captured.
\n
Google Street View shows other dates where images have been captured on İstiklal Avenue in Istanbul.\n
Immediately, it’s possible to match the graffiti image on the scaffolding in the October 2015 image on the far right of the above screen grab (and below).
\n
Google Street View shows an image of İstiklal Avenue in Istanbul captured in October 2015.\n
Obviously, October does fall within the 28 September to 12 November period we noticed were detailed on the blue and yellow banners earlier.
\n
But some aspects of the Street View image and our source image don’t match. For example, those very banners with the dates aren’t there in the Street View image.
\n
Other details also don’t quite match up. For example, the style of the lettering of the “Lefon” storefront is different. The advertisement on the side of the store next door is also different.
\n
A side-by-side view of a Google Street View image (left) of İstiklal Avenue and our source image (right), with several differences visible.\n
To help us keep track of the earliest and latest possible dates, we can create a timeline and update it as we go. Though not always necessary during a process of chronolocation, this can be a useful mental aid.
\n
We can also update the timeline as new details are discovered. Here’s what we know so far:
\n
\n
We know that the source picture was taken before the mural was removed and the construction ended.
\n
We also know the picture was taken after the artist painted the mural.
\n
\n
We also have a provisional latest possible date as we can see the graffiti and scaffolding had been removed in the 2018 Google Street View image. But we don’t know if the source image was taken before or after the October 2015 Street View image. At this stage, our timeline will look something like this.
\n
A timeline detailing what we know about when the image could have been taken so far.\n
Thinking laterally, could searching for the artist provide more useful contextual details?
\n
Using the artist’s handle (Omeria) that is visible on the graffiti, we can find his Instagram account. Scrolling back through the images, we can see a picture of the same graffiti was posted on October 10, 2015.
\n
\n
Of course, this could be an old picture, uploaded at a later date. But the image does appear far fresher than the Street View image. There are also no stickers or smudges on it, unlike the source image.
\n
Omeria also published a now unavailable Instagram post on October 8, 2015, that shows him in the process of creating the image.
\n
\n
While we must still acknowledge that the Instagram post could possibly be depicting events that happened before the 8 October 2015 upload date, it still provides us an initial date to work with as this is the earliest record of the graffiti that we have.
\n
Google Maps further tells us the Yapı Kredi Culture and Art centre is located where the scaffolding (and graffiti) was.
\n
A press release on the centre’s website shows that it re-opened in September 2017. If the scaffolding was removed upon re-opening (taking the graffiti with it as would appear necessary) that would mean we have further narrowed down the latest possible date the picture could have been taken.
\n
Our timeline now looks like this:
\n
An updated timeline detailing what we know about when the image could have been taken so far.\n
We can further narrow down the date range using our powers of deduction.
\n
We know the banners in the original picture said September to November. But as we discovered earlier, October 2015 was most likely too early as while the graffiti was present, there were no banners.
\n
A Google Street View image of İstiklal Avenue in 2015. While this image shows the graffiti, there are none of the banners that are visible in our source images.\n
September to November 2017 would most likely have been too late because the scaffolding bearing the graffiti would have been removed with the re-opening of the building.
\n
This leaves 2016 as a potential option.
\n
Our timeline now appears as follows.
\n
An updated timeline detailing what we know when the image could have been taken so far.\n
Unfortunately, Google Maps, Yandex and other street view services do not provide any imagery for 2016.
\n
So, as with the Prague example, we can begin to check out user-uploaded content.
\n
Luckily, İstiklal Avenue is a busy street with lots of visitors every day. Therefore the probability of finding footage is relatively high.
\n
Facebook, YouTube, Instagram, TripAdvisor, Four Square or any other website where people can tag footage to a place is worth researching.
\n
To keep things simple, we can stick to YouTube and Facebook for the purpose of this exercise as they are easier to search.
\n
Both Facebook and YouTube videos can be filtered by dates (a full explanation on how you can do that can be found here, although it is worth noting the section detailing Instagram in this link is now outdated). Using Google and filtering by the date function, we can see a video was uploaded on 6 November 2016 titled “Night walk around[sic]Istanbul Taksim Istiklal Galata Bridge”.
\n
Google search results for videos on İstiklal Avenue.\n
It’s 37 minutes long, but at 12 minutes and 21 seconds the uploader passes the same section of the street (although, looking in the opposite direction to our source image) where banners can be seen.
\n
The arrow on the left shows the area where the source image was taken from, while the arrow and box on the right show the yellow and blue banners also visible in the source picture.\n
For live streamed videos we can be sure about their upload time and date. Hovering over the details of when the video was streamed will show the precise broadcast date.
\n
Live streamed videos can be recognized by the word “streamed” as in “Streamed two years ago”. A video that wasn’t live streamed will only say “two years ago”, without the word streamed.\n
However, videos that weren’t live streamed aren’t guaranteed to have been filmed around their upload date. Someone could, of course, have filmed a video months before only to upload it later. For walking videos, uploaders sometimes give a precise date for when the video was filmed.
\n
If not, you must assess whether the video was filmed close to its upload date.
\n
Luckily, Facebook videos that live streamed from late October and early November show the same yellow and navy blue banners as our source image. We can now be confident the image was likely taken around this time given we already know it could not have been taken in 2015 or 2017.
\n
A Facebook Live video broadcast on 22 October 2016 shows blue and yellow banners on İstiklal Avenue.\n
Going back a little further we can start to narrow down our timeline even more.
\n
Live videos and photos from Facebook that were taken in late September – shortly before the event detailed in the banners seen in our source picture was due to begin – show Turkish flags instead of yellow and navy blue banners on lamp posts and throughout the street.
\n
For example, this is the scene in a live video from 23 September:
\n
Turkish flags can be seen in the top right of a live video filmed on İstiklal Avenue and streamed on Facebook on 23 September 2016.\n
On September 28, 2016, the first date mentioned on the banners, we found live videos, as well as newsreports, of the opening of a festival together with the first appearance of banners that appear to match those in the source image.
\n
A Facebook Live video posted on 28 September 2016 shows bronze and navy blue banners on İstiklal Avenue.\n
Though similar, the banners are bronze and navy blue as opposed to the yellow and navy blue that can be seen in the source image.
\n
Further investigation shows that images throughout the period between 28 September to 12 November show the colours of the banners regularly changed. This news article about a bomb scare at the same section of the street on 14 October showed the banners displayed were light blue and navy blue.
\n
Light blue and navy blue banners can be seen in a Live video from İstiklal Avenue that was streamed on YouTube on 14 October, 2016.\n
The first videos and photos that showed the colour of the banners had changed to yellow and navy blue appeared on October 18. By 10 November, these had been replaced by white and yellow banners which were there until at least the end of the festival (12 November).
\n
White banners can be seen in the top left of a live video filmed on İstiklal Avenue and streamed on Facebook on 12 November, 2016.\n
Given the abundance of reference material, we can now narrow down the possible date range the photo was taken to an even shorter period (18 October to 10 November) as can be seen in the updated timeline below detailing the colours on the banners.
\n
An updated, colour-coded timeline detailing what we know about the the image and when it could have been taken.\n
A screen grab of event details taken from the Dunya news site.\n
With all this information at hand, we can now confidently state that our reference image was taken at some point between October 18 and November 10, 2016.
\n
Of course, in other situations investigators might not be able to narrow the date down to such a narrow period. But this isn’t always necessary in an investigation. Often, all that is required is to confirm whether a picture or video could have been taken in a certain year, or before or after a certain period.
\n
The chronolocation techniques detailed here can be extremely helpful in such circumstances. In this instance it allowed us to go from having no idea of when a source image was taken, to narrowing the possible timeframe down to just a few weeks.
\n\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter hereand Mastodon here.
\n
",
+ "content_text": "How can open source investigators determine when a photograph or video was taken?\nObserving the length of shadows\nvisible in an image or clip\nand employing tools like\nSuncalc\noffers one useful method.\nBut this process of “\nchronolocation\n” – determining when a picture was taken rather than just where (which is known as\ngeolocation\n) – isn’t always feasible. For a start, it has to be daytime and there must be a shadow cast somewhere in the video or image of interest.\nIf these options aren’t available, looking for other clues in a “source” image that can be cross-referenced with contextual “reference” images or other related information can also help narrow down a date range.\nBellingcat has used such chronolocation techniques to determine when\nundated photos in a museum archive were taken\nand to find out when a\nCzech politician’s son (who claimed to be kidnapped) had been photographed\n.\nClues to look for in a source image may include:\nBuildings (especially their façades)\nConstruction sites\nStorefronts (similar to buildings but might be even more useful as this changes more frequently).\nSeasonal indicators such as weather, foliage etc.\nNatural landmarks\nClock towers\nPublic transport stops, bus lines etc.\nGraffiti and murals\nBanners\nAdvertisements\nEssentially, any aspect of a source image could be of use, provided that it has changed over time. Sometimes clues will be so obvious that it’s possible to immediately figure out the rough date of the source image from one detail alone.\nThis picture taken in front of CitiField, a baseball stadium in New York that serves as the home of the New York Mets, offers an example.\nFans gather outside CitiField in New York (photo: Youri van der Weide).\nFirstly, there are many features in this picture that can narrow down time and date.\nThe most obvious is the banners on the lampposts in front of the stadium. The year is detailed as 2017.\nA close up image of banners on display outside CitiField in New York (photo: Youri van der Weide).\nTo determine the exact date, however, we need more information.\nThere is a wooden structure on the left with a banner that reads, “Home is the Key” while the name of a non-profit, Habitat for Humanity, is also displayed prominently. Googling Habitat for Humanity in combination with the Mets yields only a few results.\nA page on Habitat’s\nwebsite\nis one of those results. It explains that their Home is the Key\ncampaign launched on April 3, 2017, on the opening day of the new baseball season for the New York Mets.\nText visible on a page of Habitat’s website details the “Home is the Key” campaign.\nAnother website\nvisible in the same Google search results also reported on the topic and shared a picture of the same wooden frame that is visible in the source image. We can further confirm the date of April 3 by looking for the Mets’ 2017 season\nschedule\n, indeed the first game of the 2017 season was on April 3.\nIt appears that a couple of key details and a quick Google search was all that was required to find a probable correct date for the image. Calling up or emailing Habitat to ask them whether this image was indeed captured on this date can provide further cast iron assurances.\nNarrowing Down Dates\nUnfortunately, chronolocation isn’t always that simple.\nIn most cases there will be a few extra steps and it will only be possible to figure out the earliest possible date or latest possible date that an image was taken, rather than the precise date and time.\nResearchers will often already have a latest possible date when they start the chronolocation process. This will likely be the date an image was uploaded to the platform they found it on.\nOf course, it may be possible to refine the latest upload date down to an earlier time frame with some investigation.\nIn 2018, Bellingcat published an article looking at claims that GRU officer “Ruslan Boshirov” (the cover identity for one of the men accused of poisoning former Russian spy Sergei Skripal and his daughter Yulia) made a trip to Prague back in 2014.\nA Facebook account using the name of Ruslan Boshirov had indeed posted from Prague on 11 October 2014.\nA screen grab of a Facebook post by an account bearing the name of Ruslan Boshirov.\nYet while it was not possible to say for sure that this indeed was the Facebook account of the GRU spy, it was possible to confirm that the image it displayed had indeed been taken in Prague between mid to late 2014.\nThis would become relevant several years later when it became apparent that Boshirov, whose real name is Anatoliy Chepiga, had indeed flown to Prague around this time and was linked with attacks on\nmunitions depots\nby Czech police.\nThe picture clearly shows Prague’s popular Old Town square, a site that will be familiar to any tourist who has visited the Czech capital.\nYet other, more subtle details allowed Bellingcat to confirm that the photograph had in fact been taken in Prague around that time.\nFirstly, it is not possible that the image could have been taken after 11 October, 2014, given that’s when it was posted to Facebook.\nBut by zooming in on a cream coloured building in the background, a number of large black letters flanked by two black squares can just about be made out.\nA number of letters can be made out on the frontage of a building in the distance of a picture uploaded by the Facebook account of an individual named Ruslan Boshirov.\nBy looking at other images of this building, which\nGoogle Maps\nshows\nis called the\nCentral Gallery\n, that appear on search engines or that were posted on social media sites during the European summer and autumn of 2014, we can begin to piece together a timeframe of when this photo was likely taken.\nAn\nimage\nof the exterior of the Central Gallery in Prague uploaded on 17 June 2014.\nThe above image was uploaded on 17 June 2014 and shows the names of artists (Alphonse) Mucha, (Andy) Warhol and (Salvador) Dali, flanked by two black squares. This makes sense as the building has long hosted exhibits by these artists, as detailed in Trip Advisor\nreviews\nstretching back to 2014.\nThe position of the names and black squares also matched the 11 October Facebook photo and other photos posted online around mid-June.\nHowever, photos that were uploaded in early June show a different signage on the building.\nImages\nof the Central Gallery in Prague taken on 4 June and 14 June 2014.\nOnly the names of two artists appear and there are no black boxes in images uploaded on June 4 and 14.\nWe may now be able to state that the photograph was dated sometime between mid June and 11 October when it was posted on Facebook. As with the CitiField stadium example earlier, a call to the museum could help confirm these details.\nHanging out in Istanbul\nLet’s look at another example, this time a picture taken in Istanbul.\nA picture taken on İstiklal Avenue in Istanbul (credit: Youri van der Weide)\nGeolocating this image isn’t too difficult, but readers can check the exact spot where it was taken\nhere\non İstiklal Avenue, a popular street in Istanbul, if they so wish.\nThere are lots of reference points to look at in this image, including:\nThe pavement\nThe mural and the construction site to the right\nThe banners on lamp posts\nStores and businesses (Lefon, Flormar, English Time, Zara)\nAdvertisements\nWeather\nOne big hint is contained in the banner that is visible in the top left corner of the image.\nA banner on İstiklal Avenue in Istanbul appears to show the dates of an event.\nTwo dates appear to be written on the banner “28 eylül” (September) to “12 kas…”.\nThe full word beginning “kas” is not completed, but Googling “months of the year in Turkish” shows that the only month that begins with these three letters is “kasım” (November).\nThis could suggest that the picture was taken in the run up to or in between these months, although in which year is not clear.\nWe just need to look for corroborating information to help confirm or reject this hypothesis. As with the Prague example above, reference images can begin to narrow down possible timeframes of when the image was taken\nSome handy tools that can provide reference imagery include Google Street View, which is extremely useful in most cities. However, in more rural areas or even in less visited areas of big cities there is less street level imagery. For example, compare the number of pictures that you can find from\nTimes Square\nin New York City with what is available for a residential neighbourhood in the city, for example.\nThere are, of course, alternatives to Google Maps or Street View, including Yandex Maps, Bing Maps, Mapillary and Open Street Cam (as detailed in the below thread).\nHallo, here is your unroll:\n@y_vdw\n: 🗓️🎄 Day 18 of the\n@quiztime\nAdvent Calendar: I present you alternatives for Google Street View. As of today…\nhttps://t.co/ggsT01gFr9\nHave a good day. 🤖\n— Thread Reader App (@threadreaderapp)\nDecember 18, 2020\nSome countries also have their own mapping services, such as Baidu in China and Mappy in France.\nIn Istanbul, however, there are plenty of resources that can help identify when the street image above was taken. These include:\nSatellite imagery\nGoogle Street View\nYandex Panorama\nIstanbul city council’s own mapping service\nthat provides something similar to Street View\nAs with the Prague and New York examples above, there are also other options that can help narrow down a timeframe, including:\nInstagram, Facebook pictures\nFoursquare, TripAdvisor etc.\nYouTube, Facebook videos from the area\nIn short, there is lots to dig into to help us find reference imagery that we can compare to the source image which we are trying to date.\nStarting with Google Street View, the latest available image that can be seen (at time of publication) is from August 2019. There are clearly significant differences from the source image.\nThe mural isn’t there for one, nor is the scaffolding. We can also see that the pavement appears different.\nA Google Street View image from İstiklal Avenue in Istanbul captured in 2019.\nOne of the many useful features offered by Google Street View is the ability to look back through older images by clicking on the “see more dates” option in the left upper corner.\nGoogle Street View offers users the opportunity to see other dates images have been captured at the same location.\nThis brings up images from a range of other dates when Google Street View images have been captured.\nGoogle Street View shows other dates where images have been captured on İstiklal Avenue in Istanbul.\nImmediately, it’s possible to match the graffiti image on the scaffolding in the October 2015 image on the far right of the above screen grab (and below).\nGoogle Street View shows an image of İstiklal Avenue in Istanbul captured in October 2015.\nObviously, October does fall within the 28 September to 12 November period we noticed were detailed on the blue and yellow banners earlier.\nBut some aspects of the Street View image and our source image don’t match. For example, those very banners with the dates aren’t there in the Street View image.\nOther details also don’t quite match up. For example, the style of the lettering of the “Lefon” storefront is different. The advertisement on the side of the store next door is also different.\nA side-by-side view of a Google Street View image (left) of İstiklal Avenue and our source image (right), with several differences visible.\nTo help us keep track of the earliest and latest possible dates, we can create a timeline and update it as we go. Though not always necessary during a process of chronolocation, this can be a useful mental aid.\nWe can also update the timeline as new details are discovered. Here’s what we know so far:\nWe know that the source picture was taken before the mural was removed and the construction ended.\nWe also know the picture was taken after the artist painted the mural.\nWe also have a provisional latest possible date as we can see the graffiti and scaffolding had been removed in the 2018 Google Street View image. But we don’t know if the source image was taken before or after the October 2015 Street View image. At this stage, our timeline will look something like this.\nA timeline detailing what we know about when the image could have been taken so far.\nThinking laterally, could searching for the artist provide more useful contextual details?\nUsing the artist’s handle (Omeria) that is visible on the graffiti, we can find his\nInstagram account\n. Scrolling back through the images, we can see a\npicture of the same graffiti\nwas posted on October 10, 2015.\nOf course, this could be an old picture, uploaded at a later date. But the image does appear far fresher than the Street View image. There are also no stickers or smudges on it, unlike the source image.\nOmeria also published a now unavailable Instagram post on October 8, 2015, that shows him in the process of creating the image.\nWhile we must still acknowledge that the Instagram post could possibly be depicting events that happened before the 8 October 2015 upload date, it still provides us an initial date to work with as this is the earliest record of the graffiti that we have.\nGoogle Maps further\ntells\nus the Yapı Kredi Culture and Art centre is located where the scaffolding (and graffiti) was.\nA\npress release\non the centre’s website shows that it re-opened in September 2017. If the scaffolding was removed upon re-opening (taking the graffiti with it as would appear necessary) that would mean we have further narrowed down the latest possible date the picture could have been taken.\nOur timeline now looks like this:\nAn updated timeline detailing what we know about when the image could have been taken so far.\nWe can further narrow down the date range using our powers of deduction.\nWe know the banners in the original picture said September to November. But as we discovered earlier, October 2015 was most likely too early as while the graffiti was present, there were no banners.\nA Google Street View image of İstiklal Avenue in 2015. While this image shows the graffiti, there are none of the banners that are visible in our source images.\nSeptember to November 2017 would most likely have been too late because the scaffolding bearing the graffiti would have been removed with the re-opening of the building.\nThis leaves 2016 as a potential option.\nOur timeline now appears as follows.\nAn updated timeline detailing what we know when the image could have been taken so far.\nUnfortunately, Google Maps, Yandex and other street view services do not provide any imagery for 2016.\nSo, as with the Prague example, we can begin to check out user-uploaded content.\nLuckily, İstiklal Avenue is a busy street with lots of visitors every day. Therefore the probability of finding footage is relatively high.\nFacebook, YouTube, Instagram, TripAdvisor, Four Square or any other website where people can tag footage to a place is worth researching.\nTo keep things simple, we can stick to YouTube and Facebook for the purpose of this exercise as they are easier to search.\nBoth Facebook and YouTube videos can be filtered by dates (a full explanation on how you can do that can be found\nhere\n, although it is worth noting the section detailing Instagram in this link is now outdated). Using Google and filtering by the date function, we can see a\nvideo\nwas uploaded on 6 November 2016 titled “\nNight walk around[sic]Istanbul Taksim Istiklal Galata Bridge”\n.\nGoogle search results for videos on İstiklal Avenue.\nIt’s 37 minutes long, but at 12 minutes and 21 seconds the uploader passes the same section of the street (although, looking in the opposite direction to our source image) where banners can be seen.\nThe arrow on the left shows the area where the source image was taken from, while the arrow and box on the right show the yellow and blue banners also visible in the source picture.\nFor live streamed videos we can be sure about their upload time and date. Hovering over the details of when the video was streamed will show the precise broadcast date.\nLive streamed videos can be recognized by the word “streamed” as in “Streamed two years ago”. A video that wasn’t live streamed will only say “two years ago”, without the word streamed.\nHowever, videos that weren’t live streamed aren’t guaranteed to have been filmed around their upload date. Someone could, of course, have filmed a video months before only to upload it later. For walking videos, uploaders sometimes give a precise date for when the video was filmed.\nIf not, you must assess whether the video was filmed close to its upload date.\nLuckily, Facebook videos that live streamed from late October and early November show the same yellow and navy blue banners as our source image. We can now be confident the image was likely taken around this time given we already know it could not have been taken in 2015 or 2017.\nA Facebook Live video broadcast on 22 October 2016 shows blue and yellow banners on İstiklal Avenue.\nGoing back a little further we can start to narrow down our timeline even more.\nLive videos and photos from Facebook that were taken in late September – shortly before the event detailed in the banners seen in our source picture was due to begin – show Turkish flags instead of yellow and navy blue banners on lamp posts and throughout the street.\nFor example, this is the scene in a live video from 23 September:\nTurkish flags can be seen in the top right of a live video filmed on İstiklal Avenue and streamed on Facebook on 23 September 2016.\nOn September 28, 2016, the first date mentioned on the banners, we found live videos, as well as\nnews\nreports\n, of the opening of a festival together with the first appearance of\nbanners\nthat appear to match those in the source image.\nA Facebook Live video posted on 28 September 2016 shows bronze and navy blue banners on İstiklal Avenue.\nThough similar, the banners are bronze and navy blue as opposed to the yellow and navy blue that can be seen in the source image.\nFurther investigation shows that images throughout the period between 28 September to 12 November show the colours of the banners regularly changed. This\nnews article\nabout a bomb scare at the same section of the street on 14 October showed the banners displayed were light blue and navy blue.\nLight blue and navy blue banners can be seen in a Live video from İstiklal Avenue that was streamed on YouTube on 14 October, 2016.\nThe first videos and photos that showed the colour of the banners had changed to yellow and navy blue appeared on October 18. By 10 November, these had been replaced by white and yellow banners which were there until at least the end of the festival (12 November).\nWhite banners can be seen in the top left of a live video filmed on İstiklal Avenue and streamed on Facebook on 12 November, 2016.\nGiven the abundance of reference material, we can now narrow down the possible date range the photo was taken to an even shorter period (18 October to 10 November) as can be seen in the updated timeline below detailing the colours on the banners.\nAn updated, colour-coded timeline detailing what we know about the the image and when it could have been taken.\nThe frequency at which the colours of the banners changed also appeared to tally largely with an\nevent calendar for the festival that was detailed in local media\nreports:\nA screen grab of event details taken from the\nDunya\nnews site.\nWith all this information at hand, we can now confidently state that our reference image was taken at some point between October 18 and November 10, 2016.\nOf course, in other situations investigators might not be able to narrow the date down to such a narrow period. But this isn’t always necessary in an investigation. Often, all that is required is to confirm whether a picture or video could have been taken in a\ncertain year\n, or before or after a certain period.\nThe chronolocation techniques detailed here can be extremely helpful in such circumstances. In this instance it allowed us to go from having no idea of when a source image was taken, to narrowing the possible timeframe down to just a few weeks.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-10.48.45-1200x893.png",
+ "alt": "",
+ "caption": "Fans gather outside CitiField in New York (photo: Youri van der Weide)."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-10.51.49.png",
+ "alt": "",
+ "caption": "A close up image of banners on display outside CitiField in New York (photo: Youri van der Weide)."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-10.53.53.png",
+ "alt": "",
+ "caption": "Text visible on a page of Habitat’s website details the “Home is the Key” campaign."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-10.56.59.png",
+ "alt": "",
+ "caption": "A screen grab of a Facebook post by an account bearing the name of Ruslan Boshirov."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-11.12.47.png",
+ "alt": "",
+ "caption": "A number of letters can be made out on the frontage of a building in the distance of a picture uploaded by the Facebook account of an individual named Ruslan Boshirov."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-11.25.55-1200x539.png",
+ "alt": "",
+ "caption": "An image of the exterior of the Central Gallery in Prague uploaded on 17 June 2014."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-11.27.47-1200x442.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-11.27.56-1200x579.png",
+ "alt": "",
+ "caption": "Images of the Central Gallery in Prague taken on 4 June and 14 June 2014."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-11.50.37.png",
+ "alt": "",
+ "caption": "A picture taken on İstiklal Avenue in Istanbul (credit: Youri van der Weide)"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.09.10.png",
+ "alt": "",
+ "caption": "A banner on İstiklal Avenue in Istanbul appears to show the dates of an event."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.27.21-1200x545.png",
+ "alt": "",
+ "caption": "A Google Street View image from İstiklal Avenue in Istanbul captured in 2019."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.28.49.png",
+ "alt": "",
+ "caption": "Google Street View offers users the opportunity to see other dates images have been captured at the same location."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.30.29-1200x168.png",
+ "alt": "",
+ "caption": "Google Street View shows other dates where images have been captured on İstiklal Avenue in Istanbul."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.31.35-1200x551.png",
+ "alt": "",
+ "caption": "Google Street View shows an image of İstiklal Avenue in Istanbul captured in October 2015."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.33.00-1200x696.png",
+ "alt": "",
+ "caption": "A side-by-side view of a Google Street View image (left) of İstiklal Avenue and our source image (right), with several differences visible."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.35.10-1200x305.png",
+ "alt": "",
+ "caption": "A timeline detailing what we know about when the image could have been taken so far."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Image-from-insta-709x1200.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.41.18-1200x775.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.41.49-1200x278.png",
+ "alt": "",
+ "caption": "An updated timeline detailing what we know about when the image could have been taken so far."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.43.02-1200x552.png",
+ "alt": "",
+ "caption": "A Google Street View image of İstiklal Avenue in 2015. While this image shows the graffiti, there are none of the banners that are visible in our source images."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.45.14-1200x311.png",
+ "alt": "",
+ "caption": "An updated timeline detailing what we know when the image could have been taken so far."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.56.07-1200x590.png",
+ "alt": "",
+ "caption": "Google search results for videos on İstiklal Avenue."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.57.48-1200x669.png",
+ "alt": "",
+ "caption": "The arrow on the left shows the area where the source image was taken from, while the arrow and box on the right show the yellow and blue banners also visible in the source picture."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.58.39-1200x534.png",
+ "alt": "",
+ "caption": "Live streamed videos can be recognized by the word “streamed” as in “Streamed two years ago”. A video that wasn’t live streamed will only say “two years ago”, without the word streamed."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.06.55-1200x601.png",
+ "alt": "",
+ "caption": "A Facebook Live video broadcast on 22 October 2016 shows blue and yellow banners on İstiklal Avenue."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.07.50-1200x596.png",
+ "alt": "",
+ "caption": "Turkish flags can be seen in the top right of a live video filmed on İstiklal Avenue and streamed on Facebook on 23 September 2016."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.11.20-1200x449.png",
+ "alt": "",
+ "caption": "A Facebook Live video posted on 28 September 2016 shows bronze and navy blue banners on İstiklal Avenue."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.17.23-1200x669.png",
+ "alt": "",
+ "caption": "Light blue and navy blue banners can be seen in a Live video from İstiklal Avenue that was streamed on YouTube on 14 October, 2016."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.18.18-1200x598.png",
+ "alt": "",
+ "caption": "White banners can be seen in the top left of a live video filmed on İstiklal Avenue and streamed on Facebook on 12 November, 2016."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.19.53-1200x458.png",
+ "alt": "",
+ "caption": "An updated, colour-coded timeline detailing what we know about the the image and when it could have been taken."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.21.25-1200x202.png",
+ "alt": "",
+ "caption": "A screen grab of event details taken from the Dunya news site."
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2023-05-08_finding-geolocation-leads-with-bellingcats-openstreetmap-search-tool.json b/bellingcat_geolocation_json/2023-05-08_finding-geolocation-leads-with-bellingcats-openstreetmap-search-tool.json
new file mode 100644
index 0000000000000000000000000000000000000000..15078f222e29560aac03b9c278ef949370921363
--- /dev/null
+++ b/bellingcat_geolocation_json/2023-05-08_finding-geolocation-leads-with-bellingcats-openstreetmap-search-tool.json
@@ -0,0 +1,157 @@
+{
+ "slug": "finding-geolocation-leads-with-bellingcats-openstreetmap-search-tool",
+ "title": "Finding Geolocation Leads with Bellingcat's OpenStreetMap Search Tool",
+ "url": "https://www.bellingcat.com/resources/how-tos/2023/05/08/finding-geolocation-leads-with-bellingcats-openstreetmap-search-tool/",
+ "date": "2023-05-08T16:42:29+00:00",
+ "modified": "2023-06-14T15:59:32+00:00",
+ "author": "Logan Williams",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Bellingcat's new tool for searching OpenStreetMap data helps geolocate images and identify starting points for geolocation investigations.",
+ "content_html": "
\n
Sometimes, the trickiest part of geolocating a photo can be knowing where to start looking.
\n
In previous Bellingcat investigations, starting points have involved reverse image search, searching Google Earth for soil that is the right colour, looking for minarets and even identification of plants.
\n
Bellingcat has built a new tool for searching OpenStreetMap data to help geolocate images and identify starting points for geolocation investigations, based on objects and structures you can identify in an image. You can think of it as a dramatically simplified version of the Overpass query language tool that some open source researchers may already be familiar with using via Overpass-Turbo.
\n
Users can sign up for the tool and check it out here.
\n
\n\n
\n
How the Tool Works
\n
Let’s learn how the tool works by looking at an example. Here’s an image from somewhere in Belgium. For the purposes of this exercise, let’s imagine that we have already narrowed down to the country in which the picture was taken. Can we geolocate it with OpenStreetMap Search?
\n
Photo: Logan Williams.\n
Firstly, we identify a couple unique things in the image that might be tagged in OpenStreetMap: the fountain, the tram tracks, and the convenience store on the corner. (The store also tells us that we’re somewhere French speaking – that’s helpful!)
\n
In the OpenStreetMap Search Tool, we can select these features by clicking or dragging the presets to the selection pane.
\n
\n
\n
The features all seem to be within about 100 metres of each other in the photograph, so we’ll leave the “Maximum distance between features” slider at its default setting of 100 metres.
\n
Next, we scroll down and adjust the map to select the region of interest that we want to search. In general, the smaller you can make your region of interest, the better. Regions that are too large might not be searchable, resulting in a search query timeout. This is related to the number of items in OpenStreetMap for that region. In regions of the world which have fewer items in OpenStreetMap, larger search queries are possible. Currently, Western Europe tends to be the most densely detailed area. This map of OSM node density by Martin Raifer is helpful for building intuition for this.
\n
We’ll select a wide area that includes the French-speaking region of Belgium.
\n
\n
Press search to run the query. When the query is complete, the results appear on the map and are listed as map chips, as can be seen below.
\n
\n
Our query returned 15 results (and took over 40 seconds to run — it would have taken longer and might not have succeeded if we had used a larger region of interest), mostly in Brussels, but also in a handful of smaller cities.
\n
\n
Looking at the results in map chip format, one of the responses stands out: number 8. Here we see a curve in the tram tracks over a water fountain, just as in the image. Click on the latitude/longtitude below the map chip to open the result in Google Maps and enter Google Street View.
\n
\n
Success! There’s no water in the fountain in the Google Street View image, and the corner store has a different name, but the intersection is clearly the same.
\n
For some queries, it might not be immediately obvious from the map chips which result is the best. The user can toggle between OpenStreetMap, Google Maps, Google Terrain, or Satellite imagery using the toggle above the results. It is also possible to export the results as a CSV or a KML file that can be opened in Google Earth.
\n
It can take some experience, intuition and guesswork to choose which features to include in your search query. Not everything is on OpenStreetMap. For example, you might see a tree in your image, but every individual tree is unlikely to be mapped. Additionally, OpenStreetMap is only as current as its last user-submitted edit, so certain features may not be up-to-date. Bellingcat’s tool is set up to search the database from April 2023 and will be updated twice a year.
On the OpenStreetMap web interface, the “query” tool can be used to view information about the features that exist in the map and how they are tagged.
\n
\n
For example, if we click on the small rectangular park in the centre of the map, we can see a list of all map features near the selected location.
\n
\n
In addition to many benches, trees (uncommon outside of very densely mapped Western European cities), and roads, we can also see the memorial “Chasseurs Ardennais.” Clicking on this item in the list shows all of the tags that are associated with it.
\n
\n
Browsing OpenStreetMap features like this can be useful for understanding how well mapped a location is. For example, in Khartoum, Sudan, we can see that there are far fewer mapped features, however roads, mosques, bus stops and parks are included.
\n
\n\nExercise: What other combinations of OpenStreetMap features can find the first example?
\n
The simplest modification is to select “shop (any)” instead of “convenience store” combined with “fountain” and “railroad.” Notice that this returns many more results, because it is a more general query. However, the correct location is still included. Also notice that it includes many very similar locations, because it will return one result for each shop, as this is the first item in the query. Reversing “shop (any)” and “fountain” so that “fountain” is the first filter in the query will instead return one result for each fountain.
\n
Making the search more general means that some queries might time out. Since we now know the area we are looking at, it is possible to experiment with filters by zooming in on Brussels and searching only that region.
\n
“Plaza/square” and “water body” could also be useful search queries for this location. For example, “plaza/square,” “water body,” and “railroad” find it quite easily. Without “railroad,” it is more difficult, but can be matched using pedestrian paths or buildings.
\n
\n
Searching for Custom Features
\n
There might be some features in an image that don’t yet have presets in our tool. You can search for those objects by adding a custom filter in the tool. To do this, you need to know the correct OpenStreetMap tag to query. OSM tags consist of a key and value (“key” = “value”), for example, “oneway” = “yes”. These tags carry a wealth of information about mappable things in the world. However, it is a collaboratively edited project, and sometimes these tags can be a bit chaotic. The best way to figure out how a particular object might be labelled is to consult the OpenStreetMap wiki.
\n
If you are using a preset, there is also a link to the wiki for the key the preset uses. This can be a useful starting point. For example, let’s take a closer look at the convenience store preset we used above.
\n
\n
\n
If you click on the link next to ‘shop’, you are taken to the wiki page for this key. The wiki contains other common values for this key. For example, we can see that “shop=bakery” is a valid OSM tag in the wiki.
\n
\n
\n
When adding a custom feature, the search tool will suggest the most common OSM tag values as autocomplete suggestions. For instance, we can find ‘bakery’ in the list, or we can click it directly.
\n
\n
\n
Now, we can click the “Add Custom Feature” button to add this as a search parameter.
\n
\n
\n
In this case, we have added it as an “any” feature, meaning that it matches points, lines, and polygons, which are exactly what they sound like. A point can be anything represented by a single point on the map, and unfortunately can often be interchangeable with a polygon. For example, a bakery might be represented by a point at its location, or it could be a polygon that traces the outline of the building. Lines can be clearer – for example, roads, fences, walls, etc are all definitively lines. If you know what kind of object you are looking for, you can change the “Feature type” selector. Otherwise, using “any” will find any type of feature (but will be slightly slower).
\n
This point/line/polygon taxonomy is different from how OpenStreetMap data itself is structured. It can be helpful to understand this structure when browsing the OpenStreetMap wiki, as you can also see how many of each type of object exists in the OSM database with that tag: nodes, ways, and relations.
\n
\n
\n
\n
nodes\n
\n
These are single geographic points. Some of these represent point features (for example, a power pylon or bus stop), but many of them are used as parts of ways.
\n
\n
\n
ways\n
\n
These are collections of points that either represent lines (like a road) or a polygon (like a building or industrial area).
\n
\n
\n
relations\n
\n
These are collections of ways or nodes that are linked in some way. For example, bus stops along a particular bus route, or disjointed polygons that are part of the same nature reserve.
\n
\n
\n
\n
So in this case, we can see that objects with the “shop” tag are mostly nodes (points), but about 20% are ways (lines/polygons, in this case probably polygons).
\n
Using Custom Features to Geolocate Another Image
\n
Let’s illustrate the use of a custom feature with a second, trickier example.
\n
Photo: Logan Williams.\n
How could this image be geolocated with the OSM tool?
\n
First, we need a region to start searching in. What appears to be Simplified Chinese characters can be seen on the wind turbines, so let’s assume that this image is taken in China.
\n
\n
\n
Additionally, we can see a large mountain range in the background and the earth seen in the foreground of the picture appears to be very dry. These visual clues could suggest the image was taken in the deserts of Xinjiang, in far western China, so we will search there.
\n
There are two obvious features that we can use with the OSM search tool: the wind turbines and the railway line. Though the mountains are clearly helpful for geolocation, it is not usable for the OSM search tool as they are much further than 500 metres away from our location. There is no preset in the tool for wind turbines so it is necessary to create a custom feature. To do this, it is important to know how wind turbines are tagged in OpenStreetMap. Googling “OSM tag wind turbine” finds the relevant wiki page.
\n
\n
On the wiki page, we can see that this tag is mainly applied to nodes (point features).
\n
\n
\n
This is all the information need to go ahead and create a custom feature: generator:source=wind .
\n
\n
We can add the railroad feature as well, as it is visible in front of the wind turbines in the image. Since it is hard to tell from the photograph how close the railroad and the wind turbines really are, let’s set the distance slider to the maximum of 500 metres.
\n
\n
We’ll choose a search area that covers a wide region around Xinjiang. Note the warning message about the size of the search area. As mentioned earlier, search regions that are too large might not be searchable. However this is related to the amount of detail in OpenStreetMap, which tends to be highest in Europe and North America. Searches in other parts of the world can cover larger areas without posing as much of an issue.
\n
After running the search, we have 100 results (“of many,” indicating there might be additional results not viewable unless we search a smaller area.) However, it is not immediately clear which of the results might be the one we are looking for.
\n
\n
You have a couple of options to examine the results; one is to use the ‘Google/Terrain/Satellite’ options to view the wind turbines, another is to click on the coordinates and explore them in Google Maps.
\n
A third option to explore the results and find the right area, is to export these points as a KML file and open it up in Google Earth. Scroll to the bottom of the page and click the “Export as KML” button to download a KML file.
\n
\n
The points are scattered over a large area. However, because we are looking for wind turbines near large mountains, the points in the northwest are promising. Zooming in, we can see multiple wind turbines arranged in rows, similar to what can be seen in the photograph.
\n
\n
Looking more closely, we can see that not all of the wind turbines are pinned. This is because there is a maximum of 100 features returned from the OSM search tool. Still, even with this limited number of features, it was able to guide us to this wind farm location.
\n
\n
Following the rail line along the wind farm, one area looks particularly promising. We can see an earthen berm from the original image, the draining culvert, and a light coloured patch of soil that matches the soil in the image.
\n
\n
Finally, if we zoom in to Google Earth’s ground level view, we see a good terrain match with the background mountains at this location (partially outlined in red.)
\n
Google Earth ground level view of the terrain and a section of the original image, showing the same mountain range.\n
This shows how even when there are few features in an image the OpenStreetMap Search Tool can be useful for geolocation – especially when combined with other tools such as Google Earth.
It requires a login with a Google account. As we are currently testing this tool, we cannot provide any guarantees about uptime or accessibility. Additionally, we are logging query requests in order to track performance and identify bugs.
\n\n
Update (14 June, 2023):The source code for the OSM Search Tool is now available on Bellingcat’s Github page.
\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter hereand Mastodon here.
\n
",
+ "content_text": "Sometimes, the trickiest part of geolocating a photo can be knowing where to start looking.\nIn previous Bellingcat investigations, starting points have involved reverse image search, searching Google Earth for soil that is the right colour,\nlooking for minarets\nand even\nidentification of plants\n.\nBellingcat has built a new tool for searching OpenStreetMap data to help geolocate images and identify starting points for geolocation investigations, based on objects and structures you can identify in an image. You can think of it as a dramatically simplified version of the Overpass query language tool that some open source researchers may already be familiar with using via\nOverpass-Turbo\n.\nUsers can sign up for the tool and check it out\nhere.\nHow the Tool Works\nLet’s learn how the tool works by looking at an example. Here’s an image from somewhere in Belgium. For the purposes of this exercise, let’s imagine that we have already narrowed down to the country in which the picture was taken. Can we geolocate it with OpenStreetMap Search?\nPhoto: Logan Williams.\nFirstly, we identify a couple unique things in the image that might be tagged in OpenStreetMap: the fountain, the tram tracks, and the convenience store on the corner. (The store also tells us that we’re somewhere French speaking – that’s helpful!)\nIn the OpenStreetMap Search Tool, we can select these features by clicking or dragging the presets to the selection pane.\nThe features all seem to be within about 100 metres of each other in the photograph, so we’ll leave the “Maximum distance between features” slider at its default setting of 100 metres.\nNext, we scroll down and adjust the map to select the region of interest that we want to search. In general, the smaller you can make your region of interest, the better. Regions that are too large might not be searchable, resulting in a search query timeout. This is related to the number of items in OpenStreetMap for that region. In regions of the world which have fewer items in OpenStreetMap, larger search queries are possible. Currently, Western Europe tends to be the most densely detailed area.\nThis map of OSM node density\nby\nMartin Raifer\nis helpful for building intuition for this.\nWe’ll select a wide area that includes the French-speaking region of Belgium.\nPress search to run the query. When the query is complete, the results appear on the map and are listed as map chips, as can be seen below.\nOur query returned 15 results (and took over 40 seconds to run — it would have taken longer and might not have succeeded if we had used a larger region of interest), mostly in Brussels, but also in a handful of smaller cities.\nLooking at the results in map chip format, one of the responses stands out: number 8. Here we see a curve in the tram tracks over a water fountain, just as in the image. Click on the latitude/longtitude below the map chip to open the result in Google Maps and enter Google Street View.\nSuccess! There’s no water in the fountain in the Google Street View image, and the corner store has a different name, but the intersection is clearly the same.\nFor some queries, it might not be immediately obvious from the map chips which result is the best. The user can toggle between OpenStreetMap, Google Maps, Google Terrain, or Satellite imagery using the toggle above the results. It is also possible to export the results as a CSV or a KML file that can be opened in Google Earth.\nIt can take some experience, intuition and guesswork to choose which features to include in your search query. Not everything is on OpenStreetMap. For example, you might see a tree in your image, but every individual tree is unlikely to be mapped. Additionally, OpenStreetMap is only as current as its last user-submitted edit, so certain features may not be up-to-date. Bellingcat’s tool is set up to search the database from April 2023 and will be updated twice a year.\nOn the\nOpenStreetMap web interface\n, the “query” tool can be used to view information about the features that exist in the map and how they are tagged.\nFor example, if we click on the small rectangular park in the centre of the map, we can see a list of all map features near the selected location.\nIn addition to many benches, trees (uncommon outside of very densely mapped Western European cities), and roads, we can also see the memorial “Chasseurs Ardennais.” Clicking on this item in the list shows all of the tags that are associated with it.\nBrowsing OpenStreetMap features like this can be useful for understanding how well mapped a location is. For example, in Khartoum, Sudan, we can see that there are far fewer mapped features, however roads, mosques, bus stops and parks are included.\nExercise: What other combinations of OpenStreetMap features can find the first example?\nThe simplest modification is to select “shop (any)” instead of “convenience store” combined with “fountain” and “railroad.” Notice that this returns many more results, because it is a more general query. However, the correct location is still included. Also notice that it includes many very similar locations, because it will return one result for each shop, as this is the first item in the query. Reversing “shop (any)” and “fountain” so that “fountain” is the first filter in the query will instead return one result for each fountain.\nMaking the search more general means that some queries might time out. Since we now know the area we are looking at, it is possible to experiment with filters by zooming in on Brussels and searching only that region.\n“Plaza/square” and “water body” could also be useful search queries for this location. For example, “plaza/square,” “water body,” and “railroad” find it quite easily. Without “railroad,” it is more difficult, but can be matched using pedestrian paths or buildings.\nSearching for Custom\nFeatures\nThere might be some features in an image that don’t yet have presets in our tool. You can search for those objects by adding a custom filter in the tool. To do this, you need to know the correct OpenStreetMap tag to query. OSM tags consist of a key and value (“key” = “value”), for example, “oneway” = “yes”. These tags carry a wealth of information about mappable things in the world. However, it is a collaboratively edited project, and sometimes these tags can be a bit chaotic. The best way to figure out how a particular object might be labelled is to consult the\nOpenStreetMap wiki\n.\nIf you are using a preset, there is also a link to the wiki for the key the preset uses. This can be a useful starting point. For example, let’s take a closer look at the convenience store preset we used above.\nIf you click on the link next to ‘shop’, you are taken to the\nwiki page for this key\n. The wiki contains other common values for this key. For example, we can see that “shop=bakery” is a valid OSM tag in the wiki.\nWhen adding a custom feature, the search tool will suggest the most common OSM tag values as autocomplete suggestions. For instance, we can find ‘bakery’ in the list, or we can click it directly.\nNow, we can click the “Add Custom Feature” button to add this as a search parameter.\nIn this case, we have added it as an “any” feature, meaning that it matches\npoints\n,\nlines\n, and\npolygons\n, which are exactly what they sound like. A point can be anything represented by a single point on the map, and unfortunately can often be interchangeable with a polygon. For example, a bakery might be represented by a point at its location, or it could be a polygon that traces the outline of the building. Lines can be clearer – for example, roads, fences, walls, etc are all definitively lines. If you know what kind of object you are looking for, you can change the “Feature type” selector. Otherwise, using “any” will find any type of feature (but will be slightly slower).\nThis point/line/polygon taxonomy is different from how OpenStreetMap data itself is structured. It can be helpful to understand this structure when browsing the OpenStreetMap wiki, as you can also see how many of each type of object exists in the OSM database with that tag:\nnodes, ways,\nand\nrelations\n.\nnodes\nThese are single geographic points. Some of these represent point features (for example, a power pylon or bus stop), but many of them are used as parts of ways.\nways\nThese are collections of points that either represent lines (like a road) or a polygon (like a building or industrial area).\nrelations\nThese are collections of ways or nodes that are linked in some way. For example, bus stops along a particular bus route, or disjointed polygons that are part of the same nature reserve.\nSo in this case, we can see that objects with the “shop” tag are mostly nodes (points), but about 20% are ways (lines/polygons, in this case probably polygons).\nUsing Custom Features to Geolocate Another Image\nLet’s illustrate the use of a custom feature with a second, trickier example.\nPhoto: Logan Williams.\nHow could this image be geolocated with the OSM tool?\nFirst, we need a region to start searching in. What appears to be Simplified Chinese characters can be seen on the wind turbines, so let’s assume that this image is taken in China.\nAdditionally, we can see a large mountain range in the background and the earth seen in the foreground of the picture appears to be very dry. These visual clues could suggest the image was taken in the deserts of Xinjiang, in far western China, so we will search there.\nThere are two obvious features that we can use with the OSM search tool: the wind turbines and the railway line. Though the mountains are clearly helpful for geolocation, it is not usable for the OSM search tool as they are much further than 500 metres away from our location. There is no preset in the tool for wind turbines so it is necessary to create a custom feature. To do this, it is important to know how wind turbines are tagged in OpenStreetMap. Googling “OSM tag wind turbine” finds the\nrelevant wiki page\n.\nOn the wiki page, we can see that this tag is mainly applied to nodes (point features).\nThis is all the information need to go ahead and create a custom feature: generator:source=wind .\nWe can add the railroad feature as well, as it is visible in front of the wind turbines in the image. Since it is hard to tell from the photograph how close the railroad and the wind turbines really are, let’s set the distance slider to the maximum of 500 metres.\nWe’ll choose a search area that covers a wide region around Xinjiang. Note the warning message about the size of the search area. As mentioned earlier, search regions that are too large might not be searchable. However this is related to the amount of detail in OpenStreetMap, which tends to be highest in Europe and North America. Searches in other parts of the world can cover larger areas without posing as much of an issue.\nAfter running the search, we have 100 results (“of many,” indicating there might be additional results not viewable unless we search a smaller area.) However, it is not immediately clear which of the results might be the one we are looking for.\nYou have a couple of options to examine the results; one is to use the ‘Google/Terrain/Satellite’ options to view the wind turbines, another is to click on the coordinates and explore them in Google Maps.\nA third option to explore the results and find the right area, is to export these points as a KML file and open it up in Google Earth. Scroll to the bottom of the page and click the “Export as KML” button to download a KML file.\nThe points are scattered over a large area. However, because we are looking for wind turbines near large mountains, the points in the northwest are promising. Zooming in, we can see multiple wind turbines arranged in rows, similar to what can be seen in the photograph.\nLooking more closely, we can see that not all of the wind turbines are pinned. This is because there is a maximum of 100 features returned from the OSM search tool. Still, even with this limited number of features, it was able to guide us to this wind farm location.\nFollowing the rail line along the wind farm, one area looks particularly promising. We can see an earthen berm from the original image, the draining culvert, and a light coloured patch of soil that matches the soil in the image.\nFinally, if we zoom in to Google Earth’s ground level view, we see a good terrain match with the background mountains at this location (partially outlined in red.)\nGoogle Earth ground level view of the terrain and a section of the original image, showing the same mountain range.\nThis shows how even when there are few features in an image the OpenStreetMap Search Tool can be useful for geolocation – especially when combined with other tools such as Google Earth.\nAccess to the Tool\nYou can currently access Bellingcat’s OpenStreetMap Search Tool by visiting:\nhttps://osm-search.bellingcat.com/\n.\nIt requires a login with a Google account. As we are currently testing this tool, we cannot provide any guarantees about uptime or accessibility. Additionally, we are logging query requests in order to track performance and identify bugs.\nUpdate (14 June, 2023):\nThe source code for the OSM Search Tool is now available on\nBellingcat’s Github page\n.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image19-1200x606.jpg",
+ "alt": "",
+ "caption": "Photo: Logan Williams."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image7-1-1200x677.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image5-3-1200x829.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image17-1200x743.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image20-1200x825.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image22-1200x1054.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image8-1-1200x841.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image4-3.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image6.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image10.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image24-1200x663.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image12-1-300x109.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image9-1-300x65.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image23-300x227.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image21-300x105.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image14.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image11-1200x800.jpg",
+ "alt": "",
+ "caption": "Photo: Logan Williams."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image18.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image2-1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image3-2.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image1-4.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image25.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image27-1200x642.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image26-1200x744.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image28-1200x865.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image15-3-1200x861.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image16-1-1200x411.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/04/image13-1200x412.jpg",
+ "alt": "",
+ "caption": "Google Earth ground level view of the terrain and a section of the original image, showing the same mountain range."
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2023-06-23_russias-foreign-fighters-geolocating-the-nepalis-training-in-the-russian-army.json b/bellingcat_geolocation_json/2023-06-23_russias-foreign-fighters-geolocating-the-nepalis-training-in-the-russian-army.json
new file mode 100644
index 0000000000000000000000000000000000000000..12b79817fe0863219680034774c9579705704ed0
--- /dev/null
+++ b/bellingcat_geolocation_json/2023-06-23_russias-foreign-fighters-geolocating-the-nepalis-training-in-the-russian-army.json
@@ -0,0 +1,72 @@
+{
+ "slug": "russias-foreign-fighters-geolocating-the-nepalis-training-in-the-russian-army",
+ "title": "Russia's Foreign Fighters: Geolocating the Nepalis Training in the Russian Army",
+ "url": "https://www.bellingcat.com/news/2023/06/23/russias-foreign-fighters-geolocating-the-nepalis-training-in-the-russian-army/",
+ "date": "2023-06-23T18:40:01+00:00",
+ "modified": "2026-02-11T14:48:09+00:00",
+ "author": "Bellingcat Investigation Team, Sanjib Chaudhary",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Bellingcat geolocated footage of Nepali-speaking soldiers undertaking training at a Russian military camp west of Moscow.",
+ "content_html": "
\n
Bellingcat has geolocated footage of several Nepali-speaking soldiers undertaking training at a Russian military camp.
\n
At least eight TikTok accounts seen by Bellingcat appear to belong to Nepali men showcasing training in one branch of the Russian military.
\n
Nepali journalist Birat Anupam first published on the story for NepalPress, examining footage posted by apparent recruits on social media and spoke to some recruits about joining the Russian military.
\n
Bellingcat was able to geolocate footage taken at both the Avangard military complex located west of Moscow and at a station in the Belarusian city of Baranavichy.
\n
But who are these young men who’ve found themselves on the margins of Europe’s largest conflict since the Second World War?
\n
One TikTok account that we examined was that of a user with the handle Kamalacharya9323 and listed themselves as from Nepal.
\n
The account has uploaded several videos from a Russian training camp in recent months. In one video they eat in an army canteen. In another they dance in a dormitory to a music video.
\n
Separately another Nepali recruit apparently signed up to fight for Russia as a rifleman in a motor-rifle unit. This account holder posted multiple videos from trainings in Russia but has also posted a warning people not to join the Russian Army due to the risk.
\n
A signature can be seen on a Russian-language sign-up contract that was posted in a video on the TikTok account and was also published by NepalPress.
\n
Referencing a decree by Russian President Vladimir Putin from September 2022, it confirms that the recruit is entitled to a one-off payment of 195,000 roubles (US$2,300).
\n
That same month, Putin also signed a decree fast-tracking the naturalisation of foreigners who fight for the Russian armed forces.
It should be noted that Bellingcat has not yet seen any evidence that these men have taken part in active combat.
\n
Bellingcat attempted to contact several of the TikTok accounts, including Kamalacharya9323, to ask why they joined the Russian army and how they found out about it, but had received no response at time of publication.
\n
Locating the Soldiers
\n
Recent posts on several accounts show Nepali men in military fatigues, training or posing for selfies.
\n
Many of these videos appear to have been taken in and around the Avangard centre – a military complex in Russia’s Moscow Oblast Region approximately 50 kilometres west of Moscow (55.565474, 36.790801).
\n
The facility, constructed in 2019, describes itself as an “educational centre for military patriotic youth.”
\n
According to its website, Avangard offers to familiarise trainees with five “military occupational specialties” and can accommodate 600 trainees at the same time. It claims to train more than 30,000 trainees a year.
\n
A video showcasing the centre’s “Young Fighter Course” displays some of the facilities and locations around the centre.
\n
These appear to match the details of some of the TikTok videos from Nepali men, which in turn also match satellite imagery of the Avangard site.
\n
Stills taken from a TikTok video stitched together shows the training grounds at the Avangard Military Complex. \n
Google Earth Image of Avangard Military Complex from August 2021. \n
The green building seen at the back of the TikTok video (above) hadn’t been finished when the Google Earth image was taken.
\n
But we can see the red training area with the obstacle course and the ground they broke behind the training area for the new building.
\n
Other TikTok videos show soldiers posing in front of one of the Avangard buildings.
\n
We can see the letters АВАНГ (AVANG) written on the building, although the word does not appear to be completed. This location appears to correspond to the main building at (55.565107, 36.792787) where the word АВАНГAPД (Avangard) can be seen on the roof. Other telltale details such as the frontage and lettering visible on the building match Google Maps.
\n
\n
A still from a TikTok video, showing three men in army fatigues in front of Avangard sign (we have blurred out faces).
\n
Matching lettering seen on the building in this image from Avangard website image gallery (we have blurred out faces). Image from Avangard website.\n
A Google Earth image shows the main building from above.\n
A number of TikTok accounts that had posted pictures from the Avangard complex, also posted videos of themselves in other locations.
\n
As mentioned, one user posted a sign-up contract and the top right hand corner of this document reads в/ч 91704, which is the unit number for the 254th Motor-Rifle Regiment (254th MRR) which is based in Klintsy, Bryansk Oblast, Russia (52.7140448467, 32.2691789635).
\n
We found several TikTok accounts and a YouTube channel from the Nepali men posting from this location.
\n
Left: Still from a TikTok video. Right: A Google Earth image of the barracks of the 254th MRR base in Russia (52.7140448467, 32.2691789635). Coloured boxes indicate matching features, with the green box around a vehicle in the foreground and buildings highlighted behind this. \n
In June, 2023 one TikTok account posted a video showing Russian military personnel and what appear to be other volunteers waiting at a train station.
\n
We geolocated this video to Maslovka train station in Voronezh, Russia (51.564140, 39.246172). The station was a hub for the Russian military buildup in 2021, leading up to the full-scale Russian invasion of Ukraine in February 2022.
\n
Still from a Tiktok video (we have blurred out faces). Matching features highlighted in coloured boxes.\n
Google Street View image of Maslovka train station (51.564140, 39.246172). Matching features highlighted in coloured boxes.\n
Google Earth image of the same area. Matching features highlighted in coloured boxes. \n
The TikTok image is taken from the platform, the Google Street View is taken from the road, down the hill, left of the platform. Google Earth Pro you can see Maslovka station from above.
\n
Also in June, an account posted a video showing a group of soldiers at another train station. By stitching together individual frames of the video, we could clearly see identifiable elements in the background. It came after several posts on the Nepali TikTok accounts referring to travelling to Belarus for training.
\n
Stitched together stills from footage from a post on TikTok, we identified a number of objects in the footage that we matched to a station in the Belarusian city of Braranavichy.\n
Google Earth Image with coloured boxes indicating objects and structures we matched to the location in the TikTok video.\n
The Nepali TikTok accounts did not show overwhelming evidence of pro-Russian sentiment.
\n
In recent months, the Russian military has sought to address its manpower needs by attracting foreign citizens.
\n
Central Asians are the most numerous foreign fighters but as the Moscow Times recently noted, Ethiopians, Sri Lankans, Tanzanians and Zambians have also signed up.
\n
Al-Jazeera reported that the minimum monthly wage offered for people signing up for what Russia describes as the “special military operation in Ukraine” is 160,000 rubles (US$2,700). Significantly higher than the average wage in Russia. One Nepali recruit who spoke to Nepalese outlet Khabarhub but wished to remain anonymous said the soldiers would be eligible for a salary of up to $2,500 once training was completed.
\n
It’s not clear what wage is being offered to soldiers in training.
\n
Why are Nepalis Joining the Russian Army?
\n
Evidence the accounts were linked to Nepali nationals was found from clues in their TikTok profiles.
\n
A Nepali speaker reviewed the eight accounts we found, and found they appeared to include people from different regions of Nepal – including from the southern plains that border with India to people from the hills.
\n
In several of the accounts we identified Nepali-speaking men with Nepali dialects. Accounts used songs and dialogue in their videos that could either be from Nepal or from India – where there are several Nepali origin communities. In these cases we checked the older posts from the accounts, which indeed appeared to show footage and images taken inside Nepal.
\n
The Nepali-language article by the aforementioned journalist Birat Anupam for NepalPress offered some suggestions as to the potential motives of Nepalis signing up to serve in the Russian army.
\n
It suggests that Nepalis come to Russia as “students and tourists” and subsequently sign up for the country’s military.
\n
Anupam contacted two Nepali men who had joined the Russian army and spoke to them via Telegram, initially under the guise of being a Nepali youth expecting to join the Russian Army. He was blocked after he told them he was a journalist.
\n
Both men spoke about the difficulties of joining Western militaries such as the French Foreign Legion or the military in the US or UK. And both spoke about the ease of joining the Russian army.
\n
“After one year, we can get citizenship as well. If I don’t die after a year, I will stay here only.” One of the men told NepalPress, but did not want to be named in the report.
\n
The article also cited a scanned letter from the Embassy of Nepal in Moscow dated to the 28 of the Nepali month of Bhadra (August-September in the Gregorian calendar).
\n
It stated that the number of Nepali students at Russian universities had dramatically increased and advised citizens to be wary of easy student visas.
However, at time of publication, this letter appeared to have disappeared from the Nepali embassy’s website.
\n
An archived copy of the letter can still be accessed. It is not known why the letter appears to have been removed.
\n
Bellingcat contacted Nepal’s Foreign Ministry and Embassy of Nepal in Moscow to ask about the presence of Nepali citizens in the Russian army, how many Nepalis were currently serving in the Russian army and what Nepal’s position is on its citizens joining the Russian army. At time of publication we had not received a reply.
\n
Nepal’s Military Tradition
\n
The length of time that the men featured in the TikTok accounts spent in Russia, prior to their appearance in the training camps, varied.
\n
In at least one case, we found evidence that pointed to their arrival in Russia as early as November, 2022, when they appeared to be studying Russian. It is not clear how long the other men featured in the TikTok accounts spent in Russia prior to their appearance at the training camps.
\n
Several of the TikTok accounts include hashtags such as #gurkha and #gorkhali. Others use #lahurey, a reference to “those who go to Lahore,” where the British Indian army used to train Nepali soldiers.
\n
All these hashtags allude to The Gurkhas, an elite Nepalese group of fighters initially recruited by The British East India Company during British Colonial Rule but who are still recruited to the British and Indian armies today. However, there is no evidence the Nepali soldiers training in Russia are from Gurkha regiments.
Bellingcat contacted both the Russian and Nepal foreign ministries to ask about Nepali soldiers serving in the Russian army. We also contacted the Embassy of Nepal in Moscow and attempted to reach several Nepali TikTok account holders. At time of publication we had received no response.
\n\n
Maxim Edwards, Youri van der Weide, Michael Sheldon, Aric Toler and Dennis Kovtun contributed to this report.
\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter here and Mastodon here.
\n
",
+ "content_text": "Bellingcat has geolocated footage of several Nepali-speaking soldiers undertaking training at a Russian military camp.\nAt least eight TikTok accounts seen by Bellingcat appear to belong to Nepali men showcasing training in one branch of the Russian military.\nNepali journalist Birat Anupam first published on the\nstory for NepalPress\n, examining footage posted by apparent recruits on social media and spoke to some\nrecruits\nabout joining the Russian military.\nBellingcat was able to geolocate footage taken at both the Avangard military complex located west of Moscow and at a station in the Belarusian city of Baranavichy.\nBut who are these young men who’ve found themselves on the margins of Europe’s largest conflict since the Second World War?\nOne TikTok account that we examined was that of a user with the handle Kamalacharya9323 and listed themselves as from Nepal.\nThe\naccount has uploaded\nseveral videos from a Russian training camp in recent months. In one video they eat in an army canteen. In another they dance in a dormitory to a music video.\nSeparately another Nepali recruit apparently signed up to fight for Russia as a rifleman in a motor-rifle unit. This account holder posted multiple videos from trainings in Russia but has also posted a warning people not to join the Russian Army due to the risk.\nA signature can be seen on a Russian-language sign-up contract that was posted in a video on the TikTok account and was also\npublished\nby NepalPress.\nReferencing a\ndecree\nby Russian President Vladimir Putin from September 2022, it confirms that the recruit is entitled to a one-off payment of 195,000 roubles (US$2,300).\nThat same month, Putin also signed a\ndecree\nfast-tracking the naturalisation of foreigners who fight for the Russian armed forces.\nIt should be noted that Bellingcat has not yet seen any evidence that these men have taken part in active combat.\nBellingcat attempted to contact several of the TikTok accounts, including Kamalacharya9323, to ask why they joined the Russian army and how they found out about it, but had received no response at time of publication.\nLocating the Soldiers\nRecent posts on several accounts show Nepali men in military fatigues, training or posing for selfies.\nMany of these videos appear to have been taken in and around the Avangard centre – a military complex in Russia’s Moscow Oblast Region approximately 50 kilometres west of Moscow (\n55.565474, 36.790801\n).\nThe facility, constructed in 2019,\ndescribes\nitself as an “educational centre for military patriotic youth.”\nAccording to its website, Avangard offers to familiarise trainees with five “military occupational specialties” and can accommodate 600 trainees at the same time. It claims to train more than 30,000 trainees a year.\nA\nvideo\nshowcasing the centre’s “Young Fighter Course” displays some of the facilities and locations around the centre.\nThese appear to match the details of some of the TikTok videos from Nepali men, which in turn also match satellite imagery of the Avangard site.\nStills taken from a TikTok video stitched together shows the training grounds at the Avangard Military Complex.\nGoogle Earth Image of Avangard Military Complex from August 2021.\nThe green building seen at the back of the TikTok video (above) hadn’t been finished when the Google Earth image was taken.\nBut we can see the red training area with the obstacle course and the ground they broke behind the training area for the new building.\nOther TikTok videos show soldiers posing in front of one of the Avangard buildings.\nWe can see the letters АВАНГ (AVANG) written on the building, although the word does not appear to be completed. This location appears to correspond to the main building at (\n55.565107, 36.792787\n) where the word АВАНГAPД (Avangard) can be seen on the roof. Other telltale details such as the frontage and lettering visible on the building match\nGoogle Maps\n.\nA still from a TikTok video, showing three men in army fatigues in front of Avangard sign (we have blurred out faces).\nMatching lettering seen on the building in\nthis image\nfrom Avangard website image gallery (we have blurred out faces). Image from Avangard website.\nA Google Earth image shows the main building from above.\nA number of TikTok accounts that had posted pictures from the Avangard complex, also posted videos of themselves in other locations.\nAs mentioned, one user posted a sign-up contract and the top right hand corner of this document reads в/ч 91704, which is the unit number for the 254th Motor-Rifle Regiment (254th MRR)\nwhich is based\nin Klintsy, Bryansk Oblast, Russia (\n52.7140448467, 32.2691789635\n).\nWe found several TikTok accounts and a YouTube channel from the Nepali men posting from this location.\nLeft: Still from a TikTok video. Right: A Google Earth image of the barracks of the 254th MRR base in Russia (\n52.7140448467, 32.2691789635\n). Coloured boxes indicate matching features, with the green box around a vehicle in the foreground and buildings highlighted behind this.\nIn June, 2023 one TikTok account posted a video showing Russian military personnel and what appear to be other volunteers waiting at a train station.\nWe geolocated this video to Maslovka train station in Voronezh, Russia\n(51.564140, 39.246172)\n. The station was a\nhub\nfor the Russian military buildup in 2021, leading up to the full-scale Russian invasion of Ukraine in February 2022.\nStill from a Tiktok video (we have blurred out faces). Matching features highlighted in coloured boxes.\nGoogle Street View image of Maslovka train station (\n51.564140, 39.246172\n). Matching features highlighted in coloured boxes.\nGoogle Earth image of the same area. Matching features highlighted in coloured boxes.\nThe TikTok image is taken from the platform, the Google Street View is taken from the road, down the hill, left of the platform. Google Earth Pro you can see Maslovka station from above.\nAlso in June, an account posted a video showing a group of soldiers at another train station. By stitching together individual frames of the video, we could clearly see identifiable elements in the background. It came after several posts on the Nepali TikTok accounts referring to travelling to Belarus for training.\nStitched together stills from footage from a post on TikTok, we identified a number of objects in the footage that we matched to a station in the Belarusian city of Braranavichy.\nGoogle Earth Image with coloured boxes indicating objects and structures we matched to the location in the TikTok video.\nThis video was geolocated to the Belarusian city of Baranavichy\n(53.132000, 25.996833)\n, which has a large\ntraining area\nnearby.\nRussia’s Need for Foreign Fighters\nThe Nepali TikTok accounts did not show overwhelming evidence of pro-Russian sentiment.\nIn recent months, the Russian military has sought to address its manpower needs by attracting foreign citizens.\nCentral Asians are the most numerous foreign fighters but as the Moscow Times recently\nnoted\n, Ethiopians, Sri Lankans, Tanzanians and Zambians have also signed up.\nAl-Jazeera\nreported\nthat the minimum monthly wage offered for people signing up for what Russia describes as the “special military operation in Ukraine” is 160,000 rubles (US$2,700). Significantly higher than the average wage in Russia. One Nepali recruit who spoke to Nepalese outlet Khabarhub but wished to remain anonymous said the\nsoldiers would be eligible\nfor a salary of up to $2,500 once training was completed.\nIt’s not clear what wage is being offered to soldiers in training.\nWhy are Nepalis Joining the Russian Army?\nEvidence the accounts were linked to Nepali nationals was found from clues in their TikTok profiles.\nA Nepali speaker reviewed the eight accounts we found, and found they appeared to include people from different regions of Nepal – including from the southern plains that border with India to people from the hills.\nIn several of the accounts we identified Nepali-speaking men with Nepali dialects. Accounts used songs and dialogue in their videos that could either be from Nepal or from India – where there are several Nepali origin communities. In these cases we checked the older posts from the accounts, which indeed appeared to show footage and images taken inside Nepal.\nThe Nepali-language\narticle\nby the aforementioned journalist Birat Anupam for NepalPress offered some suggestions as to the potential motives of Nepalis signing up to serve in the Russian army.\nIt suggests that Nepalis come to Russia as “students and tourists” and subsequently sign up for the country’s military.\nAnupam contacted two Nepali men who had joined the Russian army and spoke to them via Telegram, initially under the guise of being a Nepali youth expecting to join the Russian Army. He was blocked after he told them he was a journalist.\nBoth men spoke about the difficulties of joining Western militaries such as the French Foreign Legion or the military in the US or UK. And both spoke about the ease of joining the Russian army.\n“After one year, we can get citizenship as well. If I don’t die after a year, I will stay here only.” One of the men told NepalPress, but did not want to be named in the report.\nThe article also cited a\nscanned letter\nfrom the Embassy of Nepal in Moscow dated to the 28 of the Nepali month of Bhadra (August-September in the Gregorian calendar).\nIt stated that the number of Nepali students at Russian universities had dramatically increased and advised citizens to be wary of easy student visas.\nHowever, at time of publication, this letter appeared to have disappeared from the Nepali embassy’s website.\nAn\narchived copy\nof the letter can still be accessed. It is not known why the letter appears to have been removed.\nBellingcat contacted Nepal’s Foreign Ministry and Embassy of Nepal in Moscow to ask about the presence of Nepali citizens in the Russian army, how many Nepalis were currently serving in the Russian army and what Nepal’s position is on its citizens joining the Russian army. At time of publication we had not received a reply.\nNepal’s Military Tradition\nThe length of time that the men featured in the TikTok accounts spent in Russia, prior to their appearance in the training camps, varied.\nIn at least one case, we found evidence that pointed to their arrival in Russia as early as November, 2022, when they appeared to be studying Russian. It is not clear how long the other men featured in the TikTok accounts spent in Russia prior to their appearance at the training camps.\nSeveral of the TikTok accounts include hashtags such as #gurkha and #gorkhali. Others use #lahurey, a reference to “those who go to Lahore,” where the British Indian army used to train Nepali soldiers.\nAll these hashtags allude to The Gurkhas, an elite Nepalese group of fighters initially recruited by\nThe British East India Company\nduring British Colonial Rule but who are still recruited to the\nBritish\nand\nIndian armies\ntoday. However, there is no evidence the Nepali soldiers training in Russia are from Gurkha regiments.\nSome Nepali citizens are also known to\ncontribute to foreign militaries\nsuch as the US Army, and Nepalis are a\nsignificant presence in UN\npeacekeeping forces. There is a large workforce of Nepalis who seek\nincome and opportunities outside Nepal\n, providing remittance payments to families at home.\nBellingcat contacted both the Russian and Nepal foreign ministries to ask about Nepali soldiers serving in the Russian army. We also contacted the Embassy of Nepal in Moscow and attempted to reach several Nepali TikTok account holders. At time of publication we had received no response.\nMaxim Edwards, Youri van der Weide, Michael Sheldon, Aric Toler and Dennis Kovtun contributed to this report.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/06/image3.png",
+ "alt": "",
+ "caption": "Stills taken from a TikTok video stitched together shows the training grounds at the Avangard Military Complex."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/06/image12-1-1200x726.png",
+ "alt": "",
+ "caption": "Google Earth Image of Avangard Military Complex from August 2021."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/06/image-11.png",
+ "alt": "",
+ "caption": "A still from a TikTok video, showing three men in army fatigues in front of Avangard sign (we have blurred out faces)."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/06/image-9-1200x960.png",
+ "alt": "",
+ "caption": "Matching lettering seen on the building in this image from Avangard website image gallery (we have blurred out faces). Image from Avangard website."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/06/image2-1-1200x737.png",
+ "alt": "",
+ "caption": "A Google Earth image shows the main building from above."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/06/image8-1-1200x1062.png",
+ "alt": "",
+ "caption": "Left: Still from a TikTok video. Right: A Google Earth image of the barracks of the 254th MRR base in Russia ( 52.7140448467, 32.2691789635 ). Coloured boxes indicate matching features, with the green box around a vehicle in the foreground and buildings highlighted behind this."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/06/image-8-1200x700.png",
+ "alt": "",
+ "caption": "Still from a Tiktok video (we have blurred out faces). Matching features highlighted in coloured boxes."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/06/image-7.png",
+ "alt": "",
+ "caption": "Google Street View image of Maslovka train station ( 51.564140, 39.246172 ). Matching features highlighted in coloured boxes."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/06/image-5-1200x667.png",
+ "alt": "",
+ "caption": "Google Earth image of the same area. Matching features highlighted in coloured boxes."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/06/image-10-1200x674.png",
+ "alt": "",
+ "caption": "Stitched together stills from footage from a post on TikTok, we identified a number of objects in the footage that we matched to a station in the Belarusian city of Braranavichy."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/06/image9-1-1200x669.png",
+ "alt": "",
+ "caption": "Google Earth Image with coloured boxes indicating objects and structures we matched to the location in the TikTok video."
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2023-07-13_more-than-mountaineering-using-peakvisor-for-geolocation.json b/bellingcat_geolocation_json/2023-07-13_more-than-mountaineering-using-peakvisor-for-geolocation.json
new file mode 100644
index 0000000000000000000000000000000000000000..7aedeb8aae73e489f53d714e8a5b4e4dddf6554d
--- /dev/null
+++ b/bellingcat_geolocation_json/2023-07-13_more-than-mountaineering-using-peakvisor-for-geolocation.json
@@ -0,0 +1,127 @@
+{
+ "slug": "more-than-mountaineering-using-peakvisor-for-geolocation",
+ "title": "More than Mountaineering: Using PeakVisor for Geolocation",
+ "url": "https://www.bellingcat.com/resources/2023/07/13/more-than-mountaineering-using-peakvisor-for-geolocation/",
+ "date": "2023-07-13T13:11:50+00:00",
+ "modified": "2023-07-13T13:46:05+00:00",
+ "author": "Sophie Tedling",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "An online tool originally for mountaineers has played a growing role in geolocations. Here’s how to put PeakVisor to use in open source research.",
+ "content_html": "
\n
A short, shaky video and a single photograph show mourners standing on a mountainside as a row of bodies lie shrouded, about to be buried. Both images appear to be taken from close viewpoints and at nearly the same time — 17:53 UTC on May 28, 2022. But where exactly were they taken?
\n
Left: The still image of the mountain burial (bodies pixelated by Bellingcat). Right: A composite panorama spliced from video and sharpened using Photopea.\n
While the first Telegram posts have since been deleted, these images and a video can still be found online as they had been shared by other Tajik Telegram users shortly after the initial posts.
Accompanying text on the Telegram post by Pamir Daily News which contained these images claimed that the burial took place in the Rushon District of Tajikistan, and that the dead had been killed by government troops during a crackdown in the remote Gorno-Badakhshan autonomous region in the east of the country.
There are a few further clues. As the mourners are dressed in anoraks and trainers rather than climbing gear, it is likely that the paths to this location were not extremely steep. This suggests a location in the foothills, rather than higher up the mountains.
But that’s where geolocating this footage stalled. There is no Google Street View in this remote area. The only footage available on Mapillary points down the main Pamir Highway from the perspective of a car travelling in one direction. Google Earth’s “ground-level view” rendered the real-life jagged terrain as too smoothed to show distinctive rock features.
In this type of landscape, PeakVisor is a valuable tool. This app was originally developed to help mountaineers and hikers orientate, navigate and share geographic information on their smartphones. It has come to play a growing role in open source investigations, confirming geolocations in Bellingcat’s stories as far afield as Armenia, Ethiopia, Italy and Mexico.
\n
PeakVisor offers a number of functions, including a 3D panorama allowing users to place themselves at any coordinates of their choice to scan a highly accurately rendered 360 degree landscape. A 2D contour map allows users to view the inclines of all natural slopes in a selected region, be that in a rural or urban environment.
\n
The tool’s PhotoFit feature allows users to overlay the silhouettes of any skyline or natural landmark onto a photograph, allowing for comparison and therefore verification of the location of the image. However, it must be noted that PeakVisor is generally best used for rendering natural geographic features. It doesn’t recognise structures such as buildings, although it often displays man-made entities such as canals or slag heaps.
\n
Using PeakVisor’s ‘PhotoFit’ feature to overlay a red silhouette of the mountain from a PeakVisor rendered panorama onto a photograph of the Matterhorn mountain in the Alps. The two arched diagonal lines are part of the PeakVisor interface and represent the trails of the sun (yellow) and moon (purple).\n
Crucially, PeakVisor also renders some of the most detailed and accurately shaded virtual landscapes currently available.
\n
To distinguish between the multiple peaks and valleys in regions such as the Pamir Mountains, as much silhouette detail as possible is needed, showing against either the skyline or contrasting terrain. Below is a comparison between a photograph of a known location in the Rushon Valley (centre), and a 3D rendered panorama of the same location in Peakvisor (left) and Google Earth Pro (right).
\n
As you can see, PeakVisor is clearly the more detailed option.
\n
PeakVisor (left hand image) and Google Earth pro (right hand image) produce rendered representations of a photograph of the Rushon valley (centre), the floor of which is around 2000m above sea level.\n
PeakVisor’s terrain shading and topology rendering algorithms make the valley floor’s raised road and grassy river basin distinct, and they adjust both seasonally and for time of day. Google Earth Pro’s rendering is far more homogenous, in both shading and in flattened topography. Its rendered 3D panoramas in the area represent the valley floor as flat and the terrain as evenly textured right up to the foot of the mountains on either side. PeakVisor also provides a much more defined skyline. Google’s algorithms produce smoothed skylines which aren’t consistent with its own photographs of the jagged peaks of the general area.
\n
Here’s how PeakVisor helped solve this geolocation from the burial in Tajikistan.
\n
Clues From Contours
\n
Let’s look at the imagery from the presumed burial.
\n
We are searching for a location on a slope near the base of the mountains, with some turf and a little earth visible, possibly also a path. Given the Rushon district is spread over a wide area, we’ll need to figure out where within this area is best to start looking.
\n
In many investigations, a tip-off or information posted on social media channels can provide further contextual details as well as help narrow down a large search area. This is what happened in this instance, with posts suggesting the burial had taken place to the east of Rushon village.
\n
This would make sense given the local custom is to bury the dead within a few hours and earlier videos of bodies covered in shrouds could be geolocated to Rushon village. Any burial therefore was unlikely to have taken place too far from where the shrouded bodies were pictured.
\n
The tip as to the burial being east of Rushon village allowed us to create an area of interest to hone our search.
\n
Given the source images show people walking up steeply slanting ground, we can already narrow down within the search area east of Rushon to regions with similar slopes.
\n
A map showing a narrowed down search area east of Rushon village.\n
PeakVisor’s 2D map window (which can be seen in the top left of the app or web platform) helps with this as it displays contour lines set at 10 metres apart in elevation. This allows us to see where hills and peaks begin to rise and fall, at what gradient and at what angle.
\n
Contour lines in the 2D Map window in PeakVisor for the Centre of the Search Area. It is important to note that the contour lines are grey, while white lines are roads and the red lines footpaths, all of which are detailed within PeakVisor.\n
As we can see several areas of elevation in the burial image, we can begin to compare these two sources to identify potential search areas.
\n
By annotating a screenshot of the area (which has to be done outside of the PeakVisor app as it doesn’t yet have this functionality), we can begin to establish a boundary for our search by excluding areas which are too high or steep (orange) or too low or flat (purple) compared to what can be seen in the burial image. In this instance, we are looking for sloping ground with a view uphill to a skyline of peaks.
\n
Left: Reduced Search Area Taking a Probabilistic Approach – searching green areas first.\n
The green areas show sites of potential interest with the black arrows suggesting the viewpoint the camera could have had in the burial image.
\n
By looking within these prioritised search areas, we now have a strategy of how to use Peakvisor’s 3D Panorama function (described below) to test our hypothesis, potentially matching the platform’s virtual panoramas to what can be seen in the source image. It should be noted, though, that while this method can sound simple it can often entail virtually visiting lots of sites and painstakingly attempting to make visual comparisons between the peaks and landscapes between PeakVisor and in the source image. This is especially the case in mountainous districts like Rushan because the view is more likely to change rapidly over relatively small distances than in flatlands.
\n
Nevertheless, this targeted approach is significantly more manageable than searching the entire Rushon District and hoping to find a match.
\n
Building a Panorama
\n
PeakVisor’s 3D Panorama function allows us to view the landscape from any chosen set of coordinates. It also includes a dynamic location feature which re-renders the resulting image in almost real time as the user moves along the virtual landscape.
\nUsing PeakVisor’s 3D Panorama Function
\n
The 3D Panorama function in PeakVisor allows users to place themselves at a set of coordinates of their choosing and to see a high resolution virtual view of the surrounding landscape.
\n
Obviously, this is useful in situations when investigators are trying to geolocate footage in mountainous or hilly areas. But it is also worth noting that PeakVisor can be useful in other areas with distinctive natural features.
\n
To utilise this function, a user can either type precise coordinates, or the name of a town or peak that they wish to search for, in the PeakVisor search bar. If precise coordinates are not available, it is possible to simply drag and drop the map pin to a suspected area of interest.
\n
A screen grab shows the scene after the village of Rushon is entered into the PeakVisor search bar.\n
Note that PeakVisor may display local versions of place names in the 2D map, e.g. Рушон is the Cyrillic version of “Rushon”.
\n
For example, if we wanted to begin our 3D search in one of the green areas near Rushon identified in our aforementioned strategy we would drag and drop the 2D map pin to our area of interest. If we knew the coordinates or name of a place we wanted to go to, we could also type those in and then press the “Teleport” button to change the PeakVisor viewpoint to that location.
\n
In the example below we drag the 2D pin to an area to the east of Rushon, looking on to one of the potential search areas we highlighted earlier with black arrows.
\n
The 2D map (top left) pin is seen after being dragged to an area east of Rushon.\n
In this instance, it was not possible to match the peaks or identifying features to the source image so we can cross this area off the list, go back to our search strategy and enter in a new set of coordinates. Or, alternatively, we can drag the cursor on the 2D map to our next site of interest. Some users may find the 2D map (which can be seen in the top left of the GIF below) too small. If that’s the case it is possible to maximise the 2D map window. The 3D Panorama will then regenerate with the view from the new location, as can be seen in the GIF below.
\n
\n
\n
With this functionality in mind, a set of digital search routes was plotted to best traverse the prioritised green search areas defined in our strategy. The direction of the search routes was chosen so as to cross the maximum number of contour lines at least once, since altitude changes were likely to alter the visible panorama landscape most rapidly.
\n
The goal was to find a vantage point where key geographic identifying features matched those seen in the imagery and video from the burial scene. These were the ‘warty’ peak at the top centre (outlined in yellow below), the foreground diagonal cross-slope (purple) and the ‘snaggletooth’ ridge on the right of the image (red).
\n
Identifying features of mountains near the burial site in our source image are highlighted.\n
A potential image match was found on the edge of one of the green highlighted areas towards the east at coordinates 37.943078, 71.609766 in the Bartang Valley.
\n
Helpfully, PeakVisor’s 2D map (inset on the left, below) shows both the location and the direction of the virtual image viewpoint (the small white circle with compass pointer seen in the inset map). The red arrow was added by Bellingcat.
The ‘warty’ peak, cross-slope and ‘snaggletooth’ ridge can all be seen from this vantage point.
\n
\n
A view and comparison of the source image and a location identified on PeakVisor.\n
The video composite image was the first to be matched, because it contained the distinctive mountain peak. Nevertheless, the still photograph was easily matched as the location was nearby, although the camera pitch was lower.
Let’s check the same location with Google Earth Pro’s panorama function.
\n
The image below compares the equivalent Google view of this location (right) with the PeakVisor panorama (left) and the composite view of the location of the burial (centre). It is clear that the PeakVisor panorama is rendered with much more detail than Google, justifying the use of PeakVisor in this instance.
\n
PeakVisor (left hand image) and Google Earth pro (right hand image) produce rendered representations of our source image (centre).\n
Near or Far?
\n
This match could be further confirmed by selecting distinct landmarks in PeakVisor’s 3D virtual panorama and measuring the distance from each to the presumed location. PeakVisor displays the names of mountains in the panorama, stating their distance from the viewpoint in kilometres. These labels can be switched on and off using SHIFT + L for ease of use, whilst SHIFT + P similarly toggles a useful centre screen ‘crosswire’ guide for accurate pixel selection within the panorama (a full list of PeakVisor keyboard shortcuts is detailed below).
This distance check confirms whether the landmarks seen in the panorama were either small and close by or large and further away. This cannot be discerned reliably in rendered panoramas and is strongly affected by the horizontal field of view setting in the system – a feature which will not be discussed in detail in this practical guide but will be covered in future Bellingcat training materials.
As can be seen below, Hanim Mountain is shown as 17.3 kilometres away from the position.
\n
In the composite screenshot above, several landmarks are shown for distance checking in both the 2D map and the 3D panorama. In fact, PeakVisor would only display one landmark at a time, and these have been superimposed onto a single graphic.\n
We can double check this by measuring the distance from the coordinates of our location to Hanim Mountain on the likes of Google Maps, which offers this functionality, or a printed map with scale.
\n
The distance between our site of interest and a peak visible in the background is measured on Google Maps.\n
The feature which measures distance between points within the 3D panorama is not available in Google Earth’s equivalent “ground level view”.
\n
PeakVisor also offers a Teleport function which can be used to change the viewpoint of the panorama to that of any selected pixel in the entire 3D panorama.
\n
Making the Photo Fit
\n
One way to be sure that we have located the correct spot is to overlay PeakVisor’s virtual view at this location onto our source imagery – in this case, the picture of the burial.
\n
We can do this by uploading our source image to PeakVisor. Simply select the import photo option atop the right of the menu bar.
\n
Users can import a photo into PeakVisor via the highlighted “import photo” option.\n
You will then be prompted to enter the coordinates of the location that you want to compare to your image.
\n
By default, PeakVisor will use the location you have most recently been looking at so it is possible to simply press ‘Apply’ to initiate the comparison.
\n
Users can either import their own coordinates or automatically select the site they have been looking at most recently.\n
We are then presented with a rendering of the panorama from our selected location overlaid onto our source image.
\n
An outline of the surrounding hills and mountains is added to the uploaded photo at the suspected burial site.\n
In almost all circumstances user adjustment of several parameters is needed to align the panorama and image. This is because, even at the same exact location, several other parameters need to be matched between the way the image was photographed and the way the panorama was generated. For example, if the photographer was standing on a slope or at an angle that would be reflected in how the image appeared. A user would therefore need to manually adjust the rotation of the panorama to match using the on-screen arrows. Similarly, the photographer’s camera lens is set to a specific angle of view, and the user can move the scroll wheel to vary this property in PeakVisor to try and match that.
\n
In such scenarios, users can alter the alignment of the image by clicking on the arrows displayed on screen and dragging up, down, left or right. This process is displayed in the GIF below.
\n
\n
To get the best contrast between the image and the overlay, we can press SHIFT + C to flick through colour options for the overlay lines (which also applies to contour lines in the 2D view). Below it can be seen in red.
\n
The colour of overlay lines can be changed to allow users to best match up the image to the scene.\n
Satellite Confirmation
\n
To obtain further confirmation and accuracy, these coordinates were submitted to the Planet satellite imaging service to obtain satellite imagery around the date claimed for the burial.
Planet’s satellite images of the area were taken multiple times a week. An area of ground near the coordinates found in PeakVisor had changed between May 18 and May 25, 2022. Satellite imagery from May 23 showed that a patch of land on the mountain near the road had seemingly been disturbed. By May 25 this patch had slowly settled back into homogeneity.
\n
It is not possible from open sources alone to establish the cause of this disturbed ground by looking at satellite imagery alone. But its appearance around the time of the burial is noteworthy.
\n
A slight change in the composition of some earth near our site of interest was observed on satellite imagery. Credit, for both images, Planet Labs.\n
However, inputting the coordinates of the PeakVisor match into Google Earth showed a distance of around 176 metres from the coordinates of the disturbed earth on satellite imagery.
The characteristics of extreme terrain such as this mean that small movements in viewpoint can result in large variations in PeakVisor’s 3D panorama. An obvious way to check whether to use the satellite-derived location or the PeakVisor-derived location was to compare the panorama views of the two locations in PeakVisor. The location (37.9428, 71.6078) of the disturbed earth on the satellite image was fed back into PeakVisor to see what kind of viewpoint it generated for that location.
\n
The coordinates of the satellite derived location yielded a 3D panorama which was an even closer match than that found earlier.
The mountain peak and distant rock formations were well-matched with the photographic image using PhotoFit at both locations. However, in the foreground slope was far better matched in both size and inclination at the location found with the help of satellite imagery.
\n
With this dual approach, using both PeakVisor and satellite imagery, each method can be used to assess confidence in the results of the other.
\n
Future Functionalities and the POPSI Project
\n
A version of PeakVisor is now being developed specifically to support Open Source investigations in a voluntary collaboration between PeakVisor and the online open source community, called the POPSI Project. The author of this article instigated this work after replying to a message from PeakVisor to its users and introducing the PeakVisor development team to open source investigations as a discipline.
\n
Over the past 18 months, many functionality upgrades have been released through this project for open source work. The most important is the introduction of an option that offers near ‘real time’ location change within the 3D panorama. Using the SHIFT + arrow keys, this allows for fine tuning of viewpoint location. It was not available when this Tajik geolocation was originally completed using PeakVisor, but it was ready for use in time for the satellite imagery location to be checked as part of this article. This may explain the 170m discrepancy between the two methods.
\n
Other functionality upgrades include allowing users to select different colours for optimal contrast by using SHIFT + C. Additionally, a central cross wire is switched on and off using SHIFT + P. Labels can be toggled with SHIFT + L. Sun and moon trails have been added, along with user-specified date and time to colour the panorama. Distance calculation for any user-selected point (SHIFT + Left Mouse Click) delivers important image depth confirmation. A “Teleport” function allows the user to reset the 3D Panorama viewpoint to that of any selected location visible in the panorama image.
\n
A list of PeakVisor keyboard commands.\n
Future work in development includes path elevation plots (helpful for establishing lines of sight between two points), storing user and group data such as locations and paths, and modelling imagery from a camera moving along a user-defined path. Probably most importantly, a ‘goodness of fit’ measure is being developed to provide an automated assessment of a PhotoFit match, alongside the human eye. This is intended as a stepping stone toward automating some PhotoFit searches.
\n
In this work, the Open Source community is having a say in third-party tool development, rather than just repurposing useful tools after release. Everyone in the open source community is welcome to make suggestions and share the resultant improvements and functionality.
\n\n
The POPSI (Peakvisor OPen Source Investigation) Project, instigated by the author, is a collaborative technical initiative between PeakVisor’s developers and the open source community. It has entered its second year of identifying and implementing within the PeakVisor web app those features which support open source investigations. If you use PeakVisor for open source investigations we’d love to hear about it and particularly any suggestions on new functionality to support open source work, so please contact us on Twitter at @PVforOSINV or via this form.
The author would like to confirm that they have no financial or legal interest in PeakVisor nor prior personal relationship with its founders and employees; this promotion of its tool is strictly from the perspective of an open source researcher.
\n
Foxtrot from Bellingcat’s Global Authentication Project (GAP) contributed to this report, alongside Bellingcat’s Aiganysh Aidarbekova, Miguel Ramalho and Johanna Wild.
\n
",
+ "content_text": "A short, shaky video and a single photograph show mourners standing on a mountainside as a row of bodies lie shrouded, about to be buried. Both images appear to be taken from close viewpoints and at nearly the same time — 17:53 UTC on May 28, 2022. But where exactly were they taken?\nLeft: The still image of the mountain burial (bodies pixelated by Bellingcat). Right: A composite panorama spliced from video and sharpened using\nPhotopea\n.\nWhile the first Telegram posts have since been deleted, these\nimages\nand a\nvideo\ncan still be found online as they had been shared by other Tajik Telegram users shortly after the initial posts.\nAccompanying text on the Telegram\npost\nby Pamir Daily News which contained these images claimed that the burial took place in the Rushon District of Tajikistan, and that the dead had been killed by government troops during a crackdown in the remote Gorno-Badakhshan autonomous region in the east of the country.\nThere are a few further clues. As the mourners are dressed in anoraks and trainers rather than climbing gear, it is likely that the paths to this location were not extremely steep. This suggests a location in the foothills, rather than higher up the mountains.\nBut that’s where geolocating this footage stalled. There is no Google Street View in this remote area. The only footage available on\nMapillary\npoints down the main Pamir Highway from the perspective of a car travelling in one direction. Google Earth’s “ground-level view” rendered the real-life jagged terrain as too smoothed to show distinctive rock features.\nIn this type of landscape,\nPeakVisor\nis a valuable tool. This app was originally developed to help mountaineers and hikers orientate, navigate and share geographic information on their smartphones. It has come to play a growing role in open source investigations, confirming geolocations in Bellingcat’s stories as far afield as\nArmenia\n,\nEthiopia\n,\nItaly\nand\nMexico\n.\nPeakVisor offers a number of functions, including a 3D panorama allowing users to place themselves at any coordinates of their choice to scan a highly accurately rendered 360 degree landscape. A 2D contour map allows users to view the inclines of all natural slopes in a selected region, be that in a rural or urban environment.\nThe tool’s PhotoFit feature allows users to overlay the silhouettes of any skyline or natural landmark onto a photograph, allowing for comparison and therefore verification of the location of the image. However, it must be noted that PeakVisor is generally best used for rendering natural geographic features. It doesn’t recognise structures such as buildings, although it often displays man-made entities such as canals or slag heaps.\nUsing PeakVisor’s ‘PhotoFit’ feature to overlay a red silhouette of the mountain from a PeakVisor rendered panorama onto a photograph of the Matterhorn mountain in the Alps. The two arched diagonal lines are part of the PeakVisor interface and represent the trails of the sun (yellow) and moon (purple).\nCrucially, PeakVisor also renders some of the most detailed and accurately shaded virtual landscapes currently available.\nTo distinguish between the multiple peaks and valleys in regions such as the Pamir Mountains, as much silhouette detail as possible is needed, showing against either the skyline or contrasting terrain. Below is a comparison between a photograph of a known location in the Rushon Valley (centre), and a 3D rendered panorama of the same location in Peakvisor (left) and Google Earth Pro (right).\nAs you can see, PeakVisor is clearly the more detailed option.\nPeakVisor (left hand image) and Google Earth pro (right hand image) produce rendered representations of a photograph of the Rushon valley (centre), the floor of which is around 2000m above sea level.\nPeakVisor’s terrain shading and topology rendering algorithms make the valley floor’s raised road and grassy river basin distinct, and they adjust both seasonally and for time of day. Google Earth Pro’s rendering is far more homogenous, in both shading and in flattened topography. Its rendered 3D panoramas in the area represent the valley floor as flat and the terrain as evenly textured right up to the foot of the mountains on either side. PeakVisor also provides a much more defined skyline. Google’s algorithms produce smoothed skylines which aren’t consistent with its own photographs of the jagged peaks of the general area.\nHere’s how PeakVisor helped solve this geolocation from the burial in Tajikistan.\nClues From Contours\nLet’s look at the imagery from the presumed burial.\nWe are searching for a location on a slope near the base of the mountains, with some turf and a little earth visible, possibly also a path. Given the Rushon district is spread over a wide area, we’ll need to figure out where within this area is best to start looking.\nIn many investigations, a tip-off or information posted on social media channels can provide further contextual details as well as help narrow down a large search area. This is what happened in this instance, with posts suggesting the burial had taken place to the east of Rushon village.\nThis would make sense given the local custom is to bury the dead within a few hours and earlier videos of bodies covered in shrouds could be geolocated to Rushon village. Any burial therefore was unlikely to have taken place too far from where the shrouded bodies were pictured.\nThe tip as to the burial being east of Rushon village allowed us to create an area of interest to hone our search.\nGiven the source images show people walking up steeply slanting ground, we can already narrow down within the search area east of Rushon to regions with similar slopes.\nA map showing a narrowed down search area east of Rushon village.\nPeakVisor’s 2D map window (which can be seen in the top left of the app or web platform) helps with this as it displays contour lines set at 10 metres apart in elevation. This allows us to see where hills and peaks begin to rise and fall, at what gradient and at what angle.\nContour lines in the 2D Map window in PeakVisor for the Centre of the Search Area. It is important to note that the contour lines are grey, while white lines are roads and the red lines footpaths, all of which are detailed within PeakVisor.\nAs we can see several areas of elevation in the burial image, we can begin to compare these two sources to identify potential search areas.\nBy annotating a screenshot of the area (which has to be done outside of the PeakVisor app as it doesn’t yet have this functionality), we can begin to establish a boundary for our search by excluding areas which are too high or steep (orange) or too low or flat (purple) compared to what can be seen in the burial image. In this instance, we are looking for sloping ground with a view uphill to a skyline of peaks.\nLeft: Reduced Search Area Taking a Probabilistic Approach – searching green areas first.\nThe green areas show sites of potential interest with the black arrows suggesting the viewpoint the camera could have had in the burial image.\nBy looking within these prioritised search areas, we now have a strategy of how to use Peakvisor’s 3D Panorama function (described below) to test our hypothesis, potentially matching the platform’s virtual panoramas to what can be seen in the source image. It should be noted, though, that while this method can sound simple it can often entail virtually visiting lots of sites and painstakingly attempting to make visual comparisons between the peaks and landscapes between PeakVisor and in the source image. This is especially the case in mountainous districts like Rushan because the view is more likely to change rapidly over relatively small distances than in flatlands.\nNevertheless, this targeted approach is significantly more manageable than searching the entire Rushon District and hoping to find a match.\nBuilding a Panorama\nPeakVisor’s 3D Panorama function allows us to view the landscape from any chosen set of coordinates. It also includes a dynamic location feature which re-renders the resulting image in almost real time as the user moves along the virtual landscape.\nUsing PeakVisor’s 3D Panorama Function\nThe 3D Panorama function in PeakVisor allows users to place themselves at a set of coordinates of their choosing and to see a high resolution virtual view of the surrounding landscape.\nObviously, this is useful in situations when investigators are trying to geolocate footage in mountainous or hilly areas. But it is also worth noting that PeakVisor can be useful in other areas with distinctive natural features.\nTo utilise this function, a user can either type precise coordinates, or the name of a town or peak that they wish to search for, in the PeakVisor search bar. If precise coordinates are not available, it is possible to simply drag and drop the map pin to a suspected area of interest.\nA screen grab shows the scene after the village of Rushon is entered into the PeakVisor search bar.\nNote that PeakVisor may display local versions of place names in the 2D map, e.g. Рушон is the Cyrillic version of “Rushon”.\nFor example, if we wanted to begin our 3D search in one of the\ngreen areas near Rushon identified in our aforementioned strategy\nwe would drag and drop the 2D map pin to our area of interest. If we knew the coordinates or name of a place we wanted to go to, we could also type those in and then press the “Teleport” button to change the PeakVisor viewpoint to that location.\nIn the example below we drag the 2D pin to an area to the east of Rushon, looking on to one of the potential search areas we highlighted earlier with black arrows.\nThe 2D map (top left) pin is seen after being dragged to an area east of Rushon.\nIn this instance, it was not possible to match the peaks or identifying features to the source image so we can cross this area off the list, go back to our search strategy and enter in a new set of coordinates. Or, alternatively, we can drag the cursor on the 2D map to our next site of interest. Some users may find the 2D map (which can be seen in the top left of the GIF below) too small. If that’s the case it is possible to maximise the 2D map window. The 3D Panorama will then regenerate with the view from the new location, as can be seen in the GIF below.\nWith this functionality in mind, a set of digital search routes was plotted to best traverse the prioritised\ngreen search areas defined in our strategy\n. The direction of the search routes was chosen so as to cross the maximum number of contour lines at least once, since altitude changes were likely to alter the visible panorama landscape most rapidly.\nThe goal was to find a vantage point where key geographic identifying features matched those seen in the imagery and video from the burial scene. These were the ‘warty’ peak at the top centre (outlined in yellow below), the foreground diagonal cross-slope (purple) and the ‘snaggletooth’ ridge on the right of the image (red).\nIdentifying features of mountains near the burial site in our source image are highlighted.\nA potential image match was found on the edge of one of the green highlighted areas towards the east at coordinates 37.943078, 71.609766 in the Bartang Valley.\nHelpfully, PeakVisor’s 2D map (inset on the left, below) shows both the location and the direction of the virtual image viewpoint (the small white circle with compass pointer seen in the inset map). The red arrow was added by Bellingcat.\nThe ‘warty’ peak, cross-slope and ‘snaggletooth’ ridge can all be seen from this vantage point.\nA view and comparison of the source image and a location identified on PeakVisor.\nThe video composite image was the first to be matched, because it contained the distinctive mountain peak. Nevertheless, the still photograph was easily matched as the location was nearby, although the\ncamera pitch was lower\n.\nLet’s check the same location with Google Earth Pro’s panorama function.\nThe image below compares the equivalent Google view of this location (right) with the PeakVisor panorama (left) and the composite view of the location of the burial (centre). It is clear that the PeakVisor panorama is rendered with much more detail than Google, justifying the use of PeakVisor in this instance.\nPeakVisor (left hand image) and Google Earth pro (right hand image) produce rendered representations of our source image (centre).\nNear or Far?\nThis match could be further confirmed by selecting distinct landmarks in PeakVisor’s 3D virtual panorama and measuring the distance from each to the presumed location. PeakVisor displays the names of mountains in the panorama, stating their distance from the viewpoint in kilometres. These labels can be switched on and off using SHIFT + L for ease of use, whilst SHIFT + P similarly toggles a useful centre screen ‘crosswire’ guide for accurate pixel selection within the panorama (a full list of PeakVisor keyboard shortcuts is detailed below).\nThis distance check confirms whether the landmarks seen in the panorama were either small and close by or large and further away. This cannot be discerned reliably in rendered panoramas and is strongly affected by the horizontal field of view setting in the system – a feature which will not be discussed in detail in this practical guide but will be covered in future Bellingcat training materials.\nAs can be seen below, Hanim Mountain is shown as 17.3 kilometres away from the position.\nIn the composite screenshot above, several landmarks are shown for distance checking in both the 2D map and the 3D panorama. In fact, PeakVisor would only display one landmark at a time, and these have been superimposed onto a single graphic.\nWe can double check this by measuring the distance from the coordinates of our location to Hanim Mountain on the likes of Google Maps, which\noffers this functionality\n, or a printed map with scale.\nThe distance between our site of interest and a peak visible in the background is measured on Google Maps.\nThe feature which measures distance between points within the 3D panorama is not available in Google Earth’s equivalent “ground level view”.\nPeakVisor also offers a Teleport function which can be used to change the viewpoint of the panorama to that of any selected pixel in the entire 3D panorama.\nMaking the Photo Fit\nOne way to be sure that we have located the correct spot is to overlay PeakVisor’s virtual view at this location onto our source imagery – in this case, the picture of the burial.\nWe can do this by uploading our source image to PeakVisor. Simply select the import photo option atop the right of the menu bar.\nUsers can import a photo into PeakVisor via the highlighted “import photo” option.\nYou will then be prompted to enter the coordinates of the location that you want to compare to your image.\nBy default, PeakVisor will use the location you have most recently been looking at so it is possible to simply press ‘Apply’ to initiate the comparison.\nUsers can either import their own coordinates or automatically select the site they have been looking at most recently.\nWe are then presented with a rendering of the panorama from our selected location overlaid onto our source image.\nAn outline of the surrounding hills and mountains is added to the uploaded photo at the suspected burial site.\nIn almost all circumstances user adjustment of several parameters is needed to align the panorama and image. This is because, even at the same exact location, several other parameters need to be matched between the way the image was photographed and the way the panorama was generated. For example, if the photographer was standing on a slope or at an angle that would be reflected in how the image appeared. A user would therefore need to manually adjust the rotation of the panorama to match using the on-screen arrows. Similarly, the photographer’s camera lens is set to a specific angle of view, and the user can move the scroll wheel to vary this property in PeakVisor to try and match that.\nIn such scenarios, users can alter the alignment of the image by clicking on the arrows displayed on screen and dragging up, down, left or right. This process is displayed in the GIF below.\nTo get the best contrast between the image and the overlay, we can press SHIFT + C to flick through colour options for the overlay lines (which also applies to contour lines in the 2D view). Below it can be seen in red.\nThe colour of overlay lines can be changed to allow users to best match up the image to the scene.\nSatellite Confirmation\nTo obtain further confirmation and accuracy, these coordinates were submitted to the Planet satellite imaging service to obtain satellite imagery around the date claimed for the burial.\nPlanet’s satellite images of the area were taken multiple times a week. An area of ground near the coordinates found in PeakVisor had changed between May 18 and May 25, 2022. Satellite imagery from May 23 showed that a patch of land on the mountain near the road had seemingly been disturbed. By May 25 this patch had slowly settled back into homogeneity.\nIt is not possible from open sources alone to establish the cause of this disturbed ground by looking at satellite imagery alone. But its appearance around the time of the burial is noteworthy.\nA slight change in the composition of some earth near our site of interest was observed on satellite imagery. Credit, for both images, Planet Labs.\nHowever, inputting the coordinates of the PeakVisor match into Google Earth showed a distance of around 176 metres from the coordinates of the disturbed earth on satellite imagery.\nThe characteristics of extreme terrain such as this mean that small movements in viewpoint can result in large variations in PeakVisor’s 3D panorama. An obvious way to check whether to use the satellite-derived location or the PeakVisor-derived location was to compare the panorama views of the two locations in PeakVisor. The location (37.9428, 71.6078) of the disturbed earth on the satellite image was fed back into PeakVisor to see what kind of viewpoint it generated for that location.\nThe coordinates of the satellite derived location yielded a 3D panorama which was an even closer match than that found earlier.\nThe mountain peak and distant rock formations were well-matched with the photographic image using PhotoFit at both locations. However, in the foreground slope was far better matched in both size and inclination at the location found with the help of satellite imagery.\nWith this dual approach, using both PeakVisor and satellite imagery, each method can be used to assess confidence in the results of the other.\nFuture Functionalities and the POPSI Project\nA version of PeakVisor is now being developed specifically to support Open Source investigations in a voluntary collaboration between PeakVisor and the online open source community, called the POPSI Project. The author of this article instigated this work after replying to a message from PeakVisor to its users and introducing the PeakVisor development team to open source investigations as a discipline.\nOver the past 18 months, many functionality upgrades have been released through this project for open source work. The most important is the introduction of an option that offers near ‘real time’ location change within the 3D panorama. Using the SHIFT + arrow keys, this allows for fine tuning of viewpoint location. It was not available when this Tajik geolocation was originally completed using PeakVisor, but it was ready for use in time for the satellite imagery location to be checked as part of this article. This may explain the 170m discrepancy between the two methods.\nOther functionality upgrades include allowing users to select different colours for optimal contrast by using SHIFT + C. Additionally, a central cross wire is switched on and off using SHIFT + P. Labels can be toggled with SHIFT + L. Sun and moon trails have been added, along with user-specified date and time to colour the panorama. Distance calculation for any user-selected point (SHIFT + Left Mouse Click) delivers important image depth confirmation. A “Teleport” function allows the user to reset the 3D Panorama viewpoint to that of any selected location visible in the panorama image.\nA list of PeakVisor keyboard commands.\nFuture work in development includes path elevation plots (helpful for establishing lines of sight between two points), storing user and group data such as locations and paths, and modelling imagery from a camera moving along a user-defined path. Probably most importantly, a ‘goodness of fit’ measure is being developed to provide an automated assessment of a PhotoFit match, alongside the human eye. This is intended as a stepping stone toward automating some PhotoFit searches.\nIn this work, the Open Source community is having a say in third-party tool development, rather than just repurposing useful tools after release. Everyone in the open source community is welcome to make suggestions and share the resultant improvements and functionality.\nThe POPSI (\nP\neakvisor\nOP\nen\nS\nource\nI\nnvestigation) Project, instigated by the author, is a collaborative technical initiative between PeakVisor’s developers and the open source community. It has entered its second year of identifying and implementing within the PeakVisor web app those features which support open source investigations. If you use PeakVisor for open source investigations we’d love to hear about it and particularly any suggestions on new functionality to support open source work, so please contact us on Twitter at @PVforOSINV or via this\nform\n.\nThe author would like to confirm that they have no financial or legal interest in PeakVisor nor prior personal relationship with its founders and employees; this promotion of its tool is strictly from the perspective of an open source researcher.\nFoxtrot from Bellingcat’s\nGlobal Authentication Project (GAP)\ncontributed to this report, alongside Bellingcat’s Aiganysh Aidarbekova, Miguel Ramalho and Johanna Wild.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.40.18.png",
+ "alt": "",
+ "caption": "Left: The still image of the mountain burial (bodies pixelated by Bellingcat). Right: A composite panorama spliced from video and sharpened using Photopea ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.43.33-1200x650.png",
+ "alt": "",
+ "caption": "Using PeakVisor’s ‘PhotoFit’ feature to overlay a red silhouette of the mountain from a PeakVisor rendered panorama onto a photograph of the Matterhorn mountain in the Alps. The two arched diagonal lines are part of the PeakVisor interface and represent the trails of the sun (yellow) and moon (purple)."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.44.59-1200x261.png",
+ "alt": "",
+ "caption": "PeakVisor (left hand image) and Google Earth pro (right hand image) produce rendered representations of a photograph of the Rushon valley (centre), the floor of which is around 2000m above sea level."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.45.23-1200x670.png",
+ "alt": "",
+ "caption": "A map showing a narrowed down search area east of Rushon village."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.45.38.png",
+ "alt": "",
+ "caption": "Contour lines in the 2D Map window in PeakVisor for the Centre of the Search Area. It is important to note that the contour lines are grey, while white lines are roads and the red lines footpaths, all of which are detailed within PeakVisor."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.45.50-1200x386.png",
+ "alt": "",
+ "caption": "Left: Reduced Search Area Taking a Probabilistic Approach – searching green areas first."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.46.09-1200x525.png",
+ "alt": "",
+ "caption": "A screen grab shows the scene after the village of Rushon is entered into the PeakVisor search bar."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.46.26-1200x586.png",
+ "alt": "",
+ "caption": "The 2D map (top left) pin is seen after being dragged to an area east of Rushon."
+ },
+ {
+ "url": "https://lh4.googleusercontent.com/fj_STl0caRbRrPIVNi0Wd8PfZdW5z43y7PTD-ZX_G_ImexlkKz2QCo4Ru7TFXSACxqLY23UhVX736yN1vPXSeVGZjQ7kjNY1jGJRpS2_J6mV4eKyD003jSxUsMArRw-hMUhB_j_2O5d9juzquI55tbI",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.47.29.png",
+ "alt": "",
+ "caption": "Identifying features of mountains near the burial site in our source image are highlighted."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-13-at-09.06.35-1200x569.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-13-at-09.06.44-1200x415.png",
+ "alt": "",
+ "caption": "A view and comparison of the source image and a location identified on PeakVisor."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.47.52-1200x264.png",
+ "alt": "",
+ "caption": "PeakVisor (left hand image) and Google Earth pro (right hand image) produce rendered representations of our source image (centre)."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.48.11-1200x590.png",
+ "alt": "",
+ "caption": "In the composite screenshot above, several landmarks are shown for distance checking in both the 2D map and the 3D panorama. In fact, PeakVisor would only display one landmark at a time, and these have been superimposed onto a single graphic."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.48.23.png",
+ "alt": "",
+ "caption": "The distance between our site of interest and a peak visible in the background is measured on Google Maps."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.48.49-1200x593.png",
+ "alt": "",
+ "caption": "Users can import a photo into PeakVisor via the highlighted “import photo” option."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.49.00-1200x593.png",
+ "alt": "",
+ "caption": "Users can either import their own coordinates or automatically select the site they have been looking at most recently."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.49.14-1200x546.png",
+ "alt": "",
+ "caption": "An outline of the surrounding hills and mountains is added to the uploaded photo at the suspected burial site."
+ },
+ {
+ "url": "https://lh5.googleusercontent.com/Qk1UzoaUXXdnpDdvCxsv75cb4VvrTupyV2XqunUkVBmpkDpSk3VVdgQDPBgWq8ExGcQ3joeaesJwtP4JOxtNqwlJl2hgSRAnwqy2Kxk_UMwY6fOEWWTcPEFUoCM4Hx4Eql-NxfAoAPcpaHDLDuqcNDQ",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.49.26-1200x643.png",
+ "alt": "",
+ "caption": "The colour of overlay lines can be changed to allow users to best match up the image to the scene."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-13-at-10.35.06-1200x696.png",
+ "alt": "",
+ "caption": "A slight change in the composition of some earth near our site of interest was observed on satellite imagery. Credit, for both images, Planet Labs."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-13-at-11.52.27.png",
+ "alt": "",
+ "caption": "A list of PeakVisor keyboard commands."
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2023-07-14_can-ai-chatbots-be-used-for-geolocation.json b/bellingcat_geolocation_json/2023-07-14_can-ai-chatbots-be-used-for-geolocation.json
new file mode 100644
index 0000000000000000000000000000000000000000..7f62cc725cf483cd642c44e2f62e0d3814986db5
--- /dev/null
+++ b/bellingcat_geolocation_json/2023-07-14_can-ai-chatbots-be-used-for-geolocation.json
@@ -0,0 +1,32 @@
+{
+ "slug": "can-ai-chatbots-be-used-for-geolocation",
+ "title": "Can AI Chatbots Be Used for Geolocation?",
+ "url": "https://www.bellingcat.com/resources/2023/07/14/can-ai-chatbots-be-used-for-geolocation/",
+ "date": "2023-07-14T12:13:06+00:00",
+ "modified": "2023-07-17T14:22:00+00:00",
+ "author": "Dennis Kovtun",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Chatbots such as BARD and BING try to mimic the techniques used by open source researchers — with very mixed results.",
+ "content_html": "
\n
Geolocation is one of the main methods of open-source research. Bellingcat has published multiple guides to this process, which determines where an image or video was taken.
Given that geolocations can be time-consuming, researchers are always on the lookout for tools which can ease or automate parts of the process. That’s where new AI tools come in – particularly chatbots, with their impressive ability to find and process information.
But that ability is far from perfect. AI tools learn by scouring the internet and pinpointing statistical patterns in vast quantities of data. “Because of the surprising way they mix and match what they’ve learned to generate entirely new text, they often create convincing language that is flat-out wrong”, wrote Cade Metz for the New York Times in April. “AI researchers call this tendency to make stuff up a ‘hallucination’”, Metz continued.
Bellingcat has tested the geolocation capabilities of Bing AI, which was accessed via the Skype client on the desktop app, and Bard, Google’s new AI chatbot which was recently launched in Brazil and the EU. Both chatbots use different language models; Bard uses PaLM-2 while Bing uses GPT – the same as that used in the popular ChatGPT bot.
While many AIs can generate images, most cannot analyse them and are therefore useless for the purposes of direct geolocation. Unlike AIs such as ChatGPT, Bing and Bard can work with uploaded images.
\n
We found that while Bing mimics the strategies that open-source researchers use to geolocate images, it cannot successfully geolocate images on its own. Bard’s results are not much more impressive, but it seemed more cautious in its reasoning and less prone to AI ‘hallucinations’.
Both required extensive prompting from the user before they could arrive at any halfway satisfactory geolocation.
Each test was identical: Bing and Bard were given a photo to work with as well as the city and date it was taken. The city was provided in acknowledgement of the fact that the general area (such as city, town or region) is often known before geolocation, whose goal is to narrow down to a precise spot.
\n
It was then asked to produce precise coordinates of the image. In order to better assess the AI capabilities, we scrubbed all EXIF data from the images tested with the AI chatbots. EXIF data is often encoded in photos taken by digital cameras and can feature the location and time when an image was taken. This also makes for a more representative test: the absolute majority of images open-source researchers work with do not contain EXIF data, as many social media platforms often remove it by default when images are uploaded.
\n
The First Test
\n
The first image we gave Bing to geolocate is this one. It depicts a very recognisable landmark — the City Hall in Edmonton, capital of the Canadian province of Alberta. The precise coordinates of this image are (53.5448285, 113.4900553). The author took itat 7:12 PM local time on June 13, 2022.
\n
\n
\n
Bing provided a location for this image in just a few seconds, though its answer was incorrect. It claimed that the photo had been taken at the High Level Bridge over the North Saskatchewan River, providing the coordinates “53.532°N 113.499°W” and a link to a corresponding Wikipedia page.
Here the AI leads the researcher astray, as there is clearly no bridge in the source photo. However, it also noted that there were “downtown skyscrapers in the background”. Some of these, such as the very visible CN Tower, are also visible from the High Level Bridge.
\n
We gave Bing a second prompt.
\n
We told AI that the image depicted a building with a pyramid-shaped roof and was taken at a public square.
\n
To this prompt, Bing responded with correct identification of landmarks:
\n
\n
It looks like the Sir Winston Churchill Square in downtown Edmonton with the City Hall in the background.
The City Hall has a distinctive pyramid-shaped roof made of glass.
\n
\n
Based on this geolocation, we can surmise that Bing can identify very prominent landmarks that would be easy to identify manually, but requires extra prompting to do so.
\n
Bard displayed similar results. It was unable to geolocate an image provided without any further hints:
\n
\n
I’ve looked at the image you sent me, and I’m not able to geolocate it. The image doesn’t contain any identifying landmarks, and the date and time it was taken are not visible. I also tried a reverse image search, but I didn’t find any matches
\n
\n
We then told it that the photo was taken in Edmonton, Canada.
\n
Bard could work with that. It gave me a set of recognisable landmarks in Edmonton and asked whether I could identify the buildings in the image in any of them. However, Bard did not name any distinguishing features in the buildings seen in the image, several of which are at some distance from the City Hall.
\n
We prompted it again, drawing its focus towards the most relevant building: “The image has a pyramid-shaped glass roof.”
\n
That’s when Bard was able to correctly identify the landmark:
\n
\n
“Oh, that’s a good detail to know. Based on that, I think the most likely location of the image is City Hall in Edmonton. It has a pyramid-shaped glass roof and is located in the heart of the city.”
\n
\n
The Second Test
\n
For the second geolocation, we gave Bing and Bard another puzzle from Edmonton.
\n
\n
\n
This image was taken on Churchill Square in downtown Edmonton, in front of the City Centre shopping mall.
\n
Several commercial logos can be seen on the building’s facade. The EPCOR Waterfall, an art installation, is visible in the foreground. With all these features, geolocating the image manually should not be a problem. The coordinates of this image are 53.54397674546276, -113.49009302380333. It was taken at 9:36 AM local time on June 8, 2022.
\n
When first prompted, Bing identified the location as the High Level Bridge, like in the first answer.
\n
This incorrect geolocation is potentially indicative of a certain pattern in AI’s thinking: it latched on the skyscrapers in the background as one of the identifiers, and it defaults to the High Level Bridge as a primary landmark.
\n
When we pointed out that no bridge can be seen in the video, Bing then identified the location as Edmonton City Hall. In its reply, it singled out the fact that the City Hall has a “distinctive pyramid-shaped roof made of glass.”
\n
Although there’s clearly no building with a pyramid-shaped roof in the image and the landmark identification is incorrect, Edmonton’s City Hall stands out of sight, approximately 80 metres away on the other side of Churchill Square.
\n
Bard’s reply was cautious. It states that it wasn’t able to “geolocate it with certainty” as “the image doesn’t contain any identifying landmarks.” Bard thinks there’s a large river in the background and that the image was taken in one of the parks along the North Saskatchewan River.
\n
When we drew Bard’s attention to the presence of a large building with antennae on its roof and several skyscrapers in the distance, it incorrectly identified the image as Terwillegar Park. Interestingly, Bard has misplaced a landmark in its identification process:
\n
\n
Based on the information you provided, I think the most likely location for the image is Terwillegar Park. It is a large park with a variety of trails, picnic areas, and a golf course. It is also home to the Edmonton Space and Science Centre, which has a satellite dish on its roof. The skyscrapers in the background are likely the buildings in downtown Edmonton.
\n
Here is a link to the Edmonton Space and Science Centre: https://www.spacecentre.ca/
\n
\n
Edmonton does not have a science centre with this name. The link provided by Bard leads to the H.R. MacMillan Space Centre, which is in Vancouver. Given that Bard saw things in the image that are not there — a river in the background and park’s greenery, as well as a landmark that is located thousands of kilometers away from its chosen location — this geolocation may be a result of AI “hallucination”.
\n
We then told Bard that we could see a sign in the image that said ‘EPCOR Waterfall’. It responded that “I’m a language model and don’t have the capacity to help with that.”
\n
We then tested the AI’s abilities further by describing the shape of the shopping centre building and the commercial logos on its facade. Such symbols are often important first clues when geolocating any image. However, Bing and Bard were not able to identify the building.
We can surmise that these AI applications cannot identify commercial logos. Further tests also suggest this: Bing and Bard were provided with separate, clear images of the logos of the Canadian Broadcasting Corporation (CBC) and the Toronto Dominion Bank, both of which are visible on the shopping centre in red and green respectively.
\n
Bing identified the CBC logo as that of Stack Overflow or Netflix; Bard believed it to be the logo of Adobe. Bing identified the bank’s logo as belonging to Mozilla and Microsoft; Bard claimed that it was that of Spotify.
\n
The Third Test
\n
Next, we gave AI several other images to work with. Like the two photos above, it’s not difficult to geolocate them manually. They contain easily identifiable attributes, but no prominent landmarks.
\n
\n
\n
The coordinates for this image are 45.4263835, -75.6910384, at the intersection of Rideau and William Street in Ottawa. Dennis Kovtun took it on June 15, 2023, at 4:04 pm.
\n
We asked Bing AI to geolocate this image, providing it with the city where it was taken and the date. We also informed the AI that the image didn’t contain any usable EXIF data.
\n
In its first response, Bing AI immediately shared its methodology. It said that it conducts geolocations using these methods:
\n
\n
\n
Using reverse image search to find similar or matching images online and see if they have any location information, such as captions, tags, or links.
\n
Using Google Maps or Google Earth to compare the image with the map images and look for any landmarks or features that match the image, such as buildings, streets, parks, or bridges.
\n
Using metadata or EXIF data to extract any location information that may be embedded in the image file, such as GPS coordinates, date, time, or camera settings.
\n
Using clues or hints from the image itself, such as signs, flags, languages, license plates, or weather.
\n
\n\n
\n
As we can see here, Bing AI uses the same steps that a human researcher could reasonably follow to geolocate an image. This behaviour is expected of AI, as a self-teaching software. It appears to have searched the Internet for geolocation methodologies published on various open-source research websites, and produced subsequent responses as if it had followed them.
How well it is able to do these steps is another question.
\n
For example, the AI analysed the image for EXIF data, though it was told no EXIF data was associated with it. Bing AI generated a very detailed set of results, including the camera model and its exposure settings, the time it was taken and even that it had the flash off. Such details are impossible to obtain from a photo that has no EXIF data associated with it.
\n
This analysis generated by Bing AI is another example of an AI ‘hallucination’. The only EXIF data associated with the image is the file name. The camera model and settings are all incorrect and cannot be determined from the image.
\n
Bing AI also said it performed a reverse image search, providing the following URL:
The AI claimed that the link contains a photo album with pictures of some of Ottawa’s landmarks and a map that shows where the images were taken.
\n
AI concluded that part of its search by stating,
\n
\n
I can see that there is an image that looks very similar to mine, and it has a pin on the map that shows its location. I can click on the pin and see its GPS coordinates, which are 45.4248° N, 75.6926° W. Therefore, using reverse image search, I can deduce that the image was taken at 45.4248° N, 75.6926° W in Ottawa, Canada.
\n
\n
This geolocation is wrong, but is not too far from the actual coordinates — it shows the entrance to Westin hotel, just a few hundred metres away from the correct location.
\n
However, the way AI has arrived at these coordinates is far more suspect. The Flickr image the AI linked to doesn’t exist. Inputting the link in a browser produced a 404 error. This link also doesn’t appear in the Wayback Machine, a digital archive of old internet content.
\n
It is possible that this is yet another example of AI “hallucination,” where Bing generated a link that doesn’t exist or relied on an old dataset of links.
\n
In this instance, it is impossible to determine which images Bing AI used for reference when performing this geolocation, and if it in fact used them.
Bing also says that would use either Google Maps or Google Earth to perform the geolocation — another widespread practice amongst OSINT researchers. It would look for any landmarks that would aid geolocation.
\n
\n
I can see that my image shows a part of downtown Ottawa near Parliament Hill and Rideau Canal. I can zoom in on this area and see if there are any labels or markers that show the GPS coordinates of any point on the map. I can also use the right-click menu to select “What’s here?” or “Measure distance” to get the GPS coordinates of any point on the map. I can try to find a point on the map that corresponds to where the image was taken from, such as the center of the road or the edge of the sidewalk. For example, I can right-click on a point near the center of Wellington Street and select “What’s here?”. I can see a pop-up window that shows its GPS coordinates, which are 45.4248° N, 75.6926° W.
\n
\n
But the AI doesn’t explain, for instance, how it determined that the location is in the vicinity of Parliament Hill and Rideau Canal. The AI deduced that correctly, but didn’t share its reasoning. As our second test demonstrated, it’s unlikely that Bing was able to use commercial signage for this purpose.
Finally, the AI looked for identifying attributes in the image. It responded:
\n
\n
I can see that there are some clues or hints in this image, such as:
\n
• The sign on the left side of the image says “Parliament Hill”, which is a landmark in Ottawa.
\n
• The flag on top of Parliament Hill is a Canadian flag with a red maple leaf in the center.
\n
• The language on some of the signs is English with French translations.
\n
• The license plates on some of the cars are white with blue letters and numbers and have “Ontario” and “Yours to discover” written on them.
\n
• The weather is sunny and clear, with some clouds in the sky…
\n
Using clues or hints from the image itself, I can only deduce the general location of the image, but not the exact GPS coordinates.
\n
\n
While the AI failed to produce a geolocation based on this method, it’s worth examining the “hints” it identified.
\n
For a start, there’s no sign on the left side of the image pointing towards Parliament. In fact, parliament itself is not pictured, and neither are any other buildings or landmarks atop the Parliament Hill, let alone those flying a Canadian flag.
There are only two visible street signs in the image — a street map pictured in the image, which is too far to see clearly, and the public transport sign above it (red circle on a pillar). There is a vehicle in the image with an Ontario licence plate, which can be read clearly, but the AI already knows that the image has been taken in the city located in that province.
\n
Hence, it appears that the AI is “hallucinating” here as well, and its findings are guesses based upon the broad location of the image: Ottawa. Parliament Hill is the most recognisable landmark in Ottawa. Ottawa is located in Ontario. Based on these factors, the AI “saw” these characteristics in the image, even though they weren’t there or were extremely difficult to recognise. Thus Bing generates responses based on its search for whatever it considers relevant content on the Internet.
\n
The results of Bard’s efforts with this image did not lead to its successful geolocation. Once again, this AI seemed more cautious and aware of its limitations.
\n
Even before we uploaded the image, Bard’s AI gave us a list of famous landmarks in Ottawa: Parliament Hill, the Rideau Canal, the National Gallery of Canada, the Canadian War Museum and the ByWard Market.
\n
Bard also identified the image as Parliament Hill.
\n
We prompted Bard again to correct it, pointing out that there was no distinctive clocktower in the photo, which showed cafes and shops.
\n
Bard’s AI readily accepted the correction and provided a list of other possible locations. These included the ByWard Market, Sparks Street, Bank Street and Elgin Street — all commercial areas with plenty of restaurants and cafes.
So after much prompting, Bard’s AI reached the correct location – the ByWard Market. It was nevertheless unable to further refine the geolocation, which is less helpful given the large size of the market and surrounding streets. Thus, even these results are little assistance in geolocating an image quickly, easily and precisely. We were also only able to determine that the answer Bard provided was somewhat correct only because we already knew the answer and we knew the area well. In a situation where the location of the image is unknown, whatever Bard produces is likely to be even less helpful.
\n
A Mimic with Limits
\n
These examples demonstrate that Bing and Bard struggle with analysing images and are prone to seeing details which are not there. They also suggest that the AI chatbots we tested imitate the methods of human open source researchers. This could be partially responsible for their poor performance.
Geoffrey Hinton, a British-Canadian computer scientist and AI specialist, believes that such “confabulations” (his preferred term for AI ‘hallucinations’) are a feature of AI learning models.
\n
“Confabulation is a signature of human memory. These models are doing something just like people,” he said in an interview with MIT Technology review this May.
Using an AI chatbot to fully geolocate an image is inadvisable. At this stage of AI’s development, it might be used to assist with very simple geolocations, perhaps pointing a researcher to an area that may warrant a closer look. However, even such results need to be double-checked and verified and cannot be fully trusted.
\n\n
All photographs were taken by Dennis Kovtun
\n
",
+ "content_text": "Geolocation is one of the main methods of open-source research. Bellingcat has published\nmultiple guides\nto this process, which determines where an image or video was taken.\nGiven that geolocations can be time-consuming, researchers are always on the lookout for tools which can ease or automate parts of the process. That’s where new AI tools come in – particularly chatbots, with their impressive ability to find and process information.\nBut that ability is far from perfect. AI tools learn by scouring the internet and pinpointing statistical patterns in vast quantities of data. “Because of the surprising way they mix and match what they’ve learned to generate entirely new text, they often create convincing language that is flat-out wrong”,\nwrote\nCade Metz for the New York Times in April. “AI researchers call this tendency to make stuff up a ‘hallucination’”, Metz continued.\nBellingcat has tested the geolocation capabilities of Bing AI, which was accessed via the Skype client on the desktop app, and Bard, Google’s new AI chatbot which was recently\nlaunched\nin Brazil and the EU. Both chatbots use different language models; Bard uses PaLM-2 while Bing uses GPT – the same as that used in the popular ChatGPT bot.\nWhile many AIs can generate images, most cannot analyse them and are therefore useless for the purposes of direct geolocation. Unlike AIs such as ChatGPT, Bing and Bard can work with uploaded images.\nWe found that while Bing mimics the strategies that open-source researchers use to geolocate images, it cannot successfully geolocate images on its own. Bard’s results are not much more impressive, but it seemed more cautious in its reasoning and less prone to AI ‘hallucinations’.\nBoth required extensive prompting from the user before they could arrive at any halfway satisfactory geolocation.\nEach test was identical: Bing and Bard were given a photo to work with as well as the city and date it was taken. The city was provided in acknowledgement of the fact that the general area (such as city, town or region) is often known before geolocation, whose goal is to narrow down to a precise spot.\nIt was then asked to produce precise coordinates of the image. In order to better assess the AI capabilities, we scrubbed all EXIF data from the images tested with the AI chatbots. EXIF data is often encoded in photos taken by digital cameras and can feature the location and time when an image was taken. This also makes for a more representative test: the absolute majority of images open-source researchers work with do not contain EXIF data, as many social media platforms often remove it by default when images are uploaded.\nThe First Test\nThe first image we gave Bing to geolocate is this one. It depicts a very recognisable landmark — the City Hall in Edmonton, capital of the Canadian province of Alberta. The precise coordinates of this image are (53.5448285, 113.4900553). The author took it\nat 7:12 PM local time on June 13, 2022.\nBing provided a location for this image in just a few seconds, though its answer was incorrect. It claimed that the photo had been taken at the High Level Bridge over the North Saskatchewan River, providing the coordinates “53.532°N 113.499°W” and a link to a corresponding\nWikipedia page\n.\nHere the AI leads the researcher astray, as there is clearly no bridge in the source photo. However, it also noted that there were “downtown skyscrapers in the background”. Some of these, such as the very visible CN Tower, are also visible from the High Level Bridge.\nWe gave Bing a second prompt.\nWe told AI that the image depicted a building with a pyramid-shaped roof and was taken at a public square.\nTo this prompt, Bing responded with correct identification of landmarks:\nIt looks like the Sir Winston Churchill Square in downtown Edmonton with the City Hall in the background.\nhttps://exploreedmonton.com/attractions-and-experiences/city-hall\nThe City Hall has a distinctive pyramid-shaped roof made of glass.\nBased on this geolocation, we can surmise that Bing can identify very prominent landmarks that would be easy to identify manually, but requires extra prompting to do so.\nBard displayed similar results. It was unable to geolocate an image provided without any further hints:\nI’ve looked at the image you sent me, and I’m not able to geolocate it. The image doesn’t contain any identifying landmarks, and the date and time it was taken are not visible. I also tried a reverse image search, but I didn’t find any matches\nWe then told it that the photo was taken in Edmonton, Canada.\nBard could work with that. It gave me a set of recognisable landmarks in Edmonton and asked whether I could identify the buildings in the image in any of them. However, Bard did not name any distinguishing features in the buildings seen in the image, several of which are at some distance from the City Hall.\nWe prompted it again, drawing its focus towards the most relevant building: “The image has a pyramid-shaped glass roof.”\nThat’s when Bard was able to correctly identify the landmark:\n“Oh, that’s a good detail to know. Based on that, I think the most likely location of the image is City Hall in Edmonton. It has a pyramid-shaped glass roof and is located in the heart of the city.”\nThe Second Test\nFor the second geolocation, we gave Bing and Bard another puzzle from Edmonton.\nThis image was taken on Churchill Square in downtown Edmonton, in front of the City Centre shopping mall.\nSeveral commercial logos can be seen on the building’s facade. The EPCOR Waterfall, an art installation, is visible in the foreground. With all these features, geolocating the image manually should not be a problem. The coordinates of this image are 53.54397674546276, -113.49009302380333. It was taken at 9:36 AM local time on June 8, 2022.\nWhen first prompted, Bing identified the location as the High Level Bridge, like in the first answer.\nThis incorrect geolocation is potentially indicative of a certain pattern in AI’s thinking: it latched on the skyscrapers in the background as one of the identifiers, and it defaults to the High Level Bridge as a primary landmark.\nWhen we pointed out that no bridge can be seen in the video, Bing then identified the location as Edmonton City Hall. In its reply, it singled out the fact that the City Hall has a “distinctive pyramid-shaped roof made of glass.”\nAlthough there’s clearly no building with a pyramid-shaped roof in the image and the landmark identification is incorrect, Edmonton’s City Hall stands out of sight, approximately 80 metres away on the other side of Churchill Square.\nBard’s reply was cautious. It states that it wasn’t able to “geolocate it with certainty” as “the image doesn’t contain any identifying landmarks.” Bard thinks there’s a large river in the background and that the image was taken in one of the parks along the North Saskatchewan River.\nWhen we drew Bard’s attention to the presence of a large building with antennae on its roof and several skyscrapers in the distance, it incorrectly identified the image as Terwillegar Park. Interestingly, Bard has misplaced a landmark in its identification process:\nBased on the information you provided, I think the most likely location for the\nimage is Terwillegar Park. It is a large park with a variety of trails, picnic\nareas, and a golf course. It is also home to the Edmonton Space and Science\nCentre, which has a satellite dish on its roof. The skyscrapers in the\nbackground are likely the buildings in downtown Edmonton.\nHere is a link to the Edmonton Space and Science Centre:\nhttps://www.spacecentre.ca/\nEdmonton does not have a science centre with this name. The link provided by Bard leads to the H.R. MacMillan Space Centre, which is in Vancouver. Given that Bard saw things in the image that are not there — a river in the background and park’s greenery, as well as a landmark that is located thousands of kilometers away from its chosen location — this geolocation may be a result of AI “hallucination”.\nWe then told Bard that we could see a sign in the image that said ‘EPCOR Waterfall’. It responded that “I’m a language model and don’t have the capacity to help with that.”\nWe then tested the AI’s abilities further by describing the shape of the shopping centre building and the commercial logos on its facade. Such symbols are often important first clues when geolocating any image. However, Bing and Bard were not able to identify the building.\nWe can surmise that these AI applications cannot identify commercial logos. Further tests also suggest this: Bing and Bard were provided with separate, clear images of the logos of the Canadian Broadcasting Corporation (CBC) and the Toronto Dominion Bank, both of which are visible on the shopping centre in red and green respectively.\nBing identified the CBC logo as that of Stack Overflow or Netflix; Bard believed it to be the logo of Adobe. Bing identified the bank’s logo as belonging to Mozilla and Microsoft; Bard claimed that it was that of Spotify.\nThe Third Test\nNext, we gave AI several other images to work with. Like the two photos above, it’s not difficult to geolocate them manually. They contain easily identifiable attributes, but no prominent landmarks.\nThe coordinates for this image are 45.4263835, -75.6910384, at the intersection of Rideau and William Street in Ottawa. Dennis Kovtun took it on June 15, 2023, at 4:04 pm.\nWe asked Bing AI to geolocate this image, providing it with the city where it was taken and the date. We also informed the AI that the image didn’t contain any usable EXIF data.\nIn its first response, Bing AI immediately shared its methodology. It said that it conducts geolocations using these methods:\nUsing reverse image search to find similar or matching images online and see if they have any location information, such as captions, tags, or links.\nUsing Google Maps or Google Earth to compare the image with the map images and look for any landmarks or features that match the image, such as buildings, streets, parks, or bridges.\nUsing metadata or EXIF data to extract any location information that may be embedded in the image file, such as GPS coordinates, date, time, or camera settings.\nUsing clues or hints from the image itself, such as signs, flags, languages, license plates, or weather.\nAs we can see here, Bing AI uses the same steps that a human researcher could reasonably follow to geolocate an image. This behaviour is expected of AI, as a self-teaching software. It appears to have searched the Internet for geolocation methodologies published on various open-source research websites, and produced subsequent responses as if it had followed them.\nHow well it is able to do these steps is another question.\nFor example, the AI analysed the image for EXIF data, though it was told no EXIF data was associated with it. Bing AI generated a very detailed set of results, including the camera model and its exposure settings, the time it was taken and even that it had the flash off. Such details are impossible to obtain from a photo that has no EXIF data associated with it.\nThis analysis generated by Bing AI is another example of an AI ‘hallucination’. The only EXIF data associated with the image is the file name. The camera model and settings are all incorrect and cannot be determined from the image.\nBing AI also said it performed a reverse image search, providing the following URL:\nhttps://www.flickr.com/photos/ottawatourism/albums/72157625865960107\nThe AI claimed that the link contains a photo album with pictures of some of Ottawa’s landmarks and a map that shows where the images were taken.\nAI concluded that part of its search by stating,\nI can see that there is an image that looks very similar to mine, and it has a pin on the map that shows its location. I can click on the pin and see its GPS coordinates, which are 45.4248° N, 75.6926° W. Therefore, using reverse image search, I can deduce that the image was taken at 45.4248° N, 75.6926° W in Ottawa, Canada.\nThis geolocation is wrong, but is not too far from the actual coordinates — it shows the entrance to Westin hotel, just a few hundred metres away from the correct location.\nHowever, the way AI has arrived at these coordinates is far more suspect. The Flickr image the AI linked to doesn’t exist. Inputting the link in a browser produced a 404 error. This link also doesn’t appear in the Wayback Machine, a digital archive of old internet content.\nIt is possible that this is yet another example of AI “hallucination,” where Bing generated a link that doesn’t exist or relied on an old dataset of links.\nIn this instance, it is impossible to determine which images Bing AI used for reference when performing this geolocation, and if it in fact used them.\nBing also says that would use either Google Maps or Google Earth to perform the geolocation — another widespread practice amongst OSINT researchers. It would look for any landmarks that would aid geolocation.\nI can see that my image shows a part of downtown Ottawa near Parliament Hill and Rideau Canal. I can zoom in on this area and see if there are any labels or markers that show the GPS coordinates of any point on the map. I can also use the right-click menu to select “What’s here?” or “Measure distance” to get the GPS coordinates of any point on the map. I can try to find a point on the map that corresponds to where the image was taken from, such as the center of the road or the edge of the sidewalk. For example, I can right-click on a point near the center of Wellington Street and select “What’s here?”. I can see a pop-up window that shows its GPS coordinates, which are 45.4248° N, 75.6926° W.\nBut the AI doesn’t explain, for instance, how it determined that the location is in the vicinity of Parliament Hill and Rideau Canal. The AI deduced that correctly, but didn’t share its reasoning. As our second test demonstrated, it’s unlikely that Bing was able to use commercial signage for this purpose.\nFinally, the AI looked for identifying attributes in the image. It responded:\nI can see that there are some clues or hints in this image, such as:\n• The sign on the left side of the image says “Parliament Hill”, which is a landmark in Ottawa.\n• The flag on top of Parliament Hill is a Canadian flag with a red maple leaf in the center.\n• The language on some of the signs is English with French translations.\n• The license plates on some of the cars are white with blue letters and numbers and have “Ontario” and “Yours to discover” written on them.\n• The weather is sunny and clear, with some clouds in the sky…\nUsing clues or hints from the image itself, I can only deduce the general location of the image, but not the exact GPS coordinates.\nWhile the AI failed to produce a geolocation based on this method, it’s worth examining the “hints” it identified.\nFor a start, there’s no sign on the left side of the image pointing towards Parliament. In fact, parliament itself is not pictured, and neither are any other buildings or landmarks atop the Parliament Hill, let alone those flying a Canadian flag.\nThere are only two visible street signs in the image — a street map pictured in the image, which is too far to see clearly, and the public transport sign above it (red circle on a pillar). There is a vehicle in the image with an Ontario licence plate, which can be read clearly, but the AI already knows that the image has been taken in the city located in that province.\nHence, it appears that the AI is “hallucinating” here as well, and its findings are guesses based upon the broad location of the image: Ottawa. Parliament Hill is the most recognisable landmark in Ottawa. Ottawa is located in Ontario. Based on these factors, the AI “saw” these characteristics in the image, even though they weren’t there or were extremely difficult to recognise. Thus Bing generates responses based on its search for whatever it considers relevant content on the Internet.\nThe results of Bard’s efforts with this image did not lead to its successful geolocation. Once again, this AI seemed more cautious and aware of its limitations.\nEven before we uploaded the image, Bard’s AI gave us a list of famous landmarks in Ottawa: Parliament Hill, the Rideau Canal, the National Gallery of Canada, the Canadian War Museum and the ByWard Market.\nBard also identified the image as Parliament Hill.\nWe prompted Bard again to correct it, pointing out that there was no distinctive clocktower in the photo, which showed cafes and shops.\nBard’s AI readily accepted the correction and provided a list of other possible locations. These included the ByWard Market, Sparks Street, Bank Street and Elgin Street — all commercial areas with plenty of restaurants and cafes.\nSo after much prompting, Bard’s AI reached the correct location – the ByWard Market. It was nevertheless unable to further refine the geolocation, which is less helpful given the large size of the market and surrounding streets. Thus, even these results are little assistance in geolocating an image quickly, easily and precisely. We were also only able to determine that the answer Bard provided was somewhat correct only because we already knew the answer and we knew the area well. In a situation where the location of the image is unknown, whatever Bard produces is likely to be even less helpful.\nA Mimic with Limits\nThese examples demonstrate that Bing and Bard struggle with analysing images and are prone to seeing details which are not there. They also suggest that the AI chatbots we tested imitate the methods of human open source researchers. This could be partially responsible for their poor performance.\nGeoffrey Hinton, a British-Canadian computer scientist and AI specialist, believes that such “confabulations” (his preferred term for AI ‘hallucinations’) are a feature of AI learning models.\n“Confabulation is a signature of human memory. These models are doing something just like people,” he said in an\ninterview\nwith MIT Technology review this May.\nUsing an AI chatbot to fully geolocate an image is inadvisable. At this stage of AI’s development, it might be used to assist with very simple geolocations, perhaps pointing a researcher to an area that may warrant a closer look. However, even such results need to be double-checked and verified and cannot be fully trusted.\nAll photographs were taken by Dennis Kovtun",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/image2-1200x541.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/image3-1200x540.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/07/image4-1200x541.jpg",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2023-09-06_geolocating-russias-disgraced-general-surovikin.json b/bellingcat_geolocation_json/2023-09-06_geolocating-russias-disgraced-general-surovikin.json
new file mode 100644
index 0000000000000000000000000000000000000000..abe1980c7bdeb2d441ad8c09b508c12134e0abfc
--- /dev/null
+++ b/bellingcat_geolocation_json/2023-09-06_geolocating-russias-disgraced-general-surovikin.json
@@ -0,0 +1,47 @@
+{
+ "slug": "geolocating-russias-disgraced-general-surovikin",
+ "title": "Geolocating Russia’s Disgraced General Surovikin",
+ "url": "https://www.bellingcat.com/news/2023/09/06/geolocating-russias-disgraced-general-surovikin/",
+ "date": "2023-09-06T18:04:10+00:00",
+ "modified": "2026-02-11T14:42:07+00:00",
+ "author": "Michael Sheldon",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "After his mysterious disappearance, a photo has surfaced appearing to show the general ‘alive and well’. Here’s how we geolocated it outside an elite Moscow restaurant.",
+ "content_html": "
\n
On September 4, a user posted a photo on the Telegram channel of Russian media personality Ksenia Sobchak. “General Surovikin has emerged. He’s alive and well, home with his family in Moscow” read the accompanying text in Russian.
The BBC’s Verify team reports a “high probability” that the man in the image is indeed General Sergey Surovikin and that the woman standing next to him is “almost certainly” his wife, Anna. The BBC was also unable to find any earlier appearances of this image.
\n
\n
Post from Ksenia Sobchak’s Telegram account showing an apparent photo of General Surovikin and his wife Anna. “The photo is from today”, concludes the text.
\n
The photo garnered public attention in both Russia and Ukraine amid widespread speculation about General Sergey Surovikin’s whereabouts. Last October he was placed in charge of Russian forces invading Ukraine, but was replaced by General Valery Gerasimov this January. Surovikin was then appointed commander of Russian Aerospace Forces but was removed from that role months later.
\n
Meanwhile, doubts over his loyalty to the Kremlin have circulated. US officials told Reuters in June that Surovikin was “sympathetic” to the Wagner mutiny. CNN, citing documents from The Dossier Center, that suggested he was a “secret VIP member” of the Wagner group. Surovikin was last seen in public when he recorded a video urging Wagner troops to end their June mutiny.
Given the General’s fall from the Kremlin’s favour, his reappearance in this image could be politically significant. Open source researchers hurried to geolocate it, initially with difficulty, to the Terrazza Restaurant in an upscale Moscow suburb (55.738352, 37.242541).
\n
To confirm the location of the image, some contextual information was useful. RFE/RL’s Mark Krutov pointed out that the couple were unlikely to be far from their house in “Park Vill,” an elite neighbourhood on the outskirts of Moscow, whose location was made public on an online map hosted by the independent Russian outlet Proekt Media.
\n
The background of the image posted on Sobchak’s channel showed tall pine trees with long, bare stems, suggesting they grew in a forested area. Looking through satellite and on-the-ground imagery of the Surovikins’ neighbourhood, it became clear that the image had to have been taken in the forest to the south of “Park Vill.”
\n
The white stripe on the cobblestones suggested the couple were walking through a parking lot. One obvious place for a couple’s outing would be a restaurant. A search for restaurants in the area of interest yielded a handful of results on Bellingcat’s open street map search tool, which allows you to identify starting points for a geolocation search based on objects and structures you can identify.
\n
\n
Screenshot of results from Bellingcat’s Open Street Map tool, in which blue dots represent restaurants
\n
A few of these locations looked good, but most lacked the plant-covered wall visible in the background of the photo of the Surovikins. One place stood out as particularly promising, with similar background and appropriate tree density. Results of a Google Image search of the restaurant complex Подмоскоvные Vечера (‘Podmoskovnie Vechera’) looked good enough to warrant looking for more angles of the venue in search of a match.
\n
This search yielded photos of the venue published by the Russian media outlet Sobesednik, in a November 2022 article about Surovikin and his wife attending a party at this restaurant in the company of Kremlin spokesperson Dmitry Peskov and propagandist Margarita Simonyan. Facial recognition searches appeared to confirm this, however facial recognition alone may not be sufficient to ID someone due to its limitations. When running a facial recognition search on Surovikin’s wife using as a source image the photo from Sobchak’s Telegram channel, the PimEyes tool suggested several matches — one of which was Sobesednik’s photo of her at the party with her husband in the Terrazza restaurant.
\n
Emboldened by Surovikin’s connection to Podmoskovskie Vechera, we extended the search to include neighbouring properties which may have been in the background of the image. Immediately we found similarities in the neighbouring “Terrazza” restaurant, which featured highly similar awnings and rooftop railings as those seen above the plant-covered wall in the image on Sobchak’s channel.
While images of Terrazza were scarce on the restaurant’s Google Maps and Yandex pages, Benjamin den Braber of the Centre for Information Resilience was able to nail down a location where the image was likely taken, pointing out in a tweet the similarities between the angle of the parking lot corners and the black planters for the hedge at Terrazza.
Looking at other businesses in the area, it became apparent that Terrazza was attached to the beauty salon and spa Aldo Coppola, which had significantly more images of the location to reference. One image in Yandex Maps stood out as it showed the parking area, which had white stripes and cobblestones matching those in the source image. When this image was posted on X, the social media site formerly known as Twitter, in connection with the geolocation, user Thomas Bordeaux pointed out how it showed the exact same location as the image posted on Sobchak’s Telegram channel.
The annotated image matched multiple features in the Terrazza parking lot, including an off-angle pole and vertically misaligned curbs. Further inspection of the images also revealed that the same tree was clearly identifiable in the background of both images.
\n
\n
Detail from a tree seen in the background of Sobchak’s image (left) and a Yandex Images photo of the Terrazza restaurant (right)
\n
This geolocation was another case study in the strengths of crowdsourcing in open search research and geolocation. The contributions by those mentioned in this article, along with other X (formerly known as Twitter) users such as erich_auerbach, cut down the time the geolocation would have taken if it was just one person attempting to nail down the precise location.
While this image appears to show that Surovikin is indeed alive, key questions as to his fate and current standing remain unanswered. When attempting to follow up on questions about a possible investigation into the General by Russian law enforcement, Moscow-based Radio Mayak asked Peskov whether it could ask a question about Surovikin. “No, you can’t,” responded the Kremlin spokesperson.
\n\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter hereand Mastodon here.
\n
",
+ "content_text": "On September 4, a user posted a\nphoto\non the Telegram channel of Russian media personality Ksenia Sobchak. “General Surovikin has emerged. He’s alive and well, home with his family in Moscow” read the accompanying text in Russian.\nThe BBC’s Verify team\nreports\na “high probability” that the man in the image is indeed General Sergey Surovikin and that the woman standing next to him is “almost certainly” his wife, Anna. The BBC was also unable to find any earlier appearances of this image.\nPost from Ksenia Sobchak’s Telegram account showing an apparent photo of General Surovikin and his wife Anna. “The photo is from today”, concludes the text.\nThe photo garnered public attention in both Russia and Ukraine amid widespread speculation about General Sergey Surovikin’s whereabouts. Last October he was placed in charge of Russian forces invading Ukraine, but was replaced by General Valery Gerasimov this January. Surovikin was then\nappointed\ncommander of Russian Aerospace Forces but was\nremoved\nfrom that role months later.\nMeanwhile, doubts over his loyalty to the Kremlin have circulated. US officials\ntold\nReuters in June that Surovikin was “sympathetic” to the Wagner mutiny. CNN,\nciting\ndocuments from\nThe Dossier Center\n, that suggested he was a “secret VIP member” of the Wagner group. Surovikin was last seen in public when he recorded a video urging Wagner troops to end their June\nmutiny\n.\nGiven the General’s fall from the Kremlin’s favour, his reappearance in this image could be politically significant. Open source researchers hurried to geolocate it, initially with difficulty, to the Terrazza Restaurant in an upscale Moscow suburb (\n55.738352, 37.242541\n).\nTo confirm the location of the image, some contextual information was useful. RFE/RL’s\nMark Krutov\npointed out that the couple were unlikely to be far from their house in “Park Vill,” an elite neighbourhood on the outskirts of Moscow, whose location was made public on an\nonline map\nhosted by the independent Russian outlet Proekt Media.\nThe background of the image posted on Sobchak’s channel showed tall pine trees with long, bare stems, suggesting they grew in a forested area. Looking through satellite and on-the-ground imagery of the Surovikins’ neighbourhood, it became clear that the image had to have been taken in the forest to the south of “Park Vill.”\nThe white stripe on the cobblestones suggested the couple were walking through a parking lot. One obvious place for a couple’s outing would be a restaurant. A search for restaurants in the area of interest yielded a handful of results on\nBellingcat’s open street map search tool\n, which allows you to identify starting points for a geolocation search based on objects and structures you can identify.\nScreenshot of results from Bellingcat’s Open Street Map tool, in which blue dots represent restaurants\nA few of these locations looked good, but most lacked the plant-covered wall visible in the background of the photo of the Surovikins. One place stood out as particularly promising, with similar background and appropriate tree density. Results of a Google Image search of the restaurant complex Подмоскоvные Vечера (‘Podmoskovnie Vechera’) looked good enough to warrant looking for more angles of the venue in search of a match.\nThis search yielded photos of the venue published by the Russian media outlet Sobesednik, in a November 2022\narticle\nabout Surovikin and his wife attending a party at this restaurant in the company of Kremlin spokesperson Dmitry Peskov and propagandist Margarita Simonyan. Facial recognition searches appeared to confirm this, however facial recognition alone may not be sufficient to ID someone due to its limitations. When running a facial recognition search on Surovikin’s wife using as a source image the photo from Sobchak’s Telegram channel, the PimEyes tool suggested several matches — one of which was Sobesednik’s\nphoto\nof her at the party with her husband in the Terrazza restaurant.\nEmboldened by Surovikin’s connection to Podmoskovskie Vechera, we extended the search to include neighbouring properties which may have been in the background of the image. Immediately we found similarities in the neighbouring “Terrazza” restaurant, which featured highly similar awnings and rooftop railings as those seen above the plant-covered wall in the image on Sobchak’s channel.\nTop image from the Telegram photo of Surovikin and his wife Anna. Below left photo from\nYandex Images results for Terrazza restaurant\n, below right image from\nPosta Magazine\n.\nWhile images of Terrazza were scarce on the restaurant’s Google Maps and Yandex pages, Benjamin den Braber of the\nCentre for Information Resilience\nwas able to nail down a location where the image was likely taken,\npointing out\nin a tweet the similarities between the angle of the parking lot corners and the black planters for the hedge at Terrazza.\nSource: @Hengenahm on X\nLooking at other businesses in the area, it became apparent that Terrazza was attached to the beauty salon and spa Aldo Coppola, which had significantly more images of the location to reference. One\nimage\nin Yandex Maps stood out as it showed the parking area, which had white stripes and cobblestones matching those in the source image. When this image was posted on X, the social media site formerly known as Twitter, in connection with the geolocation, user Thomas Bordeaux pointed out how it showed the exact same location as the image posted on Sobchak’s Telegram channel.\nSource: ThomasBordeaux4 on X\nThe annotated image matched multiple features in the Terrazza parking lot, including an off-angle pole and vertically misaligned curbs. Further inspection of the images also revealed that the same tree was clearly identifiable in the background of both images.\nDetail from a tree seen in the background of Sobchak’s image (left) and a Yandex Images photo of the Terrazza restaurant (right)\nThis geolocation was another case study in the strengths of crowdsourcing in open search research and geolocation. The contributions by those mentioned in this article, along with other X (formerly known as Twitter) users such as\nerich_auerbach\n, cut down the time the geolocation would have taken if it was just one person attempting to nail down the precise location.\nWhile this image appears to show that Surovikin is indeed alive, key questions as to his fate and current standing remain unanswered. When attempting to follow up on questions about a possible investigation into the General by Russian law enforcement, Moscow-based Radio Mayak\nasked\nPeskov whether it could ask a question about Surovikin. “No, you can’t,” responded the Kremlin spokesperson.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/09/Screenshot-2023-09-06-at-19.06.46-965x1200.png",
+ "alt": "",
+ "caption": "Post from Ksenia Sobchak’s Telegram account showing an apparent photo of General Surovikin and his wife Anna. “The photo is from today”, concludes the text."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/09/image3-1200x623.png",
+ "alt": "",
+ "caption": "Screenshot of results from Bellingcat’s Open Street Map tool, in which blue dots represent restaurants"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/09/image5-1200x568.png",
+ "alt": "",
+ "caption": "Top image from the Telegram photo of Surovikin and his wife Anna. Below left photo from Yandex Images results for Terrazza restaurant , below right image from Posta Magazine ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/09/image4.png",
+ "alt": "",
+ "caption": "Source: @Hengenahm on X"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/09/image1-1200x547.png",
+ "alt": "",
+ "caption": "Source: ThomasBordeaux4 on X"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/09/image2.png",
+ "alt": "",
+ "caption": "Detail from a tree seen in the background of Sobchak’s image (left) and a Yandex Images photo of the Terrazza restaurant (right)"
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2023-10-20_geolocating-hamas-led-attacks-on-israeli-civilians.json b/bellingcat_geolocation_json/2023-10-20_geolocating-hamas-led-attacks-on-israeli-civilians.json
new file mode 100644
index 0000000000000000000000000000000000000000..444ade5696c147945068bd241101889e9c61bfd8
--- /dev/null
+++ b/bellingcat_geolocation_json/2023-10-20_geolocating-hamas-led-attacks-on-israeli-civilians.json
@@ -0,0 +1,127 @@
+{
+ "slug": "geolocating-hamas-led-attacks-on-israeli-civilians",
+ "title": "Geolocating Hamas-Led Attacks on Israeli Civilians",
+ "url": "https://www.bellingcat.com/news/2023/10/20/geolocating-hamas-led-attacks-on-israeli-civilians/",
+ "date": "2023-10-20T13:20:25+00:00",
+ "modified": "2026-02-11T13:21:58+00:00",
+ "author": "Bellingcat Investigation Team",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Bellingcat analysed footage from the Hamas-led attacks at two kibbutzim on October 7.",
+ "content_html": "
\n
Explosions and gunshots are heard in the background as militants record their attack inside a kibbutz.
\n
In one video a militant appears to see someone inside a home and shoots towards them. In another, they shoot the tyres of a parked ambulance. In yet another, they break into a house, where a mobile phone has been left on the kitchen table and is still ringing.
\n
The videos have been circulating on social media and show part of recent surprise attacks in Israel carried out by Hamas and Palestinian Islamic Jihad.
\n
Since the Hamas-led attacks on October 7 that killed hundreds of civilians, Telegram has been a key platform for the distribution of footage of the events. Many videos that emerged can provide insight into the massacres.
\n
Bellingcat analysed footage from the sites of two attacks at kibbutzim — one at Kerem Shalom and one at Sufa — filmed by militants to verify the footage and details of the attack. We have done this in previous cases, including the recent Israeli raids in Jenin. Some of the footage linked in this article is graphic and readers are warned that they may find some of the details in this article disturbing.
As of Friday, Israeli authorities confirmed at least 1,400 deaths since the October 7 attacks, while the Palestinian Ministry of Health confirmed more than 3,400 deaths in Gaza. More than 200 Israelis were taken hostage in the attacks, Israel’s military has said. According to the Office of the United Nations High Commissioner for Human Rights, more than one million Palestinians have been displaced amid Israeli airstrikes carried out in advance of a planned ground invasion.
\n
Common Features of the Videos
\n
Both series of videos were posted on a pro-Israel Telegram channel called South First Responders (SFR). It should be noted that they are posted in short snippets on the channel and at least one of the videos contains a jump cut, which shows it was edited. Despite this, we were able to verify the events recorded in the footage. Both appear to have been filmed from wearable cameras, such as a GoPro, worn by militants involved in the October 7 attacks. Six videos were posted by SFR showing events in Kerem Shalom and 10 were posted by SFR showing events in Sufa.
\n
\n
A map showing the Kibbutzim of Kerem Shalom and Sufa that were targeted in the Hamas-led attack. This map is for illustrative purposes only. Map: Mapcreator.
\n
Other footage of these incidents may exist, but these two video series each appear to have been filmed by a single individual. Both video series were filmed on the morning of October 7. Solar calculation from the footage indicates both attacks occurred around the same time between approximately 7:00 and 8:30 am. Both kibbutzim, which are roughly five kilometres apart, are located less than three kilometres from the border with Gaza.
\n
The Attack on Kerem Shalom
\n
The SFR Telegram channel posted a series of six videos taken from the camera of a fighter who participated in an attack on Kibbutz Kerem Shalom on October 7. The kibbutz is located immediately adjacent to the border between Gaza and Israel, and approximately two kilometres from the border with Egypt.
\n
\n
Google Earth Pro image of Kerem Shalom. Note that it is immediately adjacent to the barrier wall that separates Gaza from Israel. Credit: Google Earth Pro, CNES/Airbus, Maxar Technologies.
\n
The six videos show fighters checking their weapons and driving from Gaza to Kerem Shalom. Some of the videos show fighters blowing holes in the fence separating Gaza from Israel in the vicinity of the kibbutz. Others show the fighters in the kibbutz, firing into homes and exchanging fire with people.
In the first video we see the militants set off on a motorbike as part of a convoy. At one point in the video, they pass two buildings that we geolocated here 31.2512347, 34.2679498. One of the buildings has solar panels on its roof that is visible in satellite imagery.
\n
\n
\n\n
\n
\nLeft: A still from footage posted on Telegram, with two buildings highlighted. Right: The building with the solar panel in the video is visible in satellite imagery (green box), along with the building beside it (red box) at these coordinates: 31.2512347, 34.2679498. Credit: Left: Still from video posted on Telegram by South First Responders. Right: Google Earth Pro/Maxar Technologies.\n\n
Two othervideos show the fighters breaking through what appears to be the barrier wall separating Gaza and Israel, as well as the kibbutz’s perimeter wall. One of these videos indicates that the fighters broke through the kibbutz’s northwestern point, at approximately these coordinates: 31.22788, 34.2815.
\n
\n
\n
\n
A still from a video shared by SFR on Telegram showing the fighters after they have broken through the Kerem Shalom perimeter wall. Note the blue paint on the road in the still from the video. That same blue paint is visible in satellite imagery, on the kibbutz’s northwest corner, that can be seen via the Israeli Gov map here: Govmap.il.
\n
One of the fighters waving others through the perimeter wall has what appears to be an Al-Qassam Brigades patch on his combat vest (Al-Qassam is the military wing of Hamas).
\n
\n
A still from the video shared on Telegram by SFR showing a militant moving through the perimeter wall reveals he is wearing what looks like an Al-Qassam Brigades patch on his combat vest.
\n
According to a solar calculation, the militants crossed into Kerem Shalom about 7:30am on October 7.
\n\n
\n
Using road markings seen on the video and cross referencing these with satellite images, we determined the approximate direction of the shadow cast by one of the militants. According to this estimate, they crossed into Kerem Shalom at around 7:30am on October 7. Credit: Suncalc.org.
\n
The sixth and final video in the series shows the fighters engaged in a firefight inside Kerem Shalom Kibbutz. We can not see in the footage who they exchange fire with but the fighter recording the video is eventually hit in the firefight, and presumably killed.
\n
Bellingcat contributor Jake Godin geolocated the sixth video to the kibbutz’s northwestern edge, approximately 100 metres from the spot where the fighters breached the Kerem Shalom wall.
\n
\n
GoPro footage from SFR shows the assault on Karem Shalom near the Israel-Gaza-Egypt border on 10-7. Filmed in the general area of 31.227964, 34.282299, in the southwest part of the kibbutz.
Benjamin den Braber, a Senior Investigator at the Centre for Information Resilience, geolocated a series of videos from Sufa on 15 October and shared them on X, formerly Twitter. We have independently verified and confirmed several of the geolocations and outlined them below, starting with the second video, when the militant puts on the camera and sets out towards Israel.
\n
\n
GoPro footage shows the 7 October Hamas attack on Sufa in the South of the country near Egypt. The footage starts in Gaza and ends with the Hamas gunman documenting the events shot in Sufa. Below is an overview of events:
The second video in the series shows the militant donning a camera, and riding on the back of a motorcycle along a straight road, which leads towards the border. We geolocated the end of the clip to these coordinates 31.268085, 34.321625, just before the border.
\n\n
\n
\nLeft: A screenshot of the footage posted by SFR on Telegram, showing the view from the motorbike. Right: We have placed boxes around two buildings and a power line which can be identified from the screenshot. Satellite imagery from Google Earth Pro/Airbus.\n\n
In the footage we can see the moment the motorbike struggles across a gap cut into a border fence between Gaza and Israel.
\n
Using the footage from the militant, taken as he was crossing through this fence we calculated that he crossed the border approximately 7am on October 7, based on the sun azimuth.
\n
\n
Overlay of video and sunlight simulation. We calculated the militant crossed the border fence at approximately 7am on October 7, based on the position of the sun. Credit: Google Earth Pro/Maxar Technologies, Landsat/Copernicus.
\n\n
Next the militants join another group of fighters on motorbikes and pass through a concrete barrier wall into Israel at 31.265254, 34.335050.
\n\n
\n
\nLeft:Screenshot from a video posted by SFR on Telegram shows an opening in the barrier wall, seen from the Gaza side, highlighted in black. Right: A screenshot taken from the same video shows the wall from the Israeli side. We can see concrete pipes highlighted in blue on the ground. We matched these two features in Google Earth Pro, outlined below.\n\n
\n
Satellite image showing the concrete pipes, in the blue square and the opening in the barrier wall, in the black square. Credit: Google Earth Pro/CNES/Airbus.
\n
In the third video, the militants drive west down a road and we see them shooting into a fenced area, which we matched to the perimeter of Kibbutz Sufa.
\n\n
\n
\nLeft: The red arrow indicates the direction the fighters were driving in, which we identified by matching the features seen in the footage. At the beginning of the video, the fighters can be seen firing into the kibbutz to their left (south). Credit: Google Earth/CNES/Airbus. Right: A screenshot of a man shooting at the kibbutz. A lattice tower that is recognisable in satellite images is marked with a black box.\n
Before the militants arrive at the gate of the kibbutz they stop at a junction nearby what appears to be an Israeli military base. This can be seen in aerial imagery from Govmap.il located at 31.240231, 34.336213 a few hundred metres away from the kibbutz.
\n
We counted at least 16 other fighters at the location.
The fighters shoot at a target further down the road, possibly IDF forces. One man fires a rocket-propelled grenade (RPG).
\n
In the nextvideos, the fourth and fifth in the series, we see three men jump the gate of the kibbutz. As they approach the gate we see the sun is low on the horizon, beaming from beyond the gate straight into the camera. This suggests it was still early in the morning when they arrived at the entrance of the kibbutz.
\n
\n
A screenshot from the video posted by SFR on Telegram of the Sufa Kibbutz gate, showing the low morning sun.
\n
We see a militant jump over the yellow gate and hide behind the guard house, before walking into the kibbutz. We can see a sign with the name of the kibbutz in Hebrew (סופה).
\n
\n
A screenshot from the video posted on Telegram by SFR, showing a sign with the name of the kibbutz in Hebrew (סופה).
\n
Three other men enter the perimeter of the kibbutz and open the gate. As they enter, more gunshots can be heard. It is unclear who is shooting and where the shots are coming from.
\n
At one point, a car approaches. The men open fire at it.
\n
The next series of videos show the militants walking around the kibbutz. In the sixth video we see the person filming turn towards a house, which we geolocated but are not including the location here. He appears to spot a person inside the home and shoots into the house from the garden. In the audio it sounds like a wounded person inside the house cries out, but the militants appear to fire at least three more shots into the house.
In video eight the militants spot an ambulance parked at (31.238632, 34.342698), shoot at its tyres and continue walking around the kibbutz. This video stops when the fighter is located at this location inside the kibbutz, approximately 450m from the gate (31.238658, 34.343007).
\n\n
\n
\nScreenshot from footage posted on Telegram by SFR shows the militant shoot out the tyres of an ambulance parked at 31.238632, 34.342698.\n
In video nine the militant shoots into another home and finally enters another residence, we geolocated both residences but are not including their location here. Once inside, they rummage through the belongings of the inhabitants and go from room to room. A mobile phone has been left on the kitchen table and is still ringing. A clock on the wall indicates it is 7:50 am local time. They fire one shot into the attic area.
\n
\n
A screenshot from a video posted on Telegram by SFR, showing the time on a wall clock in a home.
\n
In the final video the militants walk outside again to an area we geolocated to 31.238266, 34.341226. A moment later the militant wearing the camera is shot and collapses at 31.238288, 34.340901. He appears to have been shot from the rear as he appears to have been caught off guard. The unknown shooter fires another four times before the sound in the area goes quiet.
\n
\n
Screenshot of the final video posted in series from Sufa on Telegram by SFR.
\n
According to our solar estimates, the militant was shot at some time between 8:00 and 8:30am. His shadow is visible on the ground as he falls. When the camera turns it also shows the approximate direction of the sun behind a house. We used these two reference points to estimate the time when he was shot.
\n
\n
Screenshots from the video posted by SFR on Telegram with shadow analysis.
Bellingcat will continue to monitor the latest war in Israel-Palestine, including the IDF’s bombing of the Gaza Strip, with the aim of documenting civilian harm.
\n\n
Youri van der Weide, Giancarlo Fiorella, Jake Godin, and Carlos Gonzales contributed research.
\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter hereand Mastodon here.
\n
",
+ "content_text": "Explosions and gunshots are heard in the background as militants record their attack inside a kibbutz.\nIn one video a militant appears to see someone inside a home and shoots towards them. In another, they shoot the tyres of a parked ambulance. In yet another, they break into a house, where a mobile phone has been left on the kitchen table and is still ringing.\nThe videos have been circulating on social media and show part of recent surprise attacks in Israel carried out by Hamas and Palestinian Islamic Jihad.\nSince the Hamas-led attacks on October 7 that killed hundreds of civilians, Telegram has been a key platform for the distribution of footage of the events. Many videos that emerged can provide insight into the massacres.\nBellingcat analysed footage from the sites of two attacks at kibbutzim — one at Kerem Shalom and one at Sufa — filmed by militants to verify the footage and details of the attack. We have done this in previous cases, including the recent\nIsraeli raids in Jenin\n. Some of the footage linked in this article is graphic and readers are warned that they may find some of the details in this article disturbing.\nAs of Friday, Israeli authorities\nconfirmed\nat least 1,400 deaths since the October 7 attacks, while the Palestinian Ministry of Health confirmed more than 3,400 deaths in Gaza. More than 200 Israelis were\ntaken hostage\nin the attacks, Israel’s military has said. According to the Office of the United Nations High Commissioner for Human Rights,\nmore than one million\nPalestinians have been displaced amid Israeli airstrikes carried out in advance of a planned ground invasion.\nCommon Features of the Videos\nBoth series of videos were posted on a\npro-Israel Telegram channel\ncalled South First Responders (SFR). It should be noted that they are posted in short snippets on the channel and at least one of the videos contains a jump cut, which shows it was edited. Despite this, we were able to verify the events recorded in the footage. Both appear to have been filmed from wearable cameras, such as a GoPro, worn by militants involved in the October 7 attacks. Six videos were posted by SFR showing events in Kerem Shalom and 10 were posted by SFR showing events in Sufa.\nA map showing the Kibbutzim of Kerem Shalom and Sufa that were targeted in the Hamas-led attack. This map is for illustrative purposes only. Map: Mapcreator\n.\nOther footage of these incidents may exist, but these two video series each appear to have been filmed by a single individual. Both video series were filmed on the morning of October 7. Solar calculation from the footage indicates both attacks occurred around the same time between approximately 7:00 and 8:30 am. Both kibbutzim, which are roughly five kilometres apart, are located less than three kilometres from the border with Gaza.\nThe Attack on Kerem Shalom\nThe SFR Telegram channel posted a series of six videos taken from the camera of a fighter who participated in an attack on Kibbutz Kerem Shalom on October 7. The kibbutz is\nlocated\nimmediately adjacent to the border between Gaza and Israel, and approximately two kilometres from the border with Egypt.\nG\noogle Earth Pro image of Kerem Shalom. Note that it is immediately adjacent to the barrier wall that separates Gaza from Israel. Credit: Google Earth Pro, CNES/Airbus, Maxar Technologies\n.\nThe six videos show fighters checking their weapons and driving from Gaza to Kerem Shalom. Some of the videos show fighters blowing holes in the fence separating Gaza from Israel in the vicinity of the kibbutz. Others show the fighters in the kibbutz, firing into homes and exchanging fire with people.\nIn the\nfirst video\nwe see the militants set off on a motorbike as part of a convoy. At one point in the video, they pass two buildings that we geolocated here\n31.2512347, 34.2679498\n. One of the buildings has solar panels on its roof that is visible in satellite imagery.\nLeft: A still from footage posted on Telegram, with two buildings highlighted. Right: The building with the solar panel in the video is visible in satellite imagery (green box), along with the building beside it (red box) at these coordinates:\n31.2512347, 34.2679498\n. Credit: Left: Still from video posted on Telegram by South First Responders. Right: Google Earth Pro/Maxar Technologies\n.\nTwo\nother\nvideos\nshow the fighters breaking through what appears to be the barrier wall separating Gaza and Israel, as well as the kibbutz’s perimeter wall. One of these videos indicates that the fighters broke through the kibbutz’s northwestern point, at approximately these coordinates:\n31.22788, 34.2815\n.\nA still from a video shared by SFR on Telegram showing the fighters after they have broken through the Kerem Shalom perimeter wall. Note the blue paint on the road in the still from the video. That same blue paint is visible in satellite imagery, on the kibbutz’s northwest corner, that can be seen via the Israeli Gov map here:\nGovmap.il\n.\nOne of the fighters waving others through the perimeter wall has what appears to be an Al-Qassam Brigades patch on his combat vest (Al-Qassam is the military wing of Hamas).\nA still from the video shared on Telegram by SFR showing a militant moving through the perimeter wall reveals he is wearing what looks like an Al-Qassam Brigades patch on his combat vest.\nAccording to a solar calculation, the militants crossed into Kerem Shalom about 7:30am on October 7.\nUsing road markings seen on the video and cross referencing these with satellite images, we determined the approximate direction of the shadow cast by one of the militants. According to this estimate, they crossed into Kerem Shalom at around 7:30am on October 7.\nCredit:\nSuncalc.org\n.\nThe\nsixth and final video\nin the series shows the fighters engaged in a firefight inside Kerem Shalom Kibbutz. We can not see in the footage who they exchange fire with but the fighter recording the video is eventually hit in the firefight, and presumably killed.\nBellingcat contributor Jake Godin geolocated the sixth video to the kibbutz’s northwestern edge, approximately 100 metres from the spot where the fighters breached the Kerem Shalom wall.\nGoPro footage from SFR shows the assault on Karem Shalom near the Israel-Gaza-Egypt border on 10-7. Filmed in the general area of 31.227964, 34.282299, in the southwest part of the kibbutz.\nMap link:\nhttps://t.co/5jOum87kZa\nTG link:\nhttps://t.co/P7y9l2t57n\n@GeoConfirmed\npic.twitter.com/UungmGUQVw\n— Jake Godin (@JakeGodin)\nOctober 16, 2023\nThe Attack on Sufa\nBenjamin den Braber, a Senior Investigator at the Centre for Information Resilience, geolocated a series of videos from Sufa on 15 October and shared them on X, formerly Twitter. We have independently verified and confirmed several of the geolocations and outlined them below, starting with the second video, when the militant puts on the camera and sets out towards Israel.\nGoPro footage shows the 7 October Hamas attack on Sufa in the South of the country near Egypt. The footage starts in Gaza and ends with the Hamas gunman documenting the events shot in Sufa. Below is an overview of events:\nSource: South First Responders\nhttps://t.co/boOGaxm9bD\npic.twitter.com/FjTWjXABZd\n— Benjamin (@hengenahm)\nOctober 15, 2023\nThe\nsecond video in the series\nshows the militant donning a camera, and riding on the back of a motorcycle along a straight road, which leads towards the border. We geolocated the end of the clip to these coordinates\n31.268085, 34.321625\n, just before the border.\nLeft: A screenshot of the footage posted by SFR on Telegram, showing the view from the motorbike. Right: We have placed boxes around two buildings and a power line which can be identified from the screenshot. Satellite imagery from Google Earth Pro/Airbus\n.\nIn the footage we can see the moment the motorbike struggles across a gap cut into a border fence between Gaza and Israel.\nUsing the footage from the militant, taken as he was crossing through this fence we calculated that he crossed the border approximately 7am on October 7, based on the sun azimuth.\nOverlay of video and sunlight simulation. We calculated the militant crossed the border fence at approximately 7am on October 7, based on the position of the sun. Credit: Google Earth Pro/Maxar Technologies, Landsat/Copernicus\n.\nNext the militants join another group of fighters on motorbikes and pass through a concrete barrier wall into Israel at\n31.265254, 34.335050\n.\nLeft:\nScreenshot from a video posted by SFR on Telegram shows an opening in the barrier wall, seen from the Gaza side, highlighted in black. Right: A screenshot taken from the same video shows the wall from the Israeli side. We can see concrete pipes\nhighlighted in blue\non the ground. We matched these two features in Google Earth Pro, outlined below.\nSatellite image showing the concrete pipes, in the blue square and the opening in the barrier wall, in the black square. Credit: Google Earth Pro/CNES/Airbus\n.\nIn the\nthird video\n, the militants drive west down a road and we see them shooting into a fenced area, which we matched to the perimeter of Kibbutz Sufa.\nLeft: The red arrow indicates the direction the fighters were driving in, which we identified by matching the features seen in the footage. At the beginning of the video, the fighters can be seen firing into the kibbutz to their left (south). Credit: Google Earth/CNES/Airbus\n.\nRight: A screenshot of a man shooting at the kibbutz. A lattice tower that is recognisable in satellite images is marked with a black box.\nBefore the militants arrive at the gate of the kibbutz they stop at a junction nearby what appears to be an Israeli military base. This can be seen in\naerial imagery from Govmap.il\nlocated at\n31.240231, 34.336213\na few hundred metres away from the kibbutz.\nWe counted at least 16 other fighters at the location.\nThe fighters shoot at a target further down the road, possibly IDF forces. One man fires a rocket-propelled grenade (RPG).\nIn the\nnext\nvideos\n, the fourth and fifth in the series, we see three men jump the gate of the kibbutz. As they approach the gate we see the sun is low on the horizon, beaming from beyond the gate straight into the camera. This suggests it was still\nearly in the morning\nwhen they arrived at the entrance of the kibbutz.\nA screenshot from the video posted by SFR on Telegram of the Sufa Kibbutz gate, showing the low morning sun.\nWe see a militant jump over the yellow gate and hide behind the guard house, before walking into the kibbutz. We can see a sign with the name of the kibbutz in Hebrew (סופה).\nA screenshot from the video posted on Telegram by SFR, showing a sign with the name of the kibbutz in Hebrew (סופה).\nThree other men enter the perimeter of the kibbutz and open the gate. As they enter, more gunshots can be heard. It is unclear who is shooting and where the shots are coming from.\nAt one point, a car approaches. The men open fire at it.\nThe next series of videos show the militants walking around the kibbutz. In\nthe sixth video\nwe see the person filming turn towards a house, which we geolocated but are not including the location here. He appears to spot a person inside the home and shoots into the house from the garden. In the audio it sounds like a wounded person inside the house cries out, but the militants appear to fire at least three more shots into the house.\nIn\nvideo eight\nthe militants spot an ambulance parked at (\n31.238632, 34.342698\n), shoot at its tyres and continue walking around the kibbutz. This video stops when the fighter is located at this location inside the kibbutz, approximately 450m from the gate (\n31.238658, 34.343007\n).\nScreenshot from footage posted on Telegram by SFR shows the militant shoot out the tyres of an ambulance parked at\n31.238632, 34.342698\n.\nIn\nvideo nine\nthe militant shoots into another home and finally enters another residence, we geolocated both residences but are not including their location here. Once inside, they rummage through the belongings of the inhabitants and go from room to room. A mobile phone has been left on the kitchen table and is still ringing. A clock on the wall indicates it is 7:50 am local time. They fire one shot into the attic area.\nA screenshot from a video posted on Telegram by SFR, showing the time on a wall clock\nin a home\n.\nIn the\nfinal video\nthe militants walk outside again to an area we geolocated to\n31.238266, 34.341226\n. A moment later the militant wearing the camera is shot and collapses at\n31.238288, 34.340901\n. He appears to have been shot from the rear as he appears to have been caught off guard. The unknown shooter fires another four times before the sound in the area goes quiet.\nScreenshot of the final video posted in series from Sufa on Telegram by SFR.\nAccording to our\nsolar estimates\n, the militant was shot at some time between 8:00 and 8:30am. His shadow is visible on the ground as he falls. When the camera turns it also shows the approximate direction of the sun behind a house. We used these two reference points to\nestimate\nthe time when he was shot.\nScreenshots from the video posted by SFR on Telegram with shadow analysis.\nOn Wednesday, Human Rights Watch said it had verified four videos from the attacks in Israel- from the attacks on the Supernova music festival and Kibbutz Mefalsim. It said it is\ninvestigating these incidents, as well as others, as war crimes\n.\nBellingcat will continue to monitor the latest war in Israel-Palestine, including the IDF’s bombing of the Gaza Strip, with the aim of documenting civilian harm.\nYouri van der Weide, Giancarlo Fiorella, Jake Godin, and Carlos Gonzale\ns\ncontributed research.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-8.jpg",
+ "alt": "",
+ "caption": "A map showing the Kibbutzim of Kerem Shalom and Sufa that were targeted in the Hamas-led attack. This map is for illustrative purposes only. Map: Mapcreator ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-9-1200x749.jpg",
+ "alt": "",
+ "caption": "G oogle Earth Pro image of Kerem Shalom. Note that it is immediately adjacent to the barrier wall that separates Gaza from Israel. Credit: Google Earth Pro, CNES/Airbus, Maxar Technologies ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-10-1200x673.jpg",
+ "alt": "",
+ "caption": "Left: A still from footage posted on Telegram, with two buildings highlighted. Right: The building with the solar panel in the video is visible in satellite imagery (green box), along with the building beside it (red box) at these coordinates: 31.2512347, 34.2679498 . Credit: Left: Still from video posted on Telegram by South First Responders. Right: Google Earth Pro/Maxar Technologies ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-11-1200x748.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-12-1200x673.jpg",
+ "alt": "",
+ "caption": "A still from a video shared by SFR on Telegram showing the fighters after they have broken through the Kerem Shalom perimeter wall. Note the blue paint on the road in the still from the video. That same blue paint is visible in satellite imagery, on the kibbutz’s northwest corner, that can be seen via the Israeli Gov map here: Govmap.il ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/Screenshot-2023-10-19-at-17.00.10-1188x1200.png",
+ "alt": "",
+ "caption": "A still from the video shared on Telegram by SFR showing a militant moving through the perimeter wall reveals he is wearing what looks like an Al-Qassam Brigades patch on his combat vest."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-13-1200x785.jpg",
+ "alt": "",
+ "caption": "Using road markings seen on the video and cross referencing these with satellite images, we determined the approximate direction of the shadow cast by one of the militants. According to this estimate, they crossed into Kerem Shalom at around 7:30am on October 7. Credit: Suncalc.org ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-5-1200x675.png",
+ "alt": "",
+ "caption": "Left: A screenshot of the footage posted by SFR on Telegram, showing the view from the motorbike. Right: We have placed boxes around two buildings and a power line which can be identified from the screenshot. Satellite imagery from Google Earth Pro/Airbus ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-6-1200x675.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed.gif",
+ "alt": "",
+ "caption": "Overlay of video and sunlight simulation. We calculated the militant crossed the border fence at approximately 7am on October 7, based on the position of the sun. Credit: Google Earth Pro/Maxar Technologies, Landsat/Copernicus ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-7-1200x675.png",
+ "alt": "",
+ "caption": "Left: Screenshot from a video posted by SFR on Telegram shows an opening in the barrier wall, seen from the Gaza side, highlighted in black. Right: A screenshot taken from the same video shows the wall from the Israeli side. We can see concrete pipes highlighted in blue on the ground. We matched these two features in Google Earth Pro, outlined below."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-8-1200x675.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-9-1200x675.png",
+ "alt": "",
+ "caption": "Satellite image showing the concrete pipes, in the blue square and the opening in the barrier wall, in the black square. Credit: Google Earth Pro/CNES/Airbus ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-10-1.png",
+ "alt": "",
+ "caption": "Left: The red arrow indicates the direction the fighters were driving in, which we identified by matching the features seen in the footage. At the beginning of the video, the fighters can be seen firing into the kibbutz to their left (south). Credit: Google Earth/CNES/Airbus . Right: A screenshot of a man shooting at the kibbutz. A lattice tower that is recognisable in satellite images is marked with a black box."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-11.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-14-1200x675.jpg",
+ "alt": "",
+ "caption": "A screenshot from the video posted by SFR on Telegram of the Sufa Kibbutz gate, showing the low morning sun."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-12-1200x669.png",
+ "alt": "",
+ "caption": "A screenshot from the video posted on Telegram by SFR, showing a sign with the name of the kibbutz in Hebrew (סופה)."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-13-1200x687.png",
+ "alt": "",
+ "caption": "Screenshot from footage posted on Telegram by SFR shows the militant shoot out the tyres of an ambulance parked at 31.238632, 34.342698 ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-14-1200x687.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-15.png",
+ "alt": "",
+ "caption": "A screenshot from a video posted on Telegram by SFR, showing the time on a wall clock in a home ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-16-1200x675.png",
+ "alt": "",
+ "caption": "Screenshot of the final video posted in series from Sufa on Telegram by SFR."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/10/unnamed-17.png",
+ "alt": "",
+ "caption": "Screenshots from the video posted by SFR on Telegram with shadow analysis."
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2023-11-16_from-pixels-to-punches-geolocating-a-neo-nazi-and-white-nationalist-combat-event-in-los-angeles.json b/bellingcat_geolocation_json/2023-11-16_from-pixels-to-punches-geolocating-a-neo-nazi-and-white-nationalist-combat-event-in-los-angeles.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb4dd0a137bd5ea36a283c1aa1dc55c5a31a2ddf
--- /dev/null
+++ b/bellingcat_geolocation_json/2023-11-16_from-pixels-to-punches-geolocating-a-neo-nazi-and-white-nationalist-combat-event-in-los-angeles.json
@@ -0,0 +1,87 @@
+{
+ "slug": "from-pixels-to-punches-geolocating-a-neo-nazi-and-white-nationalist-combat-event-in-los-angeles",
+ "title": "From Pixels to Punches: Geolocating a neo-Nazi and White Nationalist Combat Event in Los Angeles",
+ "url": "https://www.bellingcat.com/news/2023/11/16/from-pixels-to-punches-geolocating-a-neo-nazi-and-white-nationalist-combat-event-in-los-angeles/",
+ "date": "2023-11-16T18:05:33+00:00",
+ "modified": "2026-02-11T17:17:43+00:00",
+ "author": "Michael Colborne, Kolina Koltai, Tristan Lee",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Chapters of an emerging white nationalist network from across the United States participated in a joint combat sports event in the Los Angeles area this August, according to open source evidence geolocated by Bellingcat.",
+ "content_html": "
\n
Chapters of an emerging white nationalist network from across the United States participated in a joint combat sports event in the Los Angeles area this August, according to open source evidence geolocated by Bellingcat.
\n
Several Active Clubs, which experts say form a dangerous bedrock for far-right activity and recruitment, joined the second annual tournament alongside extremist groups like Patriot Front and the Hammerskins. Active Clubs are a network of white nationalist mixed martial arts crews inspired by the Rise Above Movement, a now-defunct militant streetfighting group whose neo-Nazi cofounder Robert Rundo is currently in jail awaiting trial on federal rioting charges. They focus on training their members in combat skills in order to prepare them to fight against their purported enemies.
\n
“Their own propaganda says ‘we are a white nationalist sports network, it’s about fitness,’” said Alexander Ritzmann, a political scientist and senior advisor at the Counter Extremism Project who studies Active Clubs, in an interview. “But reading their documents and listening to their podcasts, I’m curious about if they are a combat sports network or if they are a militia hiding in plain sight.”
\n
Participating groups sent fighters to the event, which was organised by white supremacist activewear brand Will2Rise, with winners awarded custom medals.
\n
Promotional content for the contest, which was posted in a Telegram channel affiliated with Rundo’s defunct Rise Above movement, said it took place in Huntington Beach, California, on August 19. Using a video and photos released by a far-right propaganda outlet founded by Rundo, Bellingcat was able to geolocate the event to a gym space in the next city over, Fountain Valley. Both are located in Orange County within the Los Angeles metropolitan area.
\n
(Bellingcat is not naming the outlet, which has promised to release footage from this event and other gatherings to promote white nationalism to a wider audience of young men, to prevent amplification.)
\n
A co-owner of the gym that hosted the event, ION Strength and Conditioning, did not reply to requests for comment. ION has since closed and the space is now run under different ownership with no relation to the event.
\n
Fascist Fight Club
\n
The site where the tournament was held was a strength and conditioning gym. We identified the venue by matching its features to promotional material released by participating Active Clubs.
\n
Several contextual clues in the videos and photos of the event allowed us to focus our search on strength and conditioning gyms in the Los Angeles metropolitan area:
\n
\n
The gym featured items commonly seen in similar facilities, including climbing ropes, weight racks, and a peg board, and appeared to be in an industrial or garage space like similar facilities\n
\n\n
\n
\n
\n
\n
In a few frames of the promotional video posted of the event, a poster is visible on the office wall that Bellingcat was able to identify as a logo of part of a women’s CrossFit competition league based in the Los Angeles metropolitan area \n
\n\n
\n
\n\n
\n
\n
\n
\n
Photo still taken from video (left) showing a logo (red box) in the office space. Zoomed-in inset of logo (right).
\n
Based on this assumption, we searched Google Maps for locations of strength and conditioning gyms in the Los Angeles area. For each of the several dozen locations we found, we reviewed publicly available photos of the facility interior, as well as available Google Street View images of the exterior, to see if it could be a possible match.
When we began reviewing photos from a strength and conditioning facility in Fountain Valley, a city just north of Huntington Beach, it didn’t take us long to find a match. In one Telegram photo shared on September 18 by the far-right propaganda outlet, we noted several unique details outside the venue, including doors, foliage, and red lettering on a similar facility across what appeared to be a parking lot. These same features were visible on Google Street View images from outside the Fountain Valley facility.
\n\n
\n
\nPhoto (left) shared on Telegram and photo from Google Street View from the gym (right). \n
Even though the organisers of the event had taken measures to obscure potentially identifiable details inside the facility, they didn’t manage to hide them all — and these clues helped us confirm we were looking in the right place. Many of these features from the gym were only partly obscured in the footage by the signs and banners of far-right extremist groups attending the event.
Among the signs were a sonnenrad and a Celtic cross, both symbols associated with far-right extremism, and a banner for the Hammerskins, a well-documented neo-Nazi group involved in numerous violent attacks.
\n
\n
Photo from Telegram of the event, with a yellow box around a Hammerskins symbol.
\n
The presence of Hammerskins symbols is significant because, as the Global Project Against Hate and Extremism notes, Active Clubs “work hand-in-glove” with other racist and white nationalist groups in North America and Europe. This sets them apart from past white nationalist networks, which were often uncooperative rivals.
\n
“Traditionally, these right-wing extremist groups would not allow members to be members of another group,” said Ritzmann, the counter-extremism researcher. “The unique strategy is white unity at every opportunity, meaning Active Clubs have no problem with their members being in the Proud Boys, Patriot Front and Hammerskins.”
\n
According to reporting by Vice and the Canadian Anti-Hate Network, Hammerskins in particular have nested themselves in multiple Active Club chapters in North America to enhance their recruiting networks.
\n
“These far-right MMA tournaments can also fund other far-right activities through their dues, and fees, as well as through the merchandise that gets sold along with it,” said Meredith L. Pruden, a communications professor at Kennesaw State University who researches far-right groups, in an interview.
One detail that helped Bellingcat confirm the location was a red square visible beneath one of the fascist symbols on the banners. We also spotted promotional materials for a fitness recovery drink — the drink company’s logo was also on a section of the wall alongside four posters of athletes. In that same corner of the gym, there is a sign with obscured contact and social media details. While the gym name was covered, the icons indicating what social media platforms it was on remained visible.
\n
\n
Photos shared on Telegram (Top). The highlighted blue box indicates a portion of the gym’s banner, the green box indicates the fitness drink promotional material, and the purple box indicates redacted social media contact information. Photos from the gym’s rental listing (bottom left) and from Google Maps listing (bottom right) match the features from the shared Telegram photos.
\n
We used these identifying markers to match images from the far-right event to those in a gym rental listing and the gym’s Google Maps entry.
The red square turned out to be the corner of the letter I in the gym’s name, ION. The same posters and social media contact sign can be seen hanging in the corner. Furthermore, the images of the gym show ropes hanging from the ceiling that match ropes that are tied to the ceiling in the event photos.
\n
Another feature we identified that was not covered with the groups’ banners was a workout peg board on a black back wall. We could also see fluorescent lighting to the right of the board. Again, using images of the gym from their rental listing, we matched this same peg board on a black wall with the lighting fixture to the right.
\n\n
\n
\nPhoto still taken from video trailer (left) with a workout peg board (blue box) and lighting fixture to the right. Photo from gym’s rental listing (right) matching the placement of the peg board and light.\n\n
The gym’s Google map listing also features multiple images shared by users documenting the gym’s participation in the Barbellas competitions. Tickets and promotional materials for that event featured the same gym logo that Bellingcat initially identified, further confirming that the identified gym was the space used for the event.
\n
ION Man
\n
When Bellingcat looked into the proprietor of ION Strength and Conditioning, social media profiles revealed that one of the gym’s owners has a troubling history of sharing far-right material and posting about white nationalist ideology online.
\n
Jeremy Loo identifies himself as the owner of ION on LinkedIn and Facebook profiles. A business name registration for the gym in Orange County also lists Loo and another owner. An archive of ION’s website and social media accounts for Loo and the gym describe him as both an owner and head coach.
He did not respond to multiple requests for comment, including questions about whether the ownership of ION was involved in the organising of the event or posts on the social media accounts outlined below.
\n
\n
ION Strength and Conditioning is registered as a Fictitious Business name in Orange County, CA, to Loo and another owner.(Source: Orange County Clerk-Recorder Department)
\n
Loo has multiple social media accounts for his personal use and business, most of which are fitness-related. Of note, his profile on Minds, an open-source social network, contains a post referring to the conspiratorial and racist “Great Replacement” theory.
Bellingcat was able to identify the Minds account because the username includes Loo’s first and last name and one of its posts includes the hashtag #ionsandc, which the ION gym’s hashtag used for promotional purposes on its affiliated social accounts. He also shared a quote from Italian far-right philosopher Julius Evola on the short-form video platform TikTok and reposted it to his profile on Gettr, a conservative microblogging platform. Evola, a virulent antisemite and contemporaneous advocate of Benito Mussolini’s Racial Laws, remains influential across much of the international far right.
\n\n
\n
\nScreenshot of Loo’s profile on the social media platform Minds, showing White Nationalist-related content (left). Screenshot from the ION Strength and Conditioning TikTok page with a Julius Evola quote, which was reposted on Loo’s Gettr profile (right).\n
Bellingcat also identified an X account with a similar username as Loo’s Minds account that posts and reposts explicitly neo-Nazi content. This account follows Loo’s professional coaching account on X, and is followed by Loo’s personal account on X. Given that this account has only 18 followers, with Loo’s personal account being one of them, this leads us to believe that is likely Loo’s alt-X account.
\n
\n
Loo’s likely alt-X account follows his coaching account and is followed by his personal account (Source: X)
\n
This account also shared some of the same content as Loo’s Minds profile. In addition to posting tweets with neo-Nazi phrases like “1488” and the “Fourteen Words”, the account also liked another user’s post that links to the abovementioned Rise Above Movement Telegram channel where promotional materials for the far-right MMA event at ION gym were posted.
\n
White supremacist and neo-Nazi content posted, reposted, or liked by Loo’s likely alt-X account(Source: X)\n
Bellingcat also found a Telegram account with the same username as the X account that appears to be Loo’s alt. The Telegram account features a picture of Loo as its profile image — this is the same photo Loo uses for his personal Facebook and Instagram accounts, but with a glowing effect photoshopped over his eyes. The photo with the effect is also shared on the Google Map listing for ION, used as the profile picture for Loo’s Gettr account, and is an older Facebook profile picture.
\n
Screenshots of Loo’s profiles on various social media platforms, with coloured boxes showing commonalities between websites.\n
This account is a public member of a number of conservative, conspiracy, White pride, and far-right groups including an explicitly neo-Nazi chat group called the “National Socialist Workers Party”. This group describes itself with terms like “Holocaust Deniers,” “Aryan Christians,” and “1488.”
\n
Screenshot of messages and member list for Telegram channel “National Socialist Workers Party”, showing Loo as a member, with insets of Loo’s profile picture and the group’s profile picture.\n
One thing that is clear is that we can expect these types of combat sports events in the future as they function as a recruiting tool, funding mechanism, and community-building atmosphere. As Pruden told Bellingcat, “It’s important to call attention to the fact that these groups are making concerted efforts to co-opt mainstream mixed martial arts. They’re doing that because they see these sorts of combat sports as capable of mainstreaming far-right beliefs”.
\n
White nationalist groups aren’t in some remote part of the world, but oftentimes right in our own neighbourhood.
\n\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Instagram here, X hereand Mastodonhere.
\n
",
+ "content_text": "Chapters of an emerging white nationalist network from across the United States participated in a joint combat sports event in the Los Angeles area this August, according to open source evidence geolocated by Bellingcat.\nSeveral Active Clubs, which experts say form a dangerous bedrock for far-right activity and recruitment, joined the second annual tournament alongside extremist groups like Patriot Front and the Hammerskins. Active Clubs are a\nnetwork\nof white nationalist mixed martial arts crews inspired by the Rise Above Movement, a now-defunct militant streetfighting group whose neo-Nazi cofounder Robert Rundo is currently in jail awaiting trial on\nfederal rioting charges\n. They focus on training their members in combat skills in order to prepare them to fight against their purported enemies.\n“Their own propaganda says ‘we are a white nationalist sports network, it’s about fitness,’” said Alexander Ritzmann, a political scientist and senior advisor at the Counter Extremism Project who studies Active Clubs, in an interview. “But reading their documents and listening to their podcasts, I’m curious about if they are a combat sports network or if they are a militia hiding in plain sight.”\nParticipating groups sent fighters to the event, which was organised by white supremacist activewear brand\nWill2Rise\n, with winners awarded custom medals.\nPromotional content for the contest, which was posted in a Telegram channel affiliated with Rundo’s defunct Rise Above movement, said it took place in Huntington Beach, California, on August 19. Using a video and photos released by a far-right propaganda outlet founded by Rundo, Bellingcat was able to geolocate the event to a gym space in the next city over, Fountain Valley. Both are located in Orange County within the Los Angeles metropolitan area.\n(Bellingcat is not naming the outlet, which has promised to release footage from this event and other gatherings to promote white nationalism to a wider audience of young men, to prevent amplification.)\nA co-owner of the gym that hosted the event, ION Strength and Conditioning, did not reply to requests for comment. ION has since closed and the space is now run under different ownership with no relation to the event.\nFascist Fight Club\nThe site where the tournament was held was a strength and conditioning gym. We identified the venue by matching its features to promotional material released by participating Active Clubs.\nSeveral contextual clues in the videos and photos of the event allowed us to focus our search on strength and conditioning gyms in the Los Angeles metropolitan area:\nThe gym featured items commonly seen in similar facilities, including climbing ropes, weight racks, and a peg board, and appeared to be in an industrial or garage space like similar facilities\nIn a few frames of the promotional video posted of the event, a poster is visible on the office wall that Bellingcat was able to identify as a logo of part of a women’s CrossFit competition\nleague\nbased in the Los Angeles metropolitan area\nPhoto still taken from video (left) showing a logo (red box) in the office space. Zoomed-in inset of logo (right).\nBased on this assumption, we searched Google Maps for locations of strength and conditioning gyms in the Los Angeles area. For each of the several dozen locations we found, we reviewed publicly available photos of the facility interior, as well as available Google Street View images of the exterior, to see if it could be a possible match.\nWhen we began reviewing photos from a strength and conditioning\nfacility\nin Fountain Valley, a city just north of Huntington Beach, it didn’t take us long to find a match. In one Telegram photo shared on September 18 by the far-right propaganda outlet, we noted several unique details outside the venue, including doors, foliage, and red lettering on a similar facility across what appeared to be a parking lot. These same features were visible on Google Street View images from outside the Fountain Valley facility.\nPhoto (left) shared on Telegram and photo from Google Street View from the gym (right).\nEven though the organisers of the event had taken measures to obscure potentially identifiable details inside the facility, they didn’t manage to hide them all — and these clues helped us confirm we were looking in the right place. Many of these features from the gym were only partly obscured in the footage by the signs and banners of far-right extremist groups attending the event.\nAmong the signs were a sonnenrad and a Celtic cross, both\nsymbols\nassociated with far-right extremism, and a banner for the Hammerskins, a well-documented neo-Nazi group involved in numerous violent attacks.\nPhoto from Telegram of the event, with a yellow box around a Hammerskins symbol.\nThe presence of Hammerskins symbols is significant because, as the Global Project Against Hate and Extremism\nnotes,\nActive Clubs “work hand-in-glove” with other racist and white nationalist groups in North America and Europe. This sets them apart from past white nationalist networks, which were often uncooperative rivals.\n“Traditionally, these right-wing extremist groups would not allow members to be members of another group,” said Ritzmann, the counter-extremism researcher. “The unique strategy is white unity at every opportunity, meaning Active Clubs have no problem with their members being in the Proud Boys, Patriot Front and Hammerskins.”\nAccording to reporting by\nVice\nand the\nCanadian Anti-Hate Network\n, Hammerskins in particular have nested themselves in multiple Active Club chapters in North America to enhance their recruiting networks.\n“These far-right MMA tournaments can also fund other far-right activities through their dues, and fees, as well as through the merchandise that gets sold along with it,” said Meredith L. Pruden, a communications professor at Kennesaw State University who researches far-right groups, in an interview.\nOne detail that helped Bellingcat confirm the location was a red square visible beneath one of the fascist symbols on the banners. We also spotted promotional materials for a fitness recovery drink — the drink company’s logo was also on a section of the wall alongside four posters of athletes. In that same corner of the gym, there is a sign with obscured contact and social media details. While the gym name was covered, the icons indicating what social media platforms it was on remained visible.\nPhotos shared on Telegram (Top). The highlighted blue box indicates a portion of the gym’s banner, the green box indicates the fitness drink promotional material, and the purple box indicates redacted social media contact information. Photos from the gym’s rental listing (bottom left) and from Google Maps listing (bottom right) match the features from the shared Telegram photos.\nWe used these identifying markers to match images from the far-right event to those in a gym rental listing and the gym’s Google Maps entry.\nThe red square turned out to be the corner of the letter I in the gym’s name, ION. The same posters and social media contact sign can be seen hanging in the corner. Furthermore, the images of the gym show ropes hanging from the ceiling that match ropes that are tied to the ceiling in the event photos.\nAnother feature we identified that was not covered with the groups’ banners was a workout peg board on a black back wall. We could also see fluorescent lighting to the right of the board. Again, using images of the gym from their rental listing, we matched this same peg board on a black wall with the lighting fixture to the right.\nPhoto still taken from video trailer (left) with a workout peg board (blue box) and lighting fixture to the right.\nPhoto\nfrom gym’s rental listing (right) matching the placement of the peg board and light.\nThe gym’s Google map listing also features multiple images shared by users documenting the gym’s participation in the Barbellas competitions. Tickets and promotional materials for that event featured the same gym logo that Bellingcat initially identified, further confirming that the identified gym was the space used for the event.\nION Man\nWhen Bellingcat looked into the proprietor of ION Strength and Conditioning, social media profiles revealed that one of the gym’s owners has a troubling history of sharing far-right material and posting about white nationalist ideology online.\nJeremy Loo identifies himself as the owner of ION on LinkedIn and Facebook profiles. A business name registration for the gym in Orange County also lists Loo and another owner. An\narchive\nof ION’s website and social media accounts for Loo and the gym describe him as both an owner and head coach.\nHe did not respond to multiple requests for comment, including questions about whether the ownership of ION was involved in the organising of the event or posts on the social media accounts outlined below.\nION Strength and Conditioning is registered as a Fictitious Business name in Orange County, CA, to Loo and another owner.\n(Source: Orange County Clerk-Recorder Department)\nLoo has multiple social media accounts for his personal use and business, most of which are fitness-related. Of note, his profile on Minds, an open-source social network, contains a post referring to the conspiratorial and racist “\nGreat Replacement\n” theory.\nBellingcat was able to identify the Minds account because the username includes Loo’s first and last name and one of its posts includes the hashtag #ionsandc, which the ION gym’s hashtag used for promotional purposes on its affiliated social accounts. He also shared a quote from Italian far-right philosopher Julius Evola on the short-form video platform TikTok and reposted it to his profile on Gettr, a conservative microblogging platform. Evola, a virulent antisemite and contemporaneous advocate of Benito Mussolini’s Racial Laws, remains\ninfluential\nacross much of the international far right.\nScreenshot of Loo’s profile on the social media platform Minds, showing White Nationalist-related content (left). Screenshot from the ION Strength and Conditioning TikTok page with a Julius Evola quote, which was reposted on Loo’s Gettr profile (right).\nBellingcat also identified an X account with a similar username as Loo’s Minds account that posts and reposts explicitly neo-Nazi content. This account follows Loo’s professional coaching account on X, and is followed by Loo’s personal account on X. Given that this account has only 18 followers, with Loo’s personal account being one of them, this leads us to believe that is likely Loo’s alt-X account.\nLoo’s likely alt-X account follows his coaching account and is followed by his personal account (Source: X)\nThis account also shared some of the same content as Loo’s Minds profile. In addition to posting\ntweets\nwith neo-Nazi phrases like “\n1488\n” and the\n“Fourteen Words\n”, the account also liked another user’s\npost\nthat links to the abovementioned Rise Above Movement Telegram channel where promotional materials for the far-right MMA event at ION gym were posted.\nWhite supremacist and neo-Nazi content posted, reposted, or liked by Loo’s likely alt-X account\n(Source: X)\nBellingcat also found a Telegram account with the same username as the X account that appears to be Loo’s alt. The Telegram account features a picture of Loo as its profile image — this is the same photo Loo uses for his personal Facebook and Instagram accounts, but with a glowing effect photoshopped over his eyes. The photo with the effect is also shared on the Google Map listing for ION, used as the profile picture for Loo’s Gettr account, and is an older Facebook profile picture.\nScreenshots of Loo’s profiles on various social media platforms, with coloured boxes showing commonalities between websites.\nThis account is a public member of a number of conservative, conspiracy, White pride, and far-right groups including an explicitly neo-Nazi chat group called the “National Socialist Workers Party”. This group describes itself with terms like “Holocaust Deniers,” “Aryan Christians,” and “1488.”\nScreenshot of messages and member list for Telegram channel “National Socialist Workers Party”, showing Loo as a member, with insets of Loo’s profile picture and the group’s profile picture.\nOne thing that is clear is that we can expect these types of combat sports events in the future as they function as a recruiting tool, funding mechanism, and community-building atmosphere. As Pruden told Bellingcat, “It’s important to call attention to the fact that these groups are making concerted efforts to co-opt mainstream mixed martial arts. They’re doing that because they see these sorts of combat sports as capable of mainstreaming far-right beliefs”.\nWhite nationalist groups aren’t in some remote part of the world, but oftentimes right in our own neighbourhood.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Instagram\nhere\n, X\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/unnamed-9-1200x285.png",
+ "alt": "",
+ "caption": "Photo still taken from video (left) showing a logo (red box) in the office space. Zoomed-in inset of logo (right)."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-4-1200x800.png",
+ "alt": "",
+ "caption": "Photo (left) shared on Telegram and photo from Google Street View from the gym (right)."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-3.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-2-1200x801.png",
+ "alt": "",
+ "caption": "Photo from Telegram of the event, with a yellow box around a Hammerskins symbol."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-7-1200x950.png",
+ "alt": "",
+ "caption": "Photos shared on Telegram (Top). The highlighted blue box indicates a portion of the gym’s banner, the green box indicates the fitness drink promotional material, and the purple box indicates redacted social media contact information. Photos from the gym’s rental listing (bottom left) and from Google Maps listing (bottom right) match the features from the shared Telegram photos."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-6-1200x675.png",
+ "alt": "",
+ "caption": "Photo still taken from video trailer (left) with a workout peg board (blue box) and lighting fixture to the right. Photo from gym’s rental listing (right) matching the placement of the peg board and light."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-5.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-9-1200x504.png",
+ "alt": "",
+ "caption": "ION Strength and Conditioning is registered as a Fictitious Business name in Orange County, CA, to Loo and another owner. (Source: Orange County Clerk-Recorder Department)"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-10.png",
+ "alt": "",
+ "caption": "Screenshot of Loo’s profile on the social media platform Minds, showing White Nationalist-related content (left). Screenshot from the ION Strength and Conditioning TikTok page with a Julius Evola quote, which was reposted on Loo’s Gettr profile (right)."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-11.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-12.png",
+ "alt": "",
+ "caption": "Loo’s likely alt-X account follows his coaching account and is followed by his personal account (Source: X)"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-13.png",
+ "alt": "",
+ "caption": "White supremacist and neo-Nazi content posted, reposted, or liked by Loo’s likely alt-X account (Source: X)"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/diagram-1200x658.png",
+ "alt": "",
+ "caption": "Screenshots of Loo’s profiles on various social media platforms, with coloured boxes showing commonalities between websites."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2023/11/image-15-1200x908.png",
+ "alt": "",
+ "caption": "Screenshot of messages and member list for Telegram channel “National Socialist Workers Party”, showing Loo as a member, with insets of Loo’s profile picture and the group’s profile picture."
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2024-03-15_geolocating-a-us-far-right-fight-night.json b/bellingcat_geolocation_json/2024-03-15_geolocating-a-us-far-right-fight-night.json
new file mode 100644
index 0000000000000000000000000000000000000000..4184ce68af6dc05c96cf216af0bb7f4b842fffe7
--- /dev/null
+++ b/bellingcat_geolocation_json/2024-03-15_geolocating-a-us-far-right-fight-night.json
@@ -0,0 +1,47 @@
+{
+ "slug": "geolocating-a-us-far-right-fight-night",
+ "title": "Geolocating a US Far-Right Fight Night",
+ "url": "https://www.bellingcat.com/resources/2024/03/15/geolocating-a-us-far-right-fight-night/",
+ "date": "2024-03-15T17:01:17+00:00",
+ "modified": "2026-02-12T10:25:53+00:00",
+ "author": "Michael Colborne",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "When white nationalists met up in a small town in north Texas to take part in yet another fight night event, we set about locating where it took place.",
+ "content_html": "
\n
When white nationalists met up in a small town in north Texas to take part in yet another fight night event, we set about locating where it took place.
On March 8, 2024, a Telegram channel associated with US far-right extremist Robert Rundo posted eight photos of the combat sports event (Bellingcat is not naming the outlet to avoid amplification). The outlet promised that a promotional “documentary” film of the event would be “coming soon” and claimed the event took place in north Texas. In addition, the outlet claimed that Rundo’s far-right fashion brand was responsible for organising the event.
\n
Bellingcat only needed the very first photo in the post, and a few minutes of web searching, to geolocate the event to a hall available for rent in Muenster, a town of fewer than 2,000 people in north Texas.
\n
Here’s how we did it:
\n
A photo posted alongside the announcement of the fight night on Telegram. The image has been cropped on the sides from the original to obscure branding and avoid amplification.\n
The first hint was the American flag.
\n
In a Telegram chat discussing the event, one person commented on the presence of the large US flag hung in the building. Someone who claimed to have attended the event replied “it was part of the venues [sic] decoration so it couldn’t really be removed.” This led Bellingcat to assume that other public photos of the venue would likely have a large American flag hanging inside.
\n
The second hint — and the one that led Bellingcat right to the venue — was a small detail on the table in the front of the image.
\n
The letters ‘MVFD’ are visible on the table (though upside down), suggesting that the table belongs to a volunteer fire department (i.e., ‘VFD: Volunteer Fire Department’).
\n
\n
A cropped and rotated version of the photo showing the letters ‘MVFD.’
\n
With this clue, Bellingcat then searched for ‘MVFD Texas’.
\n
We found results for several towns beginning with the letter ‘M’ in north Texas whose fire departments used that abbreviation, and checked available images from these towns to see if anything resembled the venue seen in the photos.
\n
Scrolling results from Muenster, Bellingcat noticed a photo of a similar-looking venue, including a similar ceiling, windows and two doors at one end.
\n\n
\n
\nGoogle result showing a venue with matching features, and the full-size image from the MVFD’s Facebook page.\n
Next, we searched for “hall Muenster Texas” and found more photos of a venue — Fuhrmann Hall at the Heritage Park Events Center, a venue available to the public to rent.
\n
We were able to match the the large American flag, the ceiling, the air conditioners and the light fixtures and doors to the images posted in Telegram.
\n
Thus, Bellingcat was able to locate the Patriot Front event to the Fuhrmann Hall located an hour’s drive from the Dallas-Fort Worth Metroplex.
\n\n
\n
\nTwo images of the Fuhrmann Hall, from a user-uploaded photo on Google Maps and an image of the hall posted on Facebook. \n
Bellingcat contacted the Heritage Park Events Center and asked them whether they were aware that they had rented their facilities to representatives of a far-right network. We had not received a response at time of publication.
\n\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter hereand Mastodon here.
\n
",
+ "content_text": "When white nationalists met up in a small town in north Texas to take part in yet another fight night event, we set about locating where it took place.\nOn March 8, 2024, a Telegram channel associated with US far-right extremist Robert Rundo posted eight photos of the combat sports event (Bellingcat is not naming the outlet to avoid amplification). The outlet promised that a promotional “documentary” film of the event would be “coming soon” and claimed the event took place in north Texas. In addition, the outlet claimed that Rundo’s far-right fashion brand was responsible for organising the event.\nBellingcat only needed the very first photo in the post, and a few minutes of web searching, to geolocate the event to a hall available for rent in Muenster, a town of fewer than 2,000 people in north Texas.\nHere’s how we did it:\nA photo posted alongside the announcement of the fight night on Telegram. The image has been cropped on the sides from the original to obscure branding and avoid amplification.\nThe first hint was the American flag.\nIn a Telegram chat discussing the event, one person commented on the presence of the large US flag hung in the building. Someone who claimed to have attended the event replied “it was part of the venues [sic] decoration so it couldn’t really be removed.” This led Bellingcat to assume that other public photos of the venue would likely have a large American flag hanging inside.\nThe second hint — and the one that led Bellingcat right to the venue — was a small detail on the table in the front of the image.\nThe letters ‘MVFD’ are visible on the table (though upside down), suggesting that the table belongs to a volunteer fire department (i.e., ‘VFD: Volunteer Fire Department’).\nA cropped and rotated version of the photo showing the letters ‘MVFD.’\nWith this clue, Bellingcat then searched for ‘MVFD Texas’.\nWe found results for several towns beginning with the letter ‘M’ in north Texas whose fire departments used that abbreviation, and checked available images from these towns to see if anything resembled the venue seen in the photos.\nScrolling results from Muenster, Bellingcat noticed a photo of a similar-looking venue, including a similar ceiling, windows and two doors at one end.\nGoogle result showing a venue with matching features, and the full-size image from the MVFD’s Facebook page.\nNext, we searched for “hall Muenster Texas” and found more photos of a venue — Fuhrmann Hall at the Heritage Park Events Center, a venue\navailable\nto the public to rent.\nWe were able to match the the large American flag, the ceiling, the air conditioners and the light fixtures and doors to the images posted in Telegram.\nThus, Bellingcat was able to locate the Patriot Front event to the Fuhrmann Hall located an hour’s drive from the Dallas-Fort Worth Metroplex.\nTwo images of the Fuhrmann Hall, from a user-uploaded photo on Google Maps and an image of the hall posted on Facebook.\nBellingcat contacted the Heritage Park Events Center and asked them whether they were aware that they had rented their facilities to representatives of a far-right network. We had not received a response at time of publication.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/image1-1.png",
+ "alt": "",
+ "caption": "A photo posted alongside the announcement of the fight night on Telegram. The image has been cropped on the sides from the original to obscure branding and avoid amplification."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/image6-1.png",
+ "alt": "",
+ "caption": "A cropped and rotated version of the photo showing the letters ‘MVFD.’"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/image5-1.png",
+ "alt": "",
+ "caption": "Google result showing a venue with matching features, and the full-size image from the MVFD’s Facebook page."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/image2-1200x900.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/Screenshot-2024-03-15-at-12.51.06-1200x931.png",
+ "alt": "",
+ "caption": "Two images of the Fuhrmann Hall, from a user-uploaded photo on Google Maps and an image of the hall posted on Facebook."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/Screenshot-2024-03-15-at-12.49.26-1-1200x446.png",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2024-03-27_fugitive-us-militant-ammon-bundy-geolocated-to-utah.json b/bellingcat_geolocation_json/2024-03-27_fugitive-us-militant-ammon-bundy-geolocated-to-utah.json
new file mode 100644
index 0000000000000000000000000000000000000000..33552b66fe3299c69ee6b38ee689462da0be6dc3
--- /dev/null
+++ b/bellingcat_geolocation_json/2024-03-27_fugitive-us-militant-ammon-bundy-geolocated-to-utah.json
@@ -0,0 +1,57 @@
+{
+ "slug": "fugitive-us-militant-ammon-bundy-geolocated-to-utah",
+ "title": "Fugitive US Militant Ammon Bundy Geolocated to Utah",
+ "url": "https://www.bellingcat.com/news/2024/03/27/fugitive-us-militant-ammon-bundy-geolocated-to-utah/",
+ "date": "2024-03-27T22:24:25+00:00",
+ "modified": "2024-03-27T23:40:59+00:00",
+ "author": "Michael Lee",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Following a November 2023 warrant, Bundy went into hiding but continued to broadcast to his supporters.",
+ "content_html": "
\n
In February 2024, American anti-government agitator Ammon Bundy posted a video with a provocative title to his YouTube channel: ‘Want to know where Ammon Bundy is?’
\n
Earlier in 2023, Bundy was ordered by an Idaho court to pay more than US$50 million in damages after accusing the state’s largest hospital of child trafficking and leading a campaign of harassment against its staff. A few months later, in November 2023, a warrant was issued for Bundy’s arrest by an Idaho judge.
\n
He promptly went into hiding with his whereabouts not publicly known — until now.
\n
\n
\n
Whether he meant to or not, Bundy provided just enough information in the February video — a few frames of a school district calendar on a refrigerator — to lead Bellingcat to his location in Washington County, in southern Utah.
\n
Bellingcat reached out to the Ada County Sheriff’s Office, which issued the arrest warrant, for comment on this piece; we received no reply at the time of publication.
“I have no concerns, I have not been hiding but working and providing for my family”, said Bundy in a message sent to Bellingcat shortly after the publication of this article. “At anytime peace officers could find me if they wish. I pray that Utah has a stronger sense to protect free speech than Idaho does”, he wrote.
\n
Who is Ammon Bundy?
\n
Ammon Bundy is a prominent American anti-government militant, a one-time candidate for governor of the state of Idaho and the son of Nevada rancher Cliven Bundy. In 2014, Cliven Bundy and his sons led an armed standoff against Bureau of Land Management rangers who were attempting to seize hundreds of cattle that the Bundy family had reportedly let graze on public land without paying taxes since 1993.
Ammon Bundy refused to participate in the St. Luke’s case against him and did not show up in court. He appeared at the Ada County Courthouse to represent himself at an arraignment hearing on his contempt charge on Tuesday, Aug. 29, 2023, in Boise, Idaho. (Sarah A. Miller/Idaho Statesman/TNS/ABACAPRESS.COM\n
Robert Futrell, a professor of sociology at the University of Nevada at Las Vegas, told Bellingcat that Ammon Bundy was someone “who sees himself leading the righteous few against the wicked — defined as those he and his family and followers see as limiting their rights.”
\n
According to Futrell, who researches social movements and political extremism in the U.S., the Bundy family including Ammon ascribe to a “unique apocalyptic version” of the Mormon faith, and see themselves as being on a mission from God. They believe, says Futrell, that the U.S. constitution was divinely inspired and that the foundation of the United States is a step toward establishing a ‘Kingdom of Heaven’ on Earth. The Bundys, said Futrell, “use this divine vision to justify all kinds of actions.”
\n
In 2023 Bundy spearheaded a series of protests against Idaho’s biggest hospital, the St. Luke’s Boise Medical Center, after the state’s health and welfare authorities removed and hospitalized a baby, reportedly the grandson of one of his associates, for malnutrition.
“The defamation lawsuit greatly diminished [Bundy’s] stature and organisation,” said Devin Burghart, president and executive director of the Institute for Research & Education on Human Rights (IREHR), who produced a report on the activities of Bundy’s network. Burghart was an expert witness for the plaintiffs in Bundy’s civil case.
A screenshot of Ammon Bundy’s arrest warrant and mugshot from the Ada County Sheriff’s Office
\n
In an email to the judge, Bundy wrote, “My entire life has been consumed by political prosecutions and I must now do what is necessary to sustain my family.”
In the aforementioned February 24, 2024, YouTube video, Bundy talks to the camera while walking through a house. In several frames near the beginning of the footage, as Bundy walks through a kitchen, what appears to be a school calendar is visible on the refrigerator.
As reported in The Atlantic, rumours quickly surfaced after Bundy’s disappearance that he had fled to southern Utah, but there were no first-hand accounts or substantive proof. On this basis, Bellingcat searched for online versions of local school calendars in the area, and found a match — a calendar for Iron County Schools in southern Utah for January to May 2024.
\n
\n
Left: The Iron County Schools calendar for January to May 2024. Right: A screenshot from a video posted by Ammon Bundy to YouTube on February 20, 2024, titled “Want to know where Ammon Bundy is?”
\n
This was not the only video Bundy posted since the November 2023 warrant.
Bundy has a devoted following on social media where he regularly posts videos to tens of thousands of followers. It took only weeks after Bundy’s disappearance for him to resume his online activity.
For example, between December 2023 and March 2024, Bundy posted four videos from an indoor location. The consistent background and wall decoration across this footage indicates that he has very likely been staying in the same place since December 2023.
\n
\n
Screenshots from videos posted by Ammon Bundy to YouTube between December 2023 and March 2024.
\n
Two of these videos provided further corroboration of Bundy’s location in southern Utah.
A day after posting the video including the school calendar, Bundy posted an 18-minute livestream to YouTube titled “Why our country will not survive.” In this video, Bundy paces outside while speaking to the camera.
Visible in these videos are a number of distinguishing features that allowed Bellingcat to narrow down Bundy’s location to a specific property in Washington County, near its border with Iron County. Bellingcat has chosen not to publicly identify the exact location in this article.
\n
The first step in the geolocation process was to locate a group of mountains in the background of several of the frames.
Bellingcat identified possible mountains near Iron County and searched for possible vantage points of those mountains using Google Earth and Google Street View imagery.
\n
By using the photofit function on PeakVisor, a tool originally created for mountaineers, Bellingcat found that the landscape in Bundy’s livestream matched to Mount Baldy in Washington County, Utah, as viewed facing south-west.
\n
\n
Top: Frame from a livestream posted by Ammon Bundy to YouTube on February 21, 2024, titled “Why our country will not survive, with overlaid PeakVisor photofit showing close agreement between the mountain peaks and the livestream frame. Bottom: Google Street View capture of the same mountains.
\n
A view seen in one of Bundy’s livestreams from February 21, 2024, allowed us to match the mountains to these hills located in Washington County, which allowed us to locate where the livestream was filmed.
\n
Livestreaming from Utah
\n
A subsequent livestream from Bundy gave more confirmation of his presence in southern Utah. On March 18, 2024, he posted another livestream to YouTube, titled ‘$10.000 is still up for grabs’.
\n
Bundy filmed the eight-minute video while driving a truck on a highway with one of his children in the passenger seat. During the video, he stated that members of his family “came to visit us last week” and that he and his child were currently “driving up north together.”
\n
\n
Screenshot from Ammon Bundy’s March 18 YouTube video. Annotation by Bellingcat.
\n
Thanks to visual clues alongside the highway as Bundy drove, Bellingcat was able to geolocate the video to a northbound stretch of Interstate 15 in southern Utah. One of these clues was a red signpost along the side of the road.
The video was filmed entirely in Beaver County, which borders Iron County to the north.
\n
Google Streetview image of the same section of Interstate 15 showing matching elements. Annotation by Bellingcat.\n
The IREHR’s Devin Burghart cautions that, despite his network diminishing in size thanks to the defamation ruling, Bundy and his followers could still be dangerous. Burghart told Bellingcat that Bundy exploited the COVID-19 pandemic to build a network of tens of thousands across the U.S. and Canada.
\n
“[Bundy] cultivated a dangerous network of militia members, conspiracists, anti-vaxxers, white nationalists, and others into an army of followers,” Burghart said. “While he may be in hiding, remnants of that network still remain.”
\n
Leah Sottile, a journalist who has covered the Bundy family extensively, described Bundy as a “far-right provocateur” and further warned that his followers could pose a danger in the future, based on their actions in the past.
\n
“Ammon Bundy has consistently sought the aid of well-armed people since he came into the public eye in 2014,” Sottile told Bellingcat. “Given his ever-increasing rhetoric and this pattern of seeking armed vigilantes to help him, I can’t see why he, or the people around him, would do anything different now.”
\n\n
This article was updated on March 28, 2024 to include comment from Ammon Bundy.
Michael Colborne and Tristan Lee contributed research.
\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter hereand Mastodon here.
\n
",
+ "content_text": "In February 2024, American anti-government agitator Ammon Bundy posted a video with a provocative title to his YouTube channel: ‘Want to know where Ammon Bundy is?’\nEarlier in 2023, Bundy was\nordered\nby an Idaho court to pay more than US$50 million in damages after accusing the state’s largest hospital of child trafficking and leading a campaign of harassment against its staff. A few months later, in November 2023, a\nwarrant\nwas issued for Bundy’s arrest by an Idaho judge.\nHe promptly\nwent into hiding\nwith his whereabouts not publicly known — until now.\nWhether he meant to or not, Bundy provided just enough information in the February video — a few frames of a school district calendar on a refrigerator — to lead Bellingcat to his location in Washington County, in southern Utah.\nBellingcat reached out to the Ada County Sheriff’s Office,\nwhich issued the arrest warrant\n, for comment on this piece; we received no reply at the time of publication.\n“I have no concerns, I have not been hiding but working and providing for my family”, said Bundy in a message sent to Bellingcat shortly after the publication of this article. “At anytime peace officers could find me if they wish. I pray that Utah has a stronger sense to protect free speech than Idaho does”, he wrote.\nWho is Ammon Bundy?\nAmmon Bundy is a prominent American anti-government militant, a one-time candidate for governor of the state of Idaho and the son of Nevada rancher Cliven Bundy. In 2014, Cliven Bundy and his sons led an\narmed standoff against Bureau of Land Management rangers\nwho were attempting to seize hundreds of cattle that the Bundy family had reportedly let graze on public land without paying taxes since 1993.\nIn 2016, Ammon Bundy led the\ntakeover of the Malheur Wildlife Refuge in Oregon\n, another standoff with law enforcement which culminated in the\ndeath of a rancher\nand the indictment of 27 of the occupiers. While several of the activists involved in both clashes with federal officials were handed prison sentences,\nthe Bundys were found not guilty of any criminal charges\nin the Malheur incident, and the Nevada standoff case\nresulted in a mistrial\n. Ammon Bundy was later convicted of trespassing and sentenced to community service after\nstorming the Idaho State Capitol\nduring a 2020 protest against coronavirus-related legislation. Rather than complete the required community service, Bundy instead\nattempted to submit hours he had spent working for his campaign for governor\n.\nAmmon Bundy refused to participate in the St. Luke’s case against him and did not show up in court. He appeared at the Ada County Courthouse to represent himself at an arraignment hearing on his contempt charge on Tuesday, Aug. 29, 2023, in Boise, Idaho. (Sarah A. Miller/Idaho Statesman/TNS/ABACAPRESS.COM\nRobert Futrell, a professor of sociology at the University of Nevada at Las Vegas, told Bellingcat that Ammon Bundy was someone “who sees himself leading the righteous few against the wicked — defined as those he and his family and followers see as limiting their rights.”\nAccording to Futrell, who researches social movements and political extremism in the U.S., the Bundy family including Ammon ascribe to a “unique apocalyptic version” of the Mormon faith, and see themselves as being on a mission from God. They believe, says Futrell, that the U.S. constitution was divinely inspired and that the foundation of the United States is a step toward establishing a ‘Kingdom of Heaven’ on Earth. The Bundys, said Futrell, “use this divine vision to justify all kinds of actions.”\nIn 2023 Bundy spearheaded a\nseries of protests\nagainst Idaho’s biggest hospital, the St. Luke’s Boise Medical Center, after the state’s health and welfare authorities removed and hospitalized a baby, reportedly the grandson of one of his associates, for malnutrition.\nThe actions of Bundy and his associates led to a\nlockdown at the\nhospital and the doxxing and threatening of\ndoctors\n,\npolice officers and child protective services (CPS) workers\nas well as a local\njudge\n.\nIn July 2023, an Idaho jury found Ammon Bundy and several associates liable for\n$52.5 million in damages\nin a defamation case stemming from the child’s hospitalisation. Throughout the legal proceedings, Bundy\nrefused to appear in district court\n, leading to his arrest in\nAugust 2023 on contempt charges\n.\n“The defamation lawsuit greatly diminished [Bundy’s] stature and organisation,” said Devin Burghart, president and executive director of the Institute for Research & Education on Human Rights (IREHR), who produced a\nreport\non the activities of Bundy’s network. Burghart was an expert witness for the plaintiffs in Bundy’s civil case.\nIn November 2023,\na new arrest warrant for a second contempt charge\nwas filed against Bundy after he failed to appear for the first day of the trial concerning his earlier contempt charge.\nA screenshot of Ammon Bundy’s arrest warrant and mugshot from the Ada County Sheriff’s Office\nIn an email to\nthe judge, Bundy wrote, “My entire life has been consumed by political prosecutions and I must now do what is necessary to sustain my family.”\nAnd then,\nAmmon Bundy disappeared\n.\n‘Want to know where Ammon Bundy is?’\nIn the aforementioned February 24, 2024, YouTube video, Bundy talks to the camera while walking through a house. In several frames near the beginning of the footage, as Bundy walks through a kitchen, what appears to be a school calendar is visible on the refrigerator.\nAs\nreported\nin The Atlantic, rumours quickly surfaced after Bundy’s disappearance that he had fled to southern Utah, but there were no first-hand accounts or substantive proof. On this basis, Bellingcat searched for online versions of local school calendars in the area, and found a match — a calendar for Iron County Schools in southern Utah for January to May 2024.\nLeft: The Iron County Schools\ncalendar\nfor January to May 2024. Right: A screenshot from a video posted by Ammon Bundy to YouTube on February 20, 2024, titled “Want to know where Ammon Bundy is?”\nThis was not the only video Bundy posted since the November 2023 warrant.\nBundy has a devoted following on social media where he regularly posts videos to tens of thousands of followers. It took only weeks after Bundy’s disappearance for him to\nresume his online activity\n.\nFor example, between December 2023 and March 2024, Bundy posted four videos from an indoor location. The consistent background and wall decoration across this footage indicates that he has very likely been staying in the same place since December 2023.\nScreenshots from videos posted by Ammon Bundy to YouTube between December 2023 and March 2024.\nTwo of these videos provided further corroboration of Bundy’s location in southern Utah.\nA day after posting the video including the school calendar, Bundy posted an 18-minute livestream to YouTube titled “Why our country will not survive.” In this video, Bundy paces outside while speaking to the camera.\nVisible in these videos are a number of distinguishing features that allowed Bellingcat to narrow down Bundy’s location to a specific property in Washington County, near its border with Iron County. Bellingcat has chosen not to publicly identify the exact location in this article.\nThe first step in the geolocation process was to locate a group of mountains in the background of several of the frames.\nBellingcat identified possible mountains near Iron County and searched for possible vantage points of those mountains using Google Earth and Google Street View imagery.\nBy using the photofit function on\nPeakVisor\n, a tool originally created for mountaineers, Bellingcat found that the landscape in Bundy’s livestream matched to Mount Baldy in Washington County, Utah, as viewed facing south-west.\nTop: Frame from a livestream posted by Ammon Bundy to YouTube on February 21, 2024, titled “Why our country will not survive, with overlaid\nPeakVisor\nphotofit showing close agreement between the mountain peaks and the livestream frame. Bottom: Google Street View capture of the same mountains.\nA view seen in one of Bundy’s livestreams from February 21, 2024, allowed us to match the mountains to these hills located in Washington County, which allowed us to locate where the livestream was filmed.\nLivestreaming from Utah\nA subsequent livestream from Bundy gave more confirmation of his presence in southern Utah. On March 18, 2024, he posted another livestream to YouTube, titled ‘$10.000 is still up for grabs’.\nBundy filmed the eight-minute video while driving a truck on a highway with one of his children in the passenger seat. During the video, he stated that members of his family “came to visit us last week” and that he and his child were currently “driving up north together.”\nScreenshot from Ammon Bundy’s March 18 YouTube video. Annotation by Bellingcat.\nThanks to visual clues alongside the highway as Bundy drove, Bellingcat was able to geolocate the video to a northbound stretch of\nInterstate 15\nin southern Utah. One of these clues was a red signpost along the side of the road.\nThe video was filmed entirely in Beaver County, which borders Iron County to the north.\nGoogle Streetview image of the same section of Interstate 15 showing matching elements. Annotation by Bellingcat.\nThe IREHR’s Devin Burghart cautions that, despite his network diminishing in size thanks to the defamation ruling, Bundy and his followers could still be dangerous. Burghart told Bellingcat that Bundy exploited the COVID-19 pandemic to build a network of tens of thousands across the U.S. and Canada.\n“[Bundy] cultivated a dangerous network of militia members, conspiracists, anti-vaxxers, white nationalists, and others into an army of followers,” Burghart said. “While he may be in hiding, remnants of that network still remain.”\nLeah Sottile, a journalist who has\ncovered\nthe Bundy family extensively, described Bundy as a “far-right provocateur” and further warned that his followers could pose a danger in the future, based on their actions in the past.\n“Ammon Bundy has consistently sought the aid of well-armed people since he came into the public eye in 2014,” Sottile told Bellingcat. “Given his ever-increasing rhetoric and this pattern of seeking armed vigilantes to help him, I can’t see why he, or the people around him, would do anything different now.”\nThis article was updated on March 28, 2024 to include comment from Ammon Bundy.\nMichael Colborne and Tristan Lee contributed research.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/Bundy-Map.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/2023-08-29T211200Z_703442659_MT1ABCPR000X63BH6_RTRMADP_3_ABACA-PRESS-2-1200x785.jpg",
+ "alt": "",
+ "caption": "Ammon Bundy refused to participate in the St. Luke’s case against him and did not show up in court. He appeared at the Ada County Courthouse to represent himself at an arraignment hearing on his contempt charge on Tuesday, Aug. 29, 2023, in Boise, Idaho. (Sarah A. Miller/Idaho Statesman/TNS/ABACAPRESS.COM"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/image14.png",
+ "alt": "",
+ "caption": "A screenshot of Ammon Bundy’s arrest warrant and mugshot from the Ada County Sheriff’s Office"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/image11-1200x1067.png",
+ "alt": "",
+ "caption": "Left: The Iron County Schools calendar for January to May 2024. Right: A screenshot from a video posted by Ammon Bundy to YouTube on February 20, 2024, titled “Want to know where Ammon Bundy is?”"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/image13-1040x1200.jpg",
+ "alt": "",
+ "caption": "Screenshots from videos posted by Ammon Bundy to YouTube between December 2023 and March 2024."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/image9-1-1200x1035.png",
+ "alt": "",
+ "caption": "Top: Frame from a livestream posted by Ammon Bundy to YouTube on February 21, 2024, titled “Why our country will not survive, with overlaid PeakVisor photofit showing close agreement between the mountain peaks and the livestream frame. Bottom: Google Street View capture of the same mountains."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/image2-1-675x1200.png",
+ "alt": "",
+ "caption": "Screenshot from Ammon Bundy’s March 18 YouTube video. Annotation by Bellingcat."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/03/image10-1200x483.png",
+ "alt": "",
+ "caption": "Google Streetview image of the same section of Interstate 15 showing matching elements. Annotation by Bellingcat."
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2024-07-16_dubai-uae-cartel-organised-crime-geolocation-open-source-guide-technique-tools.json b/bellingcat_geolocation_json/2024-07-16_dubai-uae-cartel-organised-crime-geolocation-open-source-guide-technique-tools.json
new file mode 100644
index 0000000000000000000000000000000000000000..beaa801e925f20d1997afb75a9369fdbe9bbd6a0
--- /dev/null
+++ b/bellingcat_geolocation_json/2024-07-16_dubai-uae-cartel-organised-crime-geolocation-open-source-guide-technique-tools.json
@@ -0,0 +1,52 @@
+{
+ "slug": "dubai-uae-cartel-organised-crime-geolocation-open-source-guide-technique-tools",
+ "title": "Exploring the Skyline: How we Located an Alleged Cartel Member in Dubai",
+ "url": "https://www.bellingcat.com/resources/2024/07/16/dubai-uae-cartel-organised-crime-geolocation-open-source-guide-technique-tools/",
+ "date": "2024-07-16T11:41:21+00:00",
+ "modified": "2024-07-26T09:53:52+00:00",
+ "author": "Aiganysh Aidarbekova",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "How we used features of the skyline to narrow down the whereabouts of an alleged cartel member in the UAE.",
+ "content_html": "
\n
In this guide we’ll show you how we narrowed down where an alleged cartel member was staying in a luxury Dubai skyscraper.
\n
By following the steps outlined below, Bellingcat helped establish that in 2023 Dženis Kadrić was renting a Dubai property owned by Candido Nsue Okomo the brother-in-law of Equatorial Guinea’s President. This was later confirmed by documentation obtained by investigative partners.
\n
This finding contributed to a recent investigation by KRIK, Diario Rombe and OCCRP into dirty money in Dubai real estate.
\n
Kadrić, a former police officer, was arrested in Bosnia in February of this year in relation to alleged involvement in organised crime and money laundering. He is currently under house arrest in the country.
\n
The Dubai Skyline
\n
Prior to Kadrić’s arrest, his wife’s posts of her designer outfits for her tens of thousands of followers often included a beautiful view of the Dubai skyline.
\n
This gave us a starting point for where to look.
\n
The first thing we noticed was several odd shaped pools in the background of her posts. These distinctive pools quickly led us to the tallest building in the world- the Burj Khalifa.
\n
\n
Dženis Kadrić seen in a TikTok video of his wife. Source: Tiktok.
\n
The pools seen in the background of the posts, matched those located at the bottom of the Burj Khalifa complex. We can also see a low building with a flat roof -the Dubai Opera House, and a few skyscrapers surrounding the building that we were able to identify.
\n
Sources: Instagram, Google Earth Pro.\n
Next we started to narrow down where in the building of 168 floors, the posts were taken.
\n
In the following photo, the camera is held straight, giving us a good view of Burj Vista Tower 1 and Burj Vista Tower 2. The first one is 65 storeys high, and the second one is 20 floors high, indicating the floor is somewhere between those two.
In another photo, a side of Burj Khalifa is seen with its curved wall windows, and metal rings encompassing a number of floors.
\n
A different view from the Armani Hotel which is located on the 38th floor of the Burj Khalifa shows similar metal rings.
\n
Looking at Burj Khalifa’s plans, we find that those rings are mechanical floors, and there are seven of them spread out across the Burj Khalifa.
\n
\n
A photo of Kadrić’s wife posted on Instagram with a view of the connecting wing of the Burj Khalifa. The photo shows black rings representing mechanical floors- highlighted by us in red.
\n
We are not identifying the exact floor for privacy reasons, but here are the next steps we took to narrow down the location.
\n
There were a few options to determine the floor number.
\n
The easiest option is to look at perspective angles of the floors visible.
\n
Finding Eye Level in an Image
\n
In the post above, the floors are parallel to each other, but when taking a photo (or just simply looking with your own eyes), you can see that the top floor lines are going one direction, while the floors on the bottom are going the opposite direction.
\n
These lines are called the perspective angles. In the image below you can trace the perspective angles, as the demonstrator has marked two in orange. To find the eye level you need to continue the lines until they cross- and where they cross indicates the eye level (the yellow line in the image).
\n
In other words you need to find the true horizontal line amidst all the angled perspective.
\n
In the image above, the orange lines represent the perspective lines, and where they cross we can find the eye level -marked with a yellow line. Source: Youtube, How to Find Eye Level Practical Perspective. \n
For a more detailed explanation of finding the eye level in an image, you can watch this video from which the above picture is taken.
\n
Let’s apply the same method to our image. We build perspective lines in orange by drawing lines along the floor numbers and continue them until they cross. At the point where they cross we draw a yellow line – representing the eye level. This gives us the true eye level in the image which can be used to count the floors.
\n
\n
The same photo posted above, this time with lines of perspective (orange) and the eye level marked in yellow.
\n
Creating a 3D Model
\n
Another option to narrow down the floor is by building a 3D model. The process of building a model of your area of interest has been used in a number of investigations.
\n
Due to the popularity of the Burj Khalifa there were already a few free 3D models available for us to review. Using photos and videos from social media, as well as property listings, we created a 3D model of the building using Blender, a free and open source 3D software. Given these details, we were able to recreate the view which matches the original almost perfectly.
\n
\n
\n
\n
\n\n
Thomas Bordeaux, a member of Bellingcat’s Global Authentication Project and Masters of Architecture Student with experience working with perspective matching, also independently confirmed the floor number.
\n
Final Tips
\n
In this case study, we had a good starting point for our investigation- the fact that the Burj Khalifa is such a popular building meant we could find lots of information about it via open sources- including apartments listed for sale and detailed descriptions of apartments.
\n
Left: Post on TikTok, Right: Example of a listing for an apartment in the Burj Khalifa. Sources: TikTok and Hausnal Real Estate.\n
In addition to material posted by Kadrić’s wife, we were able to further narrow down the location by using open source material about the Burj Khalifa to locate the floor and wing.
\n
Our geolocation work was confirmed by KRIK and Diario Rombe working as part of a collaborative team with OCCRP who found documents that confirmed that Dženis Kadrić was indeed listed as a renter of the apartment in the Burj Khalifa, while Candido Nsue Okomo was listed as the owner. The geolocation of the apartment helped establish a peculiar link between these two individuals and was one piece of the puzzle uncovered for OCCRP’s investigation into dirty money and real estate in Dubai.
You can read OCCRP’s full investigation here , KRIK’s story in Serbian here and Diario Rombe’s report in Spanish here.
\n\n
The 3D model featured in this story was created by Michael Sheldon.
\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter hereand Mastodon here.
\n
",
+ "content_text": "In this guide we’ll show you how we narrowed down where an alleged cartel member was staying in a luxury Dubai skyscraper.\nBy following the steps outlined below, Bellingcat helped establish that in 2023 Dženis Kadrić was renting a Dubai property owned by Candido Nsue Okomo the brother-in-law of Equatorial Guinea’s President. This was later confirmed by documentation obtained by investigative partners.\nThis finding contributed to a recent investigation by KRIK, Diario Rombe and OCCRP into dirty money in Dubai real estate.\nKadrić, a former police officer, was arrested in Bosnia in February of this year in relation to alleged involvement in organised crime and money laundering. He is currently\nunder house arrest\nin the country.\nThe Dubai Skyline\nPrior to Kadrić’s arrest, his wife’s posts of her designer outfits for her tens of thousands of followers often included a beautiful view of the Dubai skyline.\nThis gave us a starting point for where to look.\nThe first thing we noticed was several odd shaped pools in the background of her posts. These distinctive pools quickly led us to the tallest building in the world- the Burj Khalifa.\nDženis Kadrić seen in a TikTok video of his wife. Source: Tiktok.\nThe pools seen in the background of the posts, matched those located at the bottom of the Burj Khalifa complex. We can also see a low building with a flat roof -the Dubai Opera House, and a few skyscrapers surrounding the building that we were able to identify.\nSources: Instagram, Google Earth Pro.\nNext we started to narrow down where in the building of 168 floors, the posts were taken.\nIn the following photo, the camera is held straight, giving us a good view of Burj Vista Tower 1 and Burj Vista Tower 2. The first one is 65 storeys high, and the second one is 20 floors high, indicating the floor is somewhere between those two.\nSources: Instagram,\nPropSearch.ae\n.\nIn another photo, a side of Burj Khalifa is seen with its curved wall windows, and metal rings encompassing a number of floors.\nA different\nview from the Armani Hotel\nwhich is located on the 38th floor of the Burj Khalifa shows similar metal rings.\nLooking at Burj Khalifa’s plans, we find that those rings are mechanical floors, and there are seven of them spread out across the Burj Khalifa.\nA photo of Kadrić’s wife posted on Instagram with a view of the connecting wing of the Burj Khalifa. The photo shows black rings representing mechanical floors- highlighted by us in red.\nWe are not identifying the exact floor for privacy reasons, but here are the next steps we took to narrow down the location.\nThere were a few options to determine the floor number.\nThe easiest option is to look at perspective angles of the floors visible.\nFinding Eye Level in an Image\nIn the post above, the floors are parallel to each other, but when taking a photo (or just simply looking with your own eyes), you can see that the top floor lines are going one direction, while the floors on the bottom are going the opposite direction.\nThese lines are called the perspective angles. In the image below you can trace the perspective angles, as the demonstrator has marked two in orange. To find the eye level you need to continue the lines until they cross- and where they cross indicates the eye level (the yellow line in the image).\nIn other words you need to find the\ntrue horizontal line\namidst all the angled perspective.\nIn the image above, the orange lines represent the perspective lines, and where they cross we can find the eye level -marked with a yellow line. Source: Youtube,\nHow to Find Eye Level Practical Perspective\n.\nFor a more detailed explanation of finding the eye level in an image, you can watch this\nvideo\nfrom which the above picture is taken.\nLet’s apply the same method to our image. We build perspective lines in orange by drawing lines along the floor numbers and continue them until they cross. At the point where they cross we draw a yellow line – representing the eye level. This gives us the true eye level in the image which can be used to count the floors.\nThe same photo posted above, this time with lines of perspective (orange) and the eye level marked in yellow.\nCreating a 3D Model\nAnother option to narrow down the floor is by building a 3D model. The process of building a model of your area of interest has been used in a\nnumber of investigations\n.\nDue to the popularity of the Burj Khalifa there were already a few free 3D models available for us to review. Using photos and videos from social media, as well as property listings, we created a 3D model of the building using\nBlender\n, a free and open source 3D software. Given these details, we were able to recreate the view which matches the original almost perfectly.\nhttps://www.bellingcat.com/app/uploads/2024/07/burj-khalifa-3d-model-export-1.mp4\nThomas Bordeaux, a member of Bellingcat’s Global Authentication Project and Masters of Architecture Student with experience working with perspective matching, also independently confirmed the floor number.\nFinal Tips\nIn this case study, we had a good starting point for our investigation- the fact that the Burj Khalifa is such a popular building meant we could find lots of information about it via open sources- including apartments listed for sale and detailed descriptions of apartments.\nLeft: Post on TikTok, Right: Example of a listing for an apartment in the Burj Khalifa. Sources: TikTok and Hausnal Real Estate.\nIn addition to material posted by Kadrić’s wife, we were able to further narrow down the location by using open source material about the Burj Khalifa to locate the floor and wing.\nOur geolocation work was confirmed by KRIK and Diario Rombe working as part of a collaborative team with OCCRP who found documents that confirmed that Dženis Kadrić was indeed listed as a renter of the apartment in the Burj Khalifa, while Candido Nsue Okomo was listed as the owner. The geolocation of the apartment helped establish a peculiar link between these two individuals and was one piece of the puzzle uncovered for OCCRP’s investigation into dirty money and real estate in Dubai.\nView this post on Instagram\nA post shared by Bellingcat (@bellingcatofficial)\nYou can read OCCRP’s full investigation\nhere\n, KRIK’s story in Serbian\nhere\nand Diario Rombe’s report in Spanish\nhere\n.\nThe 3D model featured in this story was created by Michael Sheldon.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/07/image7.png",
+ "alt": "",
+ "caption": "Dženis Kadrić seen in a TikTok video of his wife. Source: Tiktok."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/07/image11-1200x777.png",
+ "alt": "",
+ "caption": "Sources: Instagram, Google Earth Pro."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/07/image2.png",
+ "alt": "",
+ "caption": "Sources: Instagram, PropSearch.ae ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/07/image5.png",
+ "alt": "",
+ "caption": "A photo of Kadrić’s wife posted on Instagram with a view of the connecting wing of the Burj Khalifa. The photo shows black rings representing mechanical floors- highlighted by us in red."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/07/image8-1200x741.png",
+ "alt": "",
+ "caption": "In the image above, the orange lines represent the perspective lines, and where they cross we can find the eye level -marked with a yellow line. Source: Youtube, How to Find Eye Level Practical Perspective ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/07/image4.png",
+ "alt": "",
+ "caption": "The same photo posted above, this time with lines of perspective (orange) and the eye level marked in yellow."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/07/image9-1200x691.png",
+ "alt": "",
+ "caption": "Left: Post on TikTok, Right: Example of a listing for an apartment in the Burj Khalifa. Sources: TikTok and Hausnal Real Estate."
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2024-08-22_shadow-geolocate-geolocation-locate-image-tool-open-source-bellingcat-measure.json b/bellingcat_geolocation_json/2024-08-22_shadow-geolocate-geolocation-locate-image-tool-open-source-bellingcat-measure.json
new file mode 100644
index 0000000000000000000000000000000000000000..5cb3459fa675c74a02fe662e4881290e2f18b086
--- /dev/null
+++ b/bellingcat_geolocation_json/2024-08-22_shadow-geolocate-geolocation-locate-image-tool-open-source-bellingcat-measure.json
@@ -0,0 +1,72 @@
+{
+ "slug": "shadow-geolocate-geolocation-locate-image-tool-open-source-bellingcat-measure",
+ "title": "Chasing Shadows: Geolocate Images with Bellingcat's Shadow Finder Tool",
+ "url": "https://www.bellingcat.com/resources/2024/08/22/shadow-geolocate-geolocation-locate-image-tool-open-source-bellingcat-measure/",
+ "date": "2024-08-22T09:04:18+00:00",
+ "modified": "2024-08-22T09:04:19+00:00",
+ "author": "Galen Reich, Gabòr Friesen",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Use this new tool to narrow down your search area from the entire globe to just a handful of countries and locations.",
+ "content_html": "
\n
Geolocation is often a time-consuming task.
\n
Researchers often spend hours poring over photos, scouring satellite images and sifting through street view.
But what if there was another way to quickly narrow down your search area?
\n
Bellingcat’s new Shadow Finder Tool, developed with our Discord community, helps you quickly narrow down where an image was taken, by reducing your search area from the entire globe to just a handful of countries and locations.
\n\n
How Shadows Can Help
\n
As many open source researchers know, shadows seen in images can provide extremely useful clues about when and where an image was taken.
\n
By comparing the height of an object with the length of its shadow, you can learn how high the sun is in the sky.
\n
At any moment, there is a single point on the earth where the sun is directly overhead, called the subsolar point.
At that point, a vertical stick would cast no shadow. If you travel away from that point, the sun will appear lower in the sky the further away you go. When the sun is lower in the sky, the shadow of your vertical stick will grow longer.
\n
The distance a place is from the subsolar point determines exactly how high the sun is in the sky there.
\n
At any given time, there is always a ring of places on the earth where the sun is at the same angle above the horizon.
\n
\n
\n
If you know the date and time an image was taken, and can make accurate measurements of the height of an object and the length of its shadow, you can identify the ring of possible areas where the image was taken.
\n
Tools such as SunCalc allow you to manually check a shadow’s length at specific places.
The Shadow Finder Tool removes the need for manual checks, and lets you search the entire surface of the earth with just the click of a button.
So, you don’t need to be a geometry expert, the tool does these calculations for you!
\n
Testing the Tool: A Shore Thing
\n
Let’s use Bellingcat’s Shadow Finder Tool to narrow down the possible locations of this sunny seaside image.
The first step is to work out the date and time the image was taken. There might be a clock in the frame, or helpful metadata contained in the file (use tools like Metadata2go or EXIFmeta to check for this).
It’s not always possible to work out the date and time, and metadata might not always be accurate, so keep this in mind before going any further.
\n
In this case, our sandy beach image is a snapshot from a livestream, so we know exactly when it was taken: 2024-07-10 at 10:30:46 UTC.
\n
You can enter the time in UTC (Universal Coordinated Time) or in the local time. It doesn’t matter if you don’t know the timezone. When you use local time, the tool converts the time into the local timezone for every location. Because of this, you won’t get a nice ring if you use local time, but the principle is the same.
\n
Select an Object to Measure
\n
Once you have confirmed the date and time, identify a clear, visible object and shadow in your image.
Measure the height of the object and the length of its shadow. Sometimes you’ll be able to measure the object and shadow directly in the image. Be aware that you might need to adjust your measurements based on the perspective of the image. Ideally the object you measure should be at a right angle to the camera.
\n
You can read more about measuring objects in an image and managing the challenge of perspective in this Bellingcat guide.
\n
In our example, let’s focus on the beach shower. There are larger shadows visible in the image, but this object is vertical and has a clear shadow. The shadow also falls on a near-flat surface (we can see standing water from the beach shower so any slope is small). It makes the measurements easier if the object is vertical and the surface is level.
\n
From the camera’s perspective, the shadow is almost at a right angle to the object, so any distortion of measurement is likely to be small. Because of this, we can measure the image directly.
Use an image editor (such as GIMP) or a ruler to measure the object and shadow.
In our image we find that the object height is 273 pixels and the shadow length is 104 pixels. It doesn’t matter what units you measure in (you could measure in millimetres or inches) because it is the proportion between the two measurements we are interested in.
If you measure both the object and the shadow with the same units, Bellingcat’s Shadow Finder Tool will work.
\n
Now, with your height measurements and time and date details to hand, open the Shadow Finder Tool.
\n
Enter the object height and shadow length, as well as the date and time and click ‘▶’ to run the tool.
\n
The tool will run (be patient, it can take a little time to load for the first run) and produce a map.
\n
\n
\n
\n
The map includes a yellow ring representing the likely locations of our seaside photograph.
\n
By inspecting the map more closely we can see that the ring includes several landlocked areas and only a few areas of coastline, narrowing down our search even further.
In this case, the beach is likely on the coast of Algeria, Italy, Croatia, Yemen, Somalia, or Djibouti.
\n
There are still several countries in this list, but we’ve now ruled out lots of locations around the world. We know that looking at Portugal, South Africa, or India would be a waste of time. We can now go back to the image and use other clues to narrow down the location even further.
\n
What Challenges Can Affect Measurements?
\n
There are lots of ways to measure objects and shadows in images, some more reliable than others. We have previously covered how to measure shadows and objects in these guides, here and here.
The following points highlight the ideal conditions to get the best results from the Shadow Finder Tool:
\n\n
The time and date of the image must be known
\n
The object and its shadow must be roughly at right angles to the camera
\n
The shadow must be cast on a surface level with the horizon (a flat floor)
\n
The image should not be significantly distorted by the camera lens (as with a fish-eye lens)
\n\n
Advanced Geolocation: Time and Time Again
\n
If you have several images taken at different times, you can narrow down the location even further with the Shadow Finder Tool. Here is another image taken from the same livestream the following day, a few hours later.
Drag the slider to compare the images and see how the shadows have moved.
\n
\n
If we repeat the process outlined above and run Shadow Finder Tool again, we get a second map of possible locations.
This time the ring indicates a different set of possible locations.
The locations have changed because the Earth has rotated and the subsolar point has moved. Using an image editor, we can overlay the two images including both the new ring of possible locations and the previous ring.
\n
By locating where they overlap, we can identify just two possible locations. Use the slider to see where the locations overlap.
\n
\n
The overlay shows us only two possible locations: Cameroon and Croatia. The Cameroon location is inland, so we can eliminate it and focus on Croatia.
Zooming in leads us to focus on the coast and islands around Split. At this point, there’s no substitute for hard work. There are many beaches in that part of the world, but after searching for some time, it’s possible to find the location. There’s even a street view image in which the webcam is visible.
When geolocating any image, context matters. There are often many ways to narrow down a location without using shadows. But for those images where you know the exact time and date and there is a clear shadow, you’ll find that shadows can quickly point you in the right direction of where to look and swiftly narrow down possible locations of your image.
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter hereand Mastodon here.
\n
",
+ "content_text": "Geolocation is often a time-consuming task.\nResearchers often spend hours poring over photos, scouring satellite images and sifting through street view.\nBut what if there was another way to quickly narrow down your search area?\nBellingcat’s new\nShadow Finder Tool\n, developed with our\nDiscord community\n, helps you quickly narrow down where an image was taken, by reducing your search area from the entire globe to just a handful of countries and locations.\nHow Shadows Can Help\nAs many open source researchers know, shadows seen in images can provide extremely useful clues about\nwhen\nand\nwhere\nan image was taken.\nBy comparing the height of an object with the length of its shadow, you can learn how high the sun is in the sky.\nAt any moment, there is a single point on the earth where the sun is directly overhead, called the subsolar point.\nAt that point, a vertical stick would cast no shadow. If you travel away from that point, the sun will appear lower in the sky the further away you go. When the sun is lower in the sky, the shadow of your vertical stick will grow longer.\nThe distance a place is from the subsolar point determines exactly how high the sun is in the sky there.\nAt any given time, there is always a ring of places on the earth where the sun is at the same angle above the horizon.\nIf you know the date and time an image was taken, and can make accurate measurements of the height of an object and the length of its shadow, you can identify the ring of possible areas where the image was taken.\nTools such as\nSunCalc\nallow you to manually check a shadow’s length at specific places.\nThe Shadow Finder Tool removes the need for manual checks, and lets you search the entire surface of the earth with just the click of a button.\nSo, you don’t need to be a geometry expert, the tool does these calculations for you!\nTesting the Tool: A Shore Thing\nLet’s use Bellingcat’s Shadow Finder Tool to narrow down the possible locations of this sunny seaside image.\nSource:\nSkyline Webcams\n.\nThe first step is to work out the date and time the image was taken. There might be a clock in the frame, or helpful metadata contained in the file (use tools like\nMetadata2go\nor\nEXIFmeta\nto check for this).\nIt’s not always possible to work out the date and time, and metadata might not always be accurate, so keep this in mind before going any further.\nIn this case, our sandy beach image is a snapshot from a livestream, so we know exactly when it was taken:\n2024-07-10\nat\n10:30:46 UTC\n.\nYou can enter the time in UTC (Universal Coordinated Time) or in the local time. It doesn’t matter if you don’t know the timezone. When you use local time, the tool converts the time into the local timezone for every location. Because of this, you won’t get a nice ring if you use local time, but the principle is the same.\nSelect an Object to Measure\nOnce you have confirmed the date and time, identify a clear, visible object and shadow in your image.\nMeasure the height of the object and the length of its shadow. Sometimes you’ll be able to measure the object and shadow directly in the image. Be aware that you might need to adjust your measurements based on the perspective of the image. Ideally the object you measure should be at a right angle to the camera.\nYou can read more about measuring objects in an image and managing the challenge of perspective in this\nBellingcat guide\n.\nIn our example, let’s focus on the beach shower. There are larger shadows visible in the image, but this object is vertical and has a clear shadow. The shadow also falls on a near-flat surface (we can see standing water from the beach shower so any slope is small). It makes the measurements easier if the object is vertical and the surface is level.\nFrom the camera’s perspective, the shadow is almost at a right angle to the object, so any distortion of measurement is likely to be small. Because of this, we can measure the image directly.\nSource:\nSkyline Webcams\n.\nUse an image editor (such as\nGIMP\n) or a ruler to measure the object and shadow.\nIn our image we find that the object height is 273 pixels and the shadow length is 104 pixels. It doesn’t matter what units you measure in (you could measure in millimetres or inches) because it is the proportion between the two measurements we are interested in.\nIf you measure both the object and the shadow with the same units, Bellingcat’s Shadow Finder Tool will work.\nNow, with your height measurements and time and date details to hand, open the\nShadow Finder Tool\n.\nEnter the object height and shadow length, as well as the date and time and click ‘▶’ to run the tool.\nThe tool will run (be patient, it can take a little time to load for the first run) and produce a map.\nThe map includes a yellow ring representing the likely locations of our seaside photograph.\nBy inspecting the map more closely we can see that the ring includes several landlocked areas and only a few areas of coastline, narrowing down our search even further.\nIn this case, the beach is likely on the coast of Algeria, Italy, Croatia, Yemen, Somalia, or Djibouti.\nThere are still several countries in this list, but we’ve now ruled out lots of locations around the world. We know that looking at Portugal, South Africa, or India would be a waste of time. We can now go back to the image and use other clues to narrow down the location even further.\nWhat Challenges Can Affect Measurements?\nThere are lots of ways to measure objects and shadows in images, some more reliable than others. We have previously covered how to measure shadows and objects in these guides,\nhere\nand\nhere\n.\nThe following points highlight the ideal conditions to get the best results from the Shadow Finder Tool:\nThe time and date of the image must be known\nThe object and its shadow must be roughly at right angles to the camera\nThe shadow must be cast on a surface level with the horizon (a flat floor)\nThe image should not be significantly distorted by the camera lens (as with a fish-eye lens)\nAdvanced Geolocation: Time and Time Again\nIf you have several images taken at different times, you can narrow down the location even further with the Shadow Finder Tool. Here is another image taken from the\nsame livestream\nthe following day, a few hours later.\nDrag the slider to compare the images and see how the shadows have moved.\nIf we repeat the process outlined above and run Shadow Finder Tool again, we get a second map of possible locations.\nThis time the ring indicates a different set of possible locations.\nThe locations have changed because the Earth has rotated and the subsolar point has moved. Using an image editor, we can overlay the two images including both the new ring of possible locations and the previous ring.\nBy locating where they overlap, we can identify just two possible locations. Use the slider to see where the locations overlap.\nThe overlay shows us only two possible locations: Cameroon and Croatia. The Cameroon location is inland, so we can eliminate it and focus on Croatia.\nZooming in leads us to focus on the coast and islands around Split. At this point, there’s no substitute for hard work. There are many beaches in that part of the world, but after searching for some time, it’s possible to find the location. There’s even a street view image in which the webcam is visible.\nSources: Google Street Maps and\nGoogle Street View Image\n.\nWhen geolocating any image, context matters. There are often many ways to narrow down a location without using shadows. But for those images where you know the exact time and date and there is a clear shadow, you’ll find that shadows can quickly point you in the right direction of where to look and swiftly narrow down possible locations of your image.\nYou can sign in with a Google account and start using\nBellingcat’s Shadow Finder Tool\n, now.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/08/image8-1-1200x753.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/08/image12-1200x675.jpg",
+ "alt": "",
+ "caption": "Source: Skyline Webcams ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/08/image10-1200x675.jpg",
+ "alt": "",
+ "caption": "Source: Skyline Webcams ."
+ },
+ {
+ "url": "https://lh7-qw.googleusercontent.com/docsz/AD_4nXe44jAhSjvhVnB5OCQwf2uSJ9KapxF_irbcY68PujxARgZJuoGF_Q719iCU4ifRhxAkUsl-yYniPNE6RKlmiztkeU7QlKKGxk2nqOettm7LqhkIB1_DFzyLb_f97OTleYqMhO5LWXzz7rDSvcv63Mu1j4Q?key=wO7lkzKX9VzMSrx0SMMocQ",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/08/ShadowTool.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/08/image12.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/08/image1.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/08/image3.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/08/image4.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/08/image11-1200x683.jpg",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/08/image5-1-1200x712.jpg",
+ "alt": "",
+ "caption": "Sources: Google Street Maps and Google Street View Image ."
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2024-10-17_geolocating-site-where-hamas-leader-yahya-sinwar-was-killed.json b/bellingcat_geolocation_json/2024-10-17_geolocating-site-where-hamas-leader-yahya-sinwar-was-killed.json
new file mode 100644
index 0000000000000000000000000000000000000000..275db97baf80b1015f70b68f209df751b6b7fa7b
--- /dev/null
+++ b/bellingcat_geolocation_json/2024-10-17_geolocating-site-where-hamas-leader-yahya-sinwar-was-killed.json
@@ -0,0 +1,72 @@
+{
+ "slug": "geolocating-site-where-hamas-leader-yahya-sinwar-was-killed",
+ "title": "Geolocating Site Where Hamas Leader Yahya Sinwar Was Killed",
+ "url": "https://www.bellingcat.com/news/2024/10/17/geolocating-site-where-hamas-leader-yahya-sinwar-was-killed/",
+ "date": "2024-10-17T22:26:13+00:00",
+ "modified": "2026-02-11T10:35:13+00:00",
+ "author": "Jake Godin",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "On Thursday, the Israel Defense Forces (IDF) announced the death of Hamas leader Yahya Sinwar during an operation in southern Gaza. As part of its announcement, the IDF’s media channel on Telegram shared images and video shot from the location where they said Sinwar was killed. Using the IDF video as well as the images […]",
+ "content_html": "
\n
On Thursday, the Israel Defense Forces (IDF) announced the death of Hamas leader Yahya Sinwar during an operation in southern Gaza.
\n
As part of its announcement, the IDF’s media channel on Telegram shared images and video shot from the location where they said Sinwar was killed.
\n
Using the IDF video as well as the images shared by the IDF in their Telegram post and videos geolocated previously by volunteers, it is possible to geolocate the house to Gaza’s Tal as-Sutlan area at 31.3055, 34.2467.
\n
A satellite image from Google Earth (top left) shows where an IDF video (bottom left and right) was filmed after the death of Hamas leader Yahya Sinwar. \n
The visuals offer a few clues for where to look when geolocating the house visible in the IDF videos and images. The main ones are a gazebo with a tree next to it and a red building in the background with a palm tree next to it.
\n
An IDF video shows a red building as well as a gazebo and a tree.\n
A white residential-looking tower with a triangular orange roof is also visible in the footage.
\n
A white residential building with a distinctive roof can be seen in an IDF video.\n
Finally, a building appears to be flanking a house that has a distinct set of tiles on the wall.
\n
These tiles were also visible in a photo shared by Times of Israel military correspondent Mannie Fabian. These photos showed a body bag that it was stated contained Sinwar’s corpse.
\n
A still image from an IDF video shows soldiers standing in front of a tiled wall with another residential building visible in the right hand side of the frame.\n
There were initial reports suggesting the incident that led to Sinwar’s death took place in Rafah’s Tal as-Sultan neighbourhood. We therefore began to search this area for some of the visual markers identified above.
\n
The easiest of these to find was the gazebo as it both had a tree next to it and a distinctive hexagonal shaped roof.
\n
One such area with both a gazebo and tree was located at 31.3051, 34.2466 using recent SkySat imagery from Planet Labs PBC.
\n
A satellite image shows what appears to be a gazebo and tree visible in the Tal as-Sultan area of Rafah. Credit: Planet Labs.\n
With an area of interest now further narrowed down, we began to search for previously-verified imagery or videos of this neighbourhood using an interactive map provided by Geoconfirmed, a community-based geolocating platform.
\n
The Geoconfirmed map contained two videos of this area that proved useful. The first was from late May, showing a view from near Rafah’s Emirati Hospital. In the first few seconds of the video, a white residential tower that appears to match the one in the background of the IDF photo can be seen. The tower, located at 31.3055, 34.2455 was geolocated by Geoconfirmed volunteer @KJ_X23.
\n
A comparison of a white building that was visible in an IDF video filmed at the site where Yahya Sinwar was killed (inset left) was also visible in a video filmed earlier this year and geolocated by Geoconfirmed. \n
Whatsmore, the building visible in the background of the aforementioned photo that showed IDF soldiers in front of a distinct set of tiles can be seen in the footage (at 1:47 timestamp) geolocated by Geoconfirmed.
\n
In the video, you can see the building on the left side of the screen as the camera pans to look at the house where Sinwar would ultimately be found. This house has a yellow and white pattern that can be seen in both the Telegram video and image of IDF soldiers.
\n
A building visible in an IDF video (top right inset) was also visible in a Telegram video uploaded by and IDF soldier earlier this year (bottom left inset) and geolocated by Geoconfirmed.\n
The identification of this building was further validated by another video later released by the IDF claiming to show Sinwar before he was killed.
\n
In the first second of the video, the exterior of the building is visible. Compared with the video of the IDF drive-by from September, we can match part of the damage that was still visible in the more recent video. The shape of the windows and colour of the building also match.
\n
A building visible in a Telegram video uploaded by and IDF soldier earlier this year (left) and geolocated by Geoconfirmed matched one that could be seen in a video released by the IDF that claimed to show Yahya Sinwar before he died.\n
While there was no reference imagery of the red building and tree previously detailed, the location lines up from the perspective of the camera when looking south from the building where Sinawar was killed.
\n
Both can be seen behind the gazebo and tree in the IDF video as well as in Google Earth imagery.
\n
A Google Earth satellite image shows the location of a tree and gazebo, as well as two other buildings (left), that could be matched to a video released by the IDF (right)\n
All of these steps helped us establish the house where Sinwar’s body was found. It’s not clear from IDF statements or media reporting how long Sinwar and the other Hamas fighters killed were in the building.
\n
Based on Planet Labs imagery, it’s possible to see that the IDF started operating in the area of the house in late August and early September given that was when buildings and vegetation in the vicinity of the house began to be destroyed.
\n
\n\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Twitter hereand Mastodon here.
\n
",
+ "content_text": "On Thursday, the Israel Defense Forces (IDF)\nannounced\nthe death of Hamas leader Yahya Sinwar during an operation in southern Gaza.\nAs part of its announcement, the IDF’s media channel on Telegram shared images and\nvideo\nshot from the location where they said Sinwar was killed.\nUsing the IDF video as well as the images shared by the IDF in their Telegram post and videos geolocated previously by volunteers, it is possible to geolocate the house to Gaza’s Tal as-Sutlan area at\n31.3055, 34.2467\n.\nA satellite image from Google Earth (top left) shows where an IDF\nvideo\n(bottom left and right) was filmed after the death of Hamas leader Yahya Sinwar.\nThe visuals offer a few clues for where to look when geolocating the house visible in the IDF videos and images. The main ones are a gazebo with a tree next to it and a red building in the background with a palm tree next to it.\nAn IDF\nvideo\nshows a red building as well as a gazebo and a tree.\nA white residential-looking tower with a triangular orange roof is also visible in the footage.\nA white residential building with a distinctive roof can be seen in an IDF\nvideo\n.\nFinally, a building appears to be flanking a house that has a distinct set of tiles on the wall.\nThese tiles were also visible in a\nphoto shared\nby Times of Israel military correspondent Mannie Fabian. These photos showed a body bag that it was stated contained Sinwar’s corpse.\nA still image from an IDF\nvideo\nshows soldiers standing in front of a tiled wall with another residential building visible in the right hand side of the frame.\nThere were\ninitial reports\nsuggesting the incident that led to Sinwar’s death took place in Rafah’s Tal as-Sultan neighbourhood. We therefore began to search this area for some of the visual markers identified above.\nThe easiest of these to find was the gazebo as it both had a tree next to it and a distinctive hexagonal shaped roof.\nOne such area with both a gazebo and tree was located at\n31.3051, 34.2466\nusing recent SkySat imagery from Planet Labs PBC.\nA satellite image shows what appears to be a gazebo and tree visible in the Tal as-Sultan area of Rafah. Credit: Planet Labs.\nWith an area of interest now further narrowed down, we began to search for previously-verified imagery or videos of this neighbourhood using an interactive\nmap provided by Geoconfirmed\n, a community-based geolocating platform.\nThe Geoconfirmed map contained two videos of this area that proved useful. The first was\nfrom late May\n, showing a view from near Rafah’s Emirati Hospital. In the first few seconds of the video, a white residential tower that appears to match the one in the background of the IDF photo can be seen. The tower, located at 31.3055, 34.2455 was\ngeolocated by Geoconfirmed volunteer @KJ_X23\n.\nA comparison of a white building that was visible in an IDF\nvideo\nfilmed at the site where Yahya Sinwar was killed (inset left) was also visible in a\nvideo\nfilmed earlier this year and geolocated by Geoconfirmed.\nAnother video that provides a clue is a drive-by video filmed by an IDF soldier and\nshared online in September\n. The route,\ngeolocated by Geoconfirmed volunteer @NemoAnno\n, passes right by the location near the gazebo.\nWhatsmore, the building visible in the background of the aforementioned photo that showed IDF soldiers in front of a distinct set of tiles can be seen in the footage (at 1:47 timestamp) geolocated by Geoconfirmed.\nIn the video, you can see the building on the left side of the screen as the camera pans to look at the house where Sinwar would ultimately be found. This house has a yellow and white pattern that can be seen in both the Telegram video and image of IDF soldiers.\nA building visible in an IDF\nvideo\n(top right inset) was also visible in a Telegram\nvideo\nuploaded by and IDF soldier earlier this year (bottom left inset) and geolocated by Geoconfirmed.\nThe identification of this building was further validated by another video later\nreleased by the IDF\nclaiming to show Sinwar before he was killed.\nIn the first second of the video, the exterior of the building is visible. Compared with the video of the IDF drive-by from September, we can match part of the damage that was still visible in the more recent video. The shape of the windows and colour of the building also match.\nA building visible in a Telegram\nvideo\nuploaded by and IDF soldier earlier this year (left) and geolocated by Geoconfirmed matched one that could be seen in a\nvideo\nreleased by the IDF that claimed to show Yahya Sinwar before he died.\nWhile there was no reference imagery of the red building and tree previously detailed, the location lines up from the perspective of the camera when looking south from the building where Sinawar was killed.\nBoth can be seen behind the gazebo and tree in the IDF video as well as in Google Earth imagery.\nA Google Earth satellite image shows the location of a tree and gazebo, as well as two other buildings (left), that could be matched to a\nvideo\nreleased by the IDF (right)\nAll of these steps helped us establish the house where Sinwar’s body was found. It’s not clear from IDF statements or media reporting how long Sinwar and the other Hamas fighters killed were in the building.\nBased on Planet Labs imagery, it’s possible to see that the IDF started operating in the area of the house in late August and early September given that was when buildings and vegetation in the vicinity of the house began to be destroyed.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Twitter\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/10/quick-annotation-1200x775.png",
+ "alt": "",
+ "caption": "A satellite image from Google Earth (top left) shows where an IDF video (bottom left and right) was filmed after the death of Hamas leader Yahya Sinwar."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/10/gazebo-and-red-building-annotation-1200x671.png",
+ "alt": "",
+ "caption": "An IDF video shows a red building as well as a gazebo and a tree."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/10/white-residential-tower-annotation-1200x673.png",
+ "alt": "",
+ "caption": "A white residential building with a distinctive roof can be seen in an IDF video ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/10/tiles-and-background-building-anntoated-1200x675.png",
+ "alt": "",
+ "caption": "A still image from an IDF video shows soldiers standing in front of a tiled wall with another residential building visible in the right hand side of the frame."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/10/gazebo-planet-imagery-1200x707.png",
+ "alt": "",
+ "caption": "A satellite image shows what appears to be a gazebo and tree visible in the Tal as-Sultan area of Rafah. Credit: Planet Labs."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/10/white-residential-tower-comparison-1200x673.png",
+ "alt": "",
+ "caption": "A comparison of a white building that was visible in an IDF video filmed at the site where Yahya Sinwar was killed (inset left) was also visible in a video filmed earlier this year and geolocated by Geoconfirmed."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/10/drive-by-video-comparison-VERSION-2-1-1200x1013.png",
+ "alt": "",
+ "caption": "A building visible in an IDF video (top right inset) was also visible in a Telegram video uploaded by and IDF soldier earlier this year (bottom left inset) and geolocated by Geoconfirmed."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/10/window-annotation-1200x513.png",
+ "alt": "",
+ "caption": "A building visible in a Telegram video uploaded by and IDF soldier earlier this year (left) and geolocated by Geoconfirmed matched one that could be seen in a video released by the IDF that claimed to show Yahya Sinwar before he died."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/10/red-building-annotation-1200x479.png",
+ "alt": "",
+ "caption": "A Google Earth satellite image shows the location of a tree and gazebo, as well as two other buildings (left), that could be matched to a video released by the IDF (right)"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/10/august-22-sinwar-location-annotated.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2024/10/september-9-sinwar-location-annotated.png",
+ "alt": "",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2025-06-06_have-llms-finally-mastered-geolocation.json b/bellingcat_geolocation_json/2025-06-06_have-llms-finally-mastered-geolocation.json
new file mode 100644
index 0000000000000000000000000000000000000000..70cd15f3b68a16454ff0a90bb26ab42171f4ea1a
--- /dev/null
+++ b/bellingcat_geolocation_json/2025-06-06_have-llms-finally-mastered-geolocation.json
@@ -0,0 +1,62 @@
+{
+ "slug": "have-llms-finally-mastered-geolocation",
+ "title": "Have LLMs Finally Mastered Geolocation?",
+ "url": "https://www.bellingcat.com/resources/how-tos/2025/06/06/have-llms-finally-mastered-geolocation/",
+ "date": "2025-06-06T07:33:08+00:00",
+ "modified": "2025-07-15T10:10:14+00:00",
+ "author": "Foeke Postma, Nathan Patin",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "We tasked LLMs from OpenAI, Google, Anthropic, Mistral and xAI to geolocate our unpublished holiday snaps. Here's how they did.",
+ "content_html": "
\n
An ambiguous city street, a freshly mown field, and a parked armoured vehicle were among the example photos we chose to challenge Large Language Models (LLMs) from OpenAI, Google, Anthropic, Mistral and xAI to geolocate.
\n
Back in July 2023, Bellingcat analysed the geolocation performance of OpenAI and Google’s models. Both chatbots struggled to identify images and were highly prone to hallucinations. However, since then, such models have rapidly evolved.
\n
To assess how LLMs from OpenAI, Google, Anthropic, Mistral and xAI compare today, we ran 500 geolocation tests, with 20 models each analysing the same set of 25 images.
\n
\n
We chose 25 of our own travel photos, varying in difficulty to geolocate, none of which had been published online before.
\n
Our analysis included older and “deep research” versions of the models, to track how their geolocation capabilities have developed over time. We also included Google Lens to compare whether LLMs offer a genuine improvement over traditional reverse image search. While reverse image search tools work differently from LLMs, they remain one of the most effective ways to narrow down an image’s location when starting from scratch.
\n
The Test
\n\n
We used 25 of our own travel photos, to test a range of outdoor scenes, both rural and urban areas, with and without identifiable landmarks such as buildings, mountains, signs or roads. These images were sourced from every continent, including Antarctica.
\n
The vast majority have not been reproduced here, as we intend to continue using them to evaluate newer models as they are released. Publishing them here would compromise the integrity of future tests.
\n
Each LLM was given a photo that had not been published online and contained no metadata. All models then received the same prompt: “Where was this photo taken?”, alongside the image. If an LLM asked for more information, the response was identical: “There is no supporting information. Use this photo alone.”
\n
We tested the following models:
\n
Developer
Model
Developer’s Description
Anthropic
Claude Haiku 3.5
“fastest model for daily tasks”
Claude Sonnet 3.7
“our most intelligent model yet”
Claude Sonnet 3.7 (extended thinking)
“enhanced reasoning capabilities for complex tasks”
Claude Sonnet 4.0
“smart, efficient model for everyday use”
Claude Opus 4.0
“powerful, large model for complex challenges”
Google
Gemini 2.0 Flash
“for everyday tasks plus more features”
Gemini 2.5 Flash
“uses advanced reasoning”
Gemini 2.5 Pro
“best for complex tasks”
Gemini Deep Research
“get in-depth answers”
Mistral
Pixtral Large
“frontier-level image understanding”
OpenAI
ChatGPT 4o
“great for most tasks”
ChatGPT Deep Research
“designed to perform in-depth, multi-step research using data on the public web”
ChatGPT 4.5
“good for writing and exploring ideas”
ChatGPT o3
“uses advanced reasoning”
ChatGPT o4-mini
“fastest at advanced reasoning”
ChatGPT o4-mini-high
“great at coding and visual reasoning”
xAI
Grok 3
“smartest”
Grok 3 DeepSearch
“advanced search and reasoning”
Grok 3 DeeperSearch
“extended search, more reasoning”
\n\n
This was not a comprehensive review of all available models, partly due to the speed at which new models and versions are currently being released. For example, we did not assess DeepSeek, as it currently only extracts text from images. Note that in ChatGPT, regardless of what model you select, the “deep research” function is currently powered by a version of o4-mini.
\n
\n
\n
Support Bellingcat
\n
Your donations directly contribute to our ability to publish groundbreaking investigations and uncover wrongdoing around the world.
Gemini models have been released in “preview” and “experimental” formats, as well as dated versions like “03-25” and “05-06”. To keep the comparisons manageable, we grouped these variants under their respective base models, e.g. “Gemini 2.5 Pro”.
\n
We also compared every test with the first 10 results from Google Lens’s “visual match” feature, to assess the difficulty of the tests and the usefulness of LLMs in solving them.
\n
We ranked all responses on a scale from 0 to 10, with 10 indicating an accurate and specific identification, such as a neighbourhood, trail, or landmark, and 0 indicating no attempt to identify the location at all.
\n
And the Winner is…
\n\n
ChatGPT beat Google Lens.
\n
In our tests, ChatGPT o3, o4-mini, and o4-mini-high were the only models to outperform Google Lens in identifying the correct location, though not by a large margin. All other models were less effective when it came to geolocating our test photos.
\n
\n
We scored 20 models against 25 photos, rating each from 0 (red) to 10 (dark green) for accuracy in geolocating the images.
\n
Even Google’s own LLM, Gemini, fared worse than Google Lens. Surprisingly, it also scored lower than xAI’s Grok, despite Grok’s well-documented tendency to hallucinate. Gemini’s Deep Research mode scored roughly the same as the three Grok models we tested, with DeeperSearch proving the most effective of xAI’s LLMs.
The highest-scoring models from Anthropic and Mistral lagged well behind their current competitors from OpenAI, Google, and xAI. In several cases, even Claude’s most advanced models identified only the continent, while others were able to narrow their responses down to specific parts of a city. The latest Claude model, Opus 4, performed at a similar level to Gemini 2.5 Pro.
\n
Here are some of the highlights from five of our tests.
\n
A Road in the Japanese Mountains
\n\n
The photo below was taken on the road between Takayama and Shirakawa in Japan. As well as the road and mountains, signs and buildings are also visible.
\n
\n
Test “snowy-highway” depicted a road near Takayama, Japan.
\n
Gemini 2.5 Pro’s response was not useful. It mentioned Japan, but also Europe, North and South America and Asia. It replied:
\n
“Without any clear, identifiable landmarks, distinctive signage in a recognisable language, or unique architectural styles, it’s very difficult to determine the exact country or specific location.”
\n
In contrast, o3 identified both the architectural style and signage, responding:
\n
“Best guess: a snowy mountain stretch of central-Honshu, Japan—somewhere in the Nagano/Toyama area. (Japanese-style houses, kanji on the billboard, and typical expressway barriers give it away.)”
\n
A Field on the Swiss Plateau
\n\n
This photo was taken near Zurich. It showed no easily recognisable features apart from the mountains in the distance. A reverse image search using Google Lens didn’t immediately lead to Zurich. Without any context, identifying the location of this photo manually could take some time. So how did the LLMs fare?
\n
\n
Test “field-hills” depicted a view of a field near Zurich
\n
Gemini 2.5 Pro stated that the photo showed scenery common to many parts of the world and that it couldn’t narrow it down without additional context.
\n
By contrast, ChatGPT excelled at this test. o4-mini identified the “Jura foothills in northern Switzerland”, while o4-mini-high placed the scene ”between Zürich and the Jura mountains”.
\n
These answers stood in stark contrast to those from Grok Deep Research, which, despite the visible mountains, confidently stated the photo was taken in the Netherlands. This conclusion appeared to be based on the Dutch name of the account used, “Foeke Postma”, with the model assuming the photo must have been taken there and calling it a “reasonable and well-supported inference”.
\n
An Inner-City Alley Full of Visual Clues in Singapore
\n\n
This photo of a narrow alleyway on Circular Road in Singapore provoked a wide range of responses from the LLMs and Google Lens, with scores ranging from 3 (nearby country) to 10 (correct location).
\n
\n
Test “dark-alley”, a photo taken of an alleyway in Singapore
\n\n
The test served as a good example of how LLMs can outperform Google Lens by focusing on small details in a photo to identify the exact location. Those that answered correctly referenced the writing on the mailbox on the left in the foreground, which revealed the precise address.
\n
While Google Lens returned results from all over Singapore and Malaysia, part of ChatGPT o4-mini’s response read: “This appears to be a classic Singapore shophouse arcade – in fact, if you look at the mailboxes on the left you can just make out the label ‘[correct address].’”
\n
Some of the other models noticed the mailbox but could not read the address visible in the image, falsely inferring that it pointed to other locations. Gemini 2.5 Flash responded, “The design of the mailboxes on the left, particularly the ‘G’ for Geylang, points strongly towards Singapore.” Another Gemini model, 2.5 Pro, spotted the mailbox but focused instead on what it interpreted as Thai script on a storefront, confidently answering: “The visual evidence strongly suggests the photo was taken in an alleyway in Thailand, likely in Bangkok.”
\n
The Costa Rican Coast
\n\n
One of the harder tests we gave the models to geolocate was a photo taken from Playa Longosta on the Pacific Coast of Costa Rica near Tamarindo.
\n
\n
Test “beach-forest” showed Playa Longosta, Costa Rica.
\n\n
Gemini and Claude performed the worst on this task, with most models either declining to guess or giving incorrect answers. Claude 3.7 Sonnet correctly identified Costa Rica but hedged with other locations, such as Southeast Asia. Grok was the only model to guess the exact location correctly, while several ChatGPT models (Deep Research, o3 and the o4-minis) guessed within 160km of the beach.
\n
An Armoured Vehicle on the Streets of Beirut
\n\n
This photo was taken on the streets of Beirut and features several details useful for geolocation, including an emblem on the side of the armored personnel carrier and a partially visible Lebanese flag in the background.
\n
\n
Test “street-military” depicted an armoured personnel carrier on the streets of Beirut
\n\n
Surprisingly, most models struggled with this test: Claude 4 Opus, billed as a “powerful, large model for complex challenges”, guessed “somewhere in Europe” owing to the “European-style street furniture and building design”, while Gemini and Grok could only narrow the location down to Lebanon. Half of the ChatGPT models responded with Beirut. Only two models, both ChatGPT, referenced the flag.
\n
So Have LLMs Finally Mastered Geolocation?
\n\n
LLMs can certainly help researchers to spot the details that Google Lens or they themselves might miss.
\n
One clear advantage of LLMs is their ability to search in multiple languages. They also appear to make good use of small clues, such as vegetation, architectural styles or signage. In one test, a photo of a man wearing a life vest in front of a mountain range was correctly located because the model identified part of a company name on his vest and linked it to a nearby boat tour operator.
\n
For touristic areas and scenic landscapes, Google Lens still outperformed most models. When shown a photo of Schluchsee lake in the Black Forest, Germany, Google Lens returned it as the top result, while ChatGPT was the only LLM to correctly identify the lake’s name. In contrast, in urban settings, LLMs excelled at cross-referencing subtle details, whereas Google Lens tended to fixate on larger, similar-looking structures, such as buildings or ferris wheels, which appear in many other locations.
\n
\n
Heat map to show how each model performed on all 25 tests
\n
Enhanced Reasoning Modes
\n\n
You’d assume turning on “deep research” or “extended thinking” functions would have resulted in higher scores. However, on average, Claude and ChatGPT performed worse. Only one Grok model, DeeperSearch, and one Gemini, Gemini Deep Research, showed improvement. For example, ChatGPT Deep Research was shown a photo of a coastline and took nearly 13 minutes to produce an answer that was about 50km north of the correct location. Meanwhile, o4-mini-high responded in just 39 seconds and gave an answer 15km closer.
\n
Overall, Gemini was more cautious than ChatGPT, but Claude was the most cautious of all. Claude’s “extended thinking” mode made Sonnet even more conservative than the standard version. In some cases, the regular model would hazard a guess, albeit hedged in probabilistic terms, whereas with “extended thinking” enabled for the same test, it either declined to guess or offered only vague, region-level responses.
\n
LLMs Continue to Hallucinate
\n\n
All the models, at some point, returned answers that were entirely wrong. ChatGPT was typically more confident than Gemini, often leading to better answers, but also more hallucinations.
\n
The risk of hallucinations increased when the scenery was temporary or had changed over time. In one test, for instance, a beach photo showed a large hotel and a temporary ferris wheel (installed in 2024 and dismantled during winter). Many of the models consistently pointed to a different, more frequently photographed beach with a similar ride, despite clear differences.
\n
Final Tips
\n\n
Your account and prompt history may bias results. In one case, when analysing a photo taken in the Coral Pink Sand Dunes State Park, Utah, ChatGPT o4-mini referenced previous conversations with the account holder: “The user mentioned Durango and Colorado earlier, so I suspect they might have posted a photo from a previous trip.”
\n
Similarly, Grok appeared to draw on a user’s Twitter profile, and past tweets, even without explicit prompts to do so.
\n
Video comprehension also remains limited. Most LLMs cannot search for or watch video content, cutting off a rich source of location data. They also struggle with coordinates, often returning rough or simply incorrect responses.
\n
Ultimately, LLMs are no silver bullet. They still hallucinate, and when a photo lacks detail, geolocating it will still be difficult. That said, unlike our controlled tests, real-world investigations typically involve additional context. While Google Lens accepts only keywords, LLMs can be supplied with far richer information, making them more adaptable.
\n
There is little doubt, at the rate they are evolving, LLMs will continue to play an increasingly significant role in open source research. And as newer models emerge, we will continue to test them.
\n
Infographics by Logan Williams and Merel Zoet
\n
",
+ "content_text": "An ambiguous city street, a freshly mown field, and a parked armoured vehicle were among the example photos we chose to challenge Large Language Models (LLMs) from OpenAI, Google, Anthropic, Mistral and xAI to geolocate.\nBack in\nJuly 2023\n, Bellingcat analysed the geolocation performance of OpenAI and Google’s models. Both chatbots struggled to identify images and were highly prone to hallucinations. However, since then, such models have rapidly evolved.\nTo assess how LLMs from OpenAI, Google, Anthropic, Mistral and xAI compare today, we ran 500 geolocation tests, with 20 models each analysing the same set of 25 images.\nWe chose 25 of our own travel photos, varying in difficulty to geolocate, none of which had been published online before.\nOur analysis included older and “deep research” versions of the models, to track how their geolocation capabilities have developed over time. We also included Google Lens to compare whether LLMs offer a genuine improvement over traditional reverse image search. While reverse image search tools work differently from LLMs, they remain one of the most effective ways to narrow down an image’s location when starting from scratch.\nThe Test\nWe used 25 of our own travel photos, to test a range of outdoor scenes, both rural and urban areas, with and without identifiable landmarks such as buildings, mountains, signs or roads. These images were sourced from every continent, including Antarctica.\nThe vast majority have not been reproduced here, as we intend to continue using them to evaluate newer models as they are released. Publishing them here would compromise the integrity of future tests.\nEach LLM was given a photo that had not been published online and contained no metadata. All models then received the same prompt: “Where was this photo taken?”, alongside the image. If an LLM asked for more information, the response was identical: “There is no supporting information. Use this photo alone.”\nWe tested the following models:\nDeveloper\nModel\nDeveloper’s Description\nAnthropic\nClaude Haiku 3.5\n“fastest model for daily tasks”\nClaude Sonnet 3.7\n“our most intelligent model yet”\nClaude Sonnet 3.7 (extended thinking)\n“enhanced reasoning capabilities for complex tasks”\nClaude Sonnet 4.0\n“smart, efficient model for everyday use”\nClaude Opus 4.0\n“powerful, large model for complex challenges”\nGoogle\nGemini 2.0 Flash\n“for everyday tasks plus more features”\nGemini 2.5 Flash\n“uses advanced reasoning”\nGemini 2.5 Pro\n“best for complex tasks”\nGemini Deep Research\n“get in-depth answers”\nMistral\nPixtral Large\n“frontier-level image understanding”\nOpenAI\nChatGPT 4o\n“great for most tasks”\nChatGPT Deep Research\n“designed to perform in-depth, multi-step research using data on the public web”\nChatGPT 4.5\n“good for writing and exploring ideas”\nChatGPT o3\n“uses advanced reasoning”\nChatGPT o4-mini\n“fastest at advanced reasoning”\nChatGPT o4-mini-high\n“great at coding and visual reasoning”\nxAI\nGrok 3\n“smartest”\nGrok 3 DeepSearch\n“advanced search and reasoning”\nGrok 3 DeeperSearch\n“extended search, more reasoning”\nThis was not a comprehensive review of all available models, partly due to the speed at which new models and versions are currently being released. For example, we did not assess DeepSeek, as it currently only extracts text from images. Note that in ChatGPT, regardless of what model you select, the “deep research” function is currently powered by a version of\no4-mini\n.\nSupport Bellingcat\nYour donations directly contribute to our ability to publish groundbreaking investigations and uncover wrongdoing around the world.\nDonate\nGemini models have been released in “preview” and “experimental” formats, as well as dated versions like “03-25” and “05-06”. To keep the comparisons manageable, we grouped these variants under their respective base models, e.g. “Gemini 2.5 Pro”.\nWe also compared every test with the first 10 results from Google Lens’s “visual match” feature, to assess the difficulty of the tests and the usefulness of LLMs in solving them.\nWe ranked all responses on a scale from 0 to 10, with 10 indicating an accurate and specific identification, such as a neighbourhood, trail, or landmark, and 0 indicating no attempt to identify the location at all.\nAnd the Winner is…\nChatGPT beat Google Lens.\nIn our tests, ChatGPT o3, o4-mini, and o4-mini-high were the only models to outperform Google Lens in identifying the correct location, though not by a large margin. All other models were less effective when it came to geolocating our test photos.\nWe scored 20 models against 25 photos, rating each from 0 (red) to 10 (dark green) for accuracy in geolocating the images.\nEven Google’s own LLM, Gemini, fared worse than Google Lens. Surprisingly, it also scored lower than xAI’s Grok, despite Grok’s\nwell-documented tendency to hallucinate.\nGemini’s Deep Research mode scored roughly the same as the three Grok models we tested, with DeeperSearch proving the most effective of xAI’s LLMs.\nThe highest-scoring models from Anthropic and Mistral lagged well behind their current competitors from OpenAI, Google, and xAI. In several cases, even Claude’s most advanced models identified only the continent, while others were able to narrow their responses down to specific parts of a city. The latest Claude model, Opus 4, performed at a similar level to Gemini 2.5 Pro.\nHere are some of the highlights from five of our tests.\nA Road in the Japanese Mountains\nThe photo below was taken on the road between Takayama and Shirakawa in Japan. As well as the road and mountains, signs and buildings are also visible.\nTest “snowy-highway” depicted a road near Takayama, Japan.\nGemini 2.5 Pro’s response was not useful. It mentioned Japan, but also Europe, North and South America and Asia. It replied:\n“Without any clear, identifiable landmarks, distinctive signage in a recognisable language, or unique architectural styles, it’s very difficult to determine the exact country or specific location.”\nIn contrast, o3 identified both the architectural style and signage, responding:\n“Best guess: a snowy mountain stretch of central-Honshu, Japan—somewhere in the Nagano/Toyama area. (Japanese-style houses, kanji on the billboard, and typical expressway barriers give it away.)”\nA Field on the Swiss Plateau\nThis photo was taken near Zurich. It showed no easily recognisable features apart from the mountains in the distance. A reverse image search using Google Lens didn’t immediately lead to Zurich. Without any context, identifying the location of this photo manually could take some time. So how did the LLMs fare?\nTest “field-hills” depicted a view of a field near Zurich\nGemini 2.5 Pro stated that the photo showed scenery common to many parts of the world and that it couldn’t narrow it down without additional context.\nBy contrast, ChatGPT excelled at this test. o4-mini identified the “Jura foothills in northern Switzerland”, while o4-mini-high placed the scene ”between Zürich and the Jura mountains”.\nThese answers stood in stark contrast to those from Grok Deep Research, which, despite the visible mountains, confidently stated the photo was taken in the Netherlands. This conclusion appeared to be based on the Dutch name of the account used, “Foeke Postma”, with the model assuming the photo must have been taken there and calling it a “reasonable and well-supported inference”.\nAn Inner-City Alley Full of Visual Clues in Singapore\nThis photo of a narrow alleyway on Circular Road in Singapore provoked a wide range of responses from the LLMs and Google Lens, with scores ranging from 3 (nearby country) to 10 (correct location).\nTest “dark-alley”, a photo taken of an alleyway in Singapore\nThe test served as a good example of how LLMs can outperform Google Lens by focusing on small details in a photo to identify the exact location. Those that answered correctly referenced the writing on the mailbox on the left in the foreground, which revealed the precise address.\nWhile Google Lens returned results from all over Singapore and Malaysia, part of ChatGPT o4-mini’s response read: “This appears to be a classic Singapore shophouse arcade – in fact, if you look at the mailboxes on the left you can just make out the label ‘[correct address].’”\nSome of the other models noticed the mailbox but could not read the address visible in the image, falsely inferring that it pointed to other locations. Gemini 2.5 Flash responded, “The design of the mailboxes on the left, particularly the ‘G’ for Geylang, points strongly towards Singapore.” Another Gemini model, 2.5 Pro, spotted the mailbox but focused instead on what it interpreted as Thai script on a storefront, confidently answering: “The visual evidence strongly suggests the photo was taken in an alleyway in Thailand, likely in Bangkok.”\nThe Costa Rican Coast\nOne of the harder tests we gave the models to geolocate was a photo taken from Playa Longosta on the Pacific Coast of Costa Rica near Tamarindo.\nTest “beach-forest” showed Playa Longosta, Costa Rica.\nGemini and Claude performed the worst on this task, with most models either declining to guess or giving incorrect answers. Claude 3.7 Sonnet correctly identified Costa Rica but hedged with other locations, such as Southeast Asia. Grok was the only model to guess the exact location correctly, while several ChatGPT models (Deep Research, o3 and the o4-minis) guessed within 160km of the beach.\nAn Armoured Vehicle on the Streets of Beirut\nThis photo was taken on the streets of Beirut and features several details useful for geolocation, including an emblem on the side of the armored personnel carrier and a partially visible Lebanese flag in the background.\nTest “street-military” depicted an armoured personnel carrier on the streets of Beirut\nSurprisingly, most models struggled with this test: Claude 4 Opus, billed as a “powerful, large model for complex challenges”, guessed “somewhere in Europe” owing to the “European-style street furniture and building design”, while Gemini and Grok could only narrow the location down to Lebanon. Half of the ChatGPT models responded with Beirut. Only two models, both ChatGPT, referenced the flag.\nSo Have LLMs Finally Mastered Geolocation?\nLLMs can certainly help researchers to spot the details that Google Lens or they themselves might miss.\nOne clear advantage of LLMs is their ability to search in multiple languages. They also\nappear to make good use of small clues, such as vegetation, architectural styles or signage. In one test, a photo of a man wearing a life vest in front of a mountain range was correctly located because the model identified part of a company name on his vest and linked it to a nearby boat tour operator.\nFor touristic areas and scenic landscapes, Google Lens still outperformed most models. When shown a photo of Schluchsee lake in the Black Forest, Germany, Google Lens returned it as the top result, while ChatGPT was the only LLM to correctly identify the lake’s name. In contrast, in urban settings, LLMs excelled at cross-referencing subtle details, whereas Google Lens tended to fixate on larger, similar-looking structures, such as buildings or ferris wheels, which appear in many other locations.\nHeat map to show how each model performed on all 25 tests\nEnhanced Reasoning Modes\nYou’d assume turning on “deep research” or “extended thinking” functions would have resulted in higher scores. However, on average, Claude and ChatGPT performed worse. Only one Grok model, DeeperSearch, and one Gemini, Gemini Deep Research, showed improvement. For example, ChatGPT Deep Research was shown a photo of a coastline and took nearly 13 minutes to produce an answer that was about 50km north of the correct location. Meanwhile, o4-mini-high responded in just 39 seconds and gave an answer 15km closer.\nOverall, Gemini was more cautious than ChatGPT, but Claude was the most cautious of all. Claude’s “extended thinking” mode made Sonnet even more conservative than the standard version. In some cases, the regular model would hazard a guess, albeit hedged in probabilistic terms, whereas with “extended thinking” enabled for the same test, it either declined to guess or offered only vague, region-level responses.\nLLMs Continue to Hallucinate\nAll the models, at some point, returned answers that were entirely wrong. ChatGPT was typically more confident than Gemini, often leading to better answers, but also more hallucinations.\nThe risk of hallucinations increased when the scenery was temporary or had changed over time. In one test, for instance, a beach photo showed a large hotel and a temporary ferris wheel (installed in 2024 and dismantled during winter). Many of the models consistently pointed to a different, more frequently photographed beach with a similar ride, despite clear differences.\nFinal Tips\nYour account and prompt history may bias results. In one case, when analysing a photo taken in the Coral Pink Sand Dunes State Park, Utah, ChatGPT o4-mini referenced previous conversations with the account holder: “The user mentioned Durango and Colorado earlier, so I suspect they might have posted a photo from a previous trip.”\nSimilarly, Grok appeared to draw on a user’s Twitter profile, and past tweets, even without explicit prompts to do so.\nVideo comprehension also remains limited. Most LLMs cannot search for or watch video content, cutting off a rich source of location data. They also struggle with coordinates, often returning rough or simply incorrect responses.\nUltimately, LLMs are no silver bullet. They still hallucinate, and when a photo lacks detail, geolocating it will still be difficult. That said, unlike our controlled tests, real-world investigations typically involve additional context. While Google Lens accepts only keywords, LLMs can be supplied with far richer information, making them more adaptable.\nThere is little doubt, at the rate they are evolving, LLMs will continue to play an increasingly significant role in open source research. And as newer models emerge, we will continue to test them.\nInfographics by Logan Williams and Merel Zoet",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/06/Untitled-design-1-1200x960.jpg",
+ "alt": "",
+ "caption": "We chose 25 of our own travel photos, varying in difficulty to geolocate, none of which had been published online before."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/05/question-mark.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://public.flourish.studio/visualisation/23466601/thumbnail",
+ "alt": "chart visualization",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/06/snowy-highway-Copy-1200x900.jpg",
+ "alt": "",
+ "caption": "Test “snowy-highway” depicted a road near Takayama, Japan."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/06/field-hills-Copy-1200x904.jpg",
+ "alt": "",
+ "caption": "Test “field-hills” depicted a view of a field near Zurich"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/06/alley-903x1200.jpg",
+ "alt": "",
+ "caption": "Test “dark-alley”, a photo taken of an alleyway in Singapore"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/06/beach-forest-1200x900.jpg",
+ "alt": "",
+ "caption": "Test “beach-forest” showed Playa Longosta, Costa Rica."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/06/apc.jpg",
+ "alt": "",
+ "caption": "Test “street-military” depicted an armoured personnel carrier on the streets of Beirut"
+ },
+ {
+ "url": "https://public.flourish.studio/visualisation/23470918/thumbnail",
+ "alt": "chart visualization",
+ "caption": ""
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2025-08-14_llms-vs-geolocation-gpt-5-performs-worse-than-other-ai-models.json b/bellingcat_geolocation_json/2025-08-14_llms-vs-geolocation-gpt-5-performs-worse-than-other-ai-models.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e7a651b921062ec3ce0f216b7357f789710fe02
--- /dev/null
+++ b/bellingcat_geolocation_json/2025-08-14_llms-vs-geolocation-gpt-5-performs-worse-than-other-ai-models.json
@@ -0,0 +1,42 @@
+{
+ "slug": "llms-vs-geolocation-gpt-5-performs-worse-than-other-ai-models",
+ "title": "LLMs Vs. Geolocation: GPT-5 Performs Worse Than Other AI Models",
+ "url": "https://www.bellingcat.com/resources/2025/08/14/llms-vs-geolocation-gpt-5-performs-worse-than-other-ai-models/",
+ "date": "2025-08-14T10:42:44+00:00",
+ "modified": "2025-11-25T10:45:18+00:00",
+ "author": "Foeke Postma",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Bellingcat challenged the latest models—Grok, GPT and Gemini—on geolocation tasks, revealing GPT-5 as a clear downgrade.",
+ "content_html": "
\n
In June, Bellingcat ran 500 geolocation tests, comparing LLMs from various companies against each other, as well as Google Lens – a staple tool for finding the location of photos.
\n
At the time, ChatGPT o4-mini-high emerged as the clear winner, with Google Lens outperforming most other models. Just two months later, with new versions of these AI tools available, we re-ran the trial – this time including Google “AI Mode,” GPT-5, GPT-5 Thinking, and Grok 4 into the mix.
\n
These five photos were excluded from our most recent trial as they were published in our previous article.\n
The original test used 25 of Bellingcat’s own holiday photos. From cities to remote countryside, the images included scenes both with and without recognisable features – such as roads, signage, mountains, or architecture. Images were sourced from every continent.
\n
For the updated trial, five test photos were excluded, as they had appeared in a previous article, thus compromising the integrity of the results.
\n
All 24 models’ responses were ranked on a scale from 0 to 10, with 10 indicating an accurate and specific identification (such as a neighbourhood, trail, or landmark) and 0 indicating no attempt to identify the location at all.
\n
\n\n
Google AI Mode was shown to be the most capable geolocation tool overall.
\n
Grok 4 gave both better and worse answers compared to Grok 3 but, on average, scored marginally higher. However, it was still less accurate than older versions of Gemini and GPT.
\n
GPT-5, even in ‘Thinking’ and ‘Pro’ modes, was a considerable downgrade when compared with the capabilities demonstrated by GPT o4-mini-high. In one example, of a city street with skyscrapers in the background, o4-mini-high correctly identified the street, while GPT-5 in Thinking mode pointed to the wrong country.
\n
\n
\n
Support Bellingcat
\n
Your donations directly contribute to our ability to publish groundbreaking investigations and uncover wrongdoing around the world.
Despite delivering faster answers, GPT-5 appeared to sacrifice accuracy. A surprising number of errors and a general sense of disappointment in the new model have also been reported by other users.
\n
Bellingcat tested GPT-5 and its ‘Thinking’ mode via the Plus subscription, which costs roughly the same as access to 04-mini-high prior to its retirement. Five of the most difficult test images were also run through GPT-5 Pro. But even Pro, with a premium price tag of €200 per month, failed to geolocate the photos any more accurately than GPT 04-mini-high.
\n
A Beach, a Hotel and a Ferris Wheel
\n
The disparity between Google and the GPT models became even more apparent in Test 25 – a photo of a shoreline hotel in Noordwijk, the Netherlands, with a Ferris wheel rising just beyond the dunes.
\n
\n
Test 25: A photo of Noordwijk beach in the Netherlands. Credit: Bellingcat.
\n
In the previous trial, most older models – including those from GPT, Claude, Gemini and Grok – accurately identified the country as the Netherlands but failed to locate the town. Many latched onto the Ferris wheel but pointed instead to the seaside town of Scheveningen, which also has a Ferris wheel, though situated on a pier, not among the sand dunes.
\n
However, the most recent models, GPT-5 Pro and Thinking, were even less accurate, identifying a beach in France – an entirely different country.
\n
Unfortunately for open source researchers, following the release of GPT-5, OpenAI removed the option to select older models such as o4-mini-high. After a wave of negative feedback, OpenAI reinstated GPT-4o as the default model for paid subscribers. However, the most capable geolocation models identified in Bellingcat’s testing remain inaccessible.
\n
Google AI Mode, on the other hand, was the first, and only model so far, to correctly identify Noordwijk as the location in Test 25.
\n
Though AI Mode is powered by a version of Gemini 2.5, it outperformed Gemini 2.5 Pro Deep Research in these tests. Described by Google as its “most powerful AI search, with more advanced reasoning and multimodality,” AI Mode geolocated test images with greater accuracy than any GPT models, including our previous winner, o4-mini-high.
The majority of models, at some point, returned a hallucination. Users should not rely solely on the answers provided by LLMs. Even the best options, including Google AI Mode, still, at times, confidently point to the wrong location.
\n
The difference in models’ capabilities compared with just two months ago shows how quickly this field is evolving. However, OpenAI’s recent changes also suggest that progress is not guaranteed, and that AI’s ability to geolocate may plateau or even worsen over time. As new models emerge, Bellingcat will continue to test them.
\n
Thanks to Nathan Patin for contributing to the original benchmark tests.
\n\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Bluesky here and Instagram here.
\n
",
+ "content_text": "In June,\nBellingcat ran 500 geolocation tests\n, comparing LLMs from various companies against each other, as well as Google Lens – a staple tool for finding the location of photos.\nAt the time, ChatGPT o4-mini-high emerged as the clear winner, with Google Lens outperforming most other models. Just two months later, with new versions of these AI tools available, we re-ran the trial – this time including Google “AI Mode,” GPT-5, GPT-5 Thinking, and Grok 4 into the mix.\nThese five photos were excluded from our most recent trial as they were published in our\nprevious article\n.\nThe\noriginal test\nused 25 of Bellingcat’s own holiday photos. From cities to remote countryside, the images included scenes both with and without recognisable features – such as roads, signage, mountains, or architecture. Images were sourced from every continent.\nFor the updated trial, five test photos were excluded, as they had appeared in a previous article, thus compromising the integrity of the results.\nAll 24 models’ responses were ranked on a scale from 0 to 10, with 10 indicating an accurate and specific identification (such as a neighbourhood, trail, or landmark) and 0 indicating no attempt to identify the location at all.\nGoogle AI Mode was shown to be the most capable geolocation tool overall.\nGrok 4 gave both better and worse answers compared to Grok 3 but, on average, scored marginally higher. However, it was still less accurate than older versions of Gemini and GPT.\nGPT-5, even in ‘Thinking’ and ‘Pro’ modes, was a considerable downgrade when compared with the capabilities demonstrated by GPT o4-mini-high. In one example, of a city street with skyscrapers in the background, o4-mini-high correctly identified the street, while GPT-5 in Thinking mode pointed to the wrong country.\nSupport Bellingcat\nYour donations directly contribute to our ability to publish groundbreaking investigations and uncover wrongdoing around the world.\nDonate\nDespite delivering faster answers, GPT-5 appeared to sacrifice accuracy. A surprising number of errors and a general sense of disappointment in the new model have also been\nreported by other users\n.\nBellingcat tested GPT-5 and its ‘Thinking’ mode via the Plus subscription, which costs roughly the same as access to 04-mini-high prior to its retirement. Five of the most difficult test images were also run through GPT-5 Pro. But even Pro, with a premium price tag of €200 per month, failed to geolocate the photos any more accurately than GPT 04-mini-high.\nA Beach, a Hotel and a Ferris Wheel\nThe disparity between Google and the GPT models became even more apparent in Test 25 – a photo of a shoreline hotel in Noordwijk, the Netherlands, with a Ferris wheel rising just beyond the dunes.\nTest 25: A photo of Noordwijk beach in the Netherlands. Credit: Bellingcat\n.\nIn the previous trial, most older models – including those from GPT, Claude, Gemini and Grok – accurately identified the country as the Netherlands but failed to locate the town. Many latched onto the Ferris wheel but pointed instead to the seaside town of Scheveningen, which also has a Ferris wheel, though situated on a pier, not among the sand dunes.\nHowever, the most recent models, GPT-5 Pro and Thinking, were even less accurate, identifying a beach in France – an entirely different country.\nUnfortunately for open source researchers, following the release of GPT-5, OpenAI removed the option to select older models such as o4-mini-high. After a wave of negative feedback, OpenAI reinstated GPT-4o as the default model for paid subscribers. However, the most capable geolocation models identified in Bellingcat’s testing remain inaccessible.\nGoogle AI Mode, on the other hand, was the first, and only model so far, to correctly identify Noordwijk as the location in Test 25.\nThough AI Mode is powered by a version of Gemini 2.5, it outperformed Gemini 2.5 Pro Deep Research in these tests.\nDescribed by Google\nas its “most powerful AI search, with more advanced reasoning and multimodality,” AI Mode geolocated test images with greater accuracy than any GPT models, including our previous winner, o4-mini-high.\nAI Mode is currently only available\nin India, United Kingdom and the United States.\nCredit: Google\n.\nThe majority of models, at some point, returned a hallucination. Users should not rely solely on the answers provided by LLMs. Even the best options, including Google AI Mode, still, at times, confidently point to the wrong location.\nThe difference in models’ capabilities compared with just two months ago shows how quickly this field is evolving. However, OpenAI’s recent changes also suggest that progress is not guaranteed, and that AI’s ability to geolocate may plateau or even worsen over time. As new models emerge, Bellingcat will continue to test them.\nThanks to Nathan Patin for contributing to the original benchmark tests.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Bluesky\nhere\nand Instagram\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/08/Untitled-design-1-1200x675.jpg",
+ "alt": "",
+ "caption": "These five photos were excluded from our most recent trial as they were published in our previous article ."
+ },
+ {
+ "url": "https://public.flourish.studio/visualisation/24646406/thumbnail",
+ "alt": "chart visualization",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/05/question-mark.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/08/test25-1200x904.jpg",
+ "alt": "",
+ "caption": "Test 25: A photo of Noordwijk beach in the Netherlands. Credit: Bellingcat ."
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/08/Screenshot-2025-08-14-at-11.05.24.jpg",
+ "alt": "",
+ "caption": "Credit: Google ."
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/2025-11-05_geolocating-darfur-killings-of-those-escaping-al-fashir.json b/bellingcat_geolocation_json/2025-11-05_geolocating-darfur-killings-of-those-escaping-al-fashir.json
new file mode 100644
index 0000000000000000000000000000000000000000..2ec8f26ee4d6fe39c289fa88aa91b3d1b3420ee6
--- /dev/null
+++ b/bellingcat_geolocation_json/2025-11-05_geolocating-darfur-killings-of-those-escaping-al-fashir.json
@@ -0,0 +1,47 @@
+{
+ "slug": "geolocating-darfur-killings-of-those-escaping-al-fashir",
+ "title": "Geolocating Darfur Killings of Those Escaping Al Fashir",
+ "url": "https://www.bellingcat.com/news/2025/11/05/geolocating-darfur-killings-of-those-escaping-al-fashir/",
+ "date": "2025-11-05T13:35:27+00:00",
+ "modified": "2026-02-11T09:58:19+00:00",
+ "author": "Sebastian Vandermeersch, Merel Zoet",
+ "source": "Bellingcat",
+ "task_type": "geolocation_reasoning",
+ "categories": [],
+ "tags": [],
+ "excerpt": "Warning: Contains written descriptions of extreme violence including murder. All graphic footage has been omitted. A Bellingcat investigation has identified and verified the exact location of a mass killing of more than 80 people that took place near Al Fashir during the recent siege and takeover of the Sudanese city by the paramilitary Rapid Support […]",
+ "content_html": "
\n
Warning: Contains written descriptions of extreme violence including murder. All graphic footage has been omitted.
\n
A Bellingcat investigation has identified and verified the exact location of a mass killing of more than 80 people that took place near Al Fashir during the recent siege and takeover of the Sudanese city by the paramilitary Rapid Support Forces (RSF). More than a dozen videos which have emerged online since Oct. 27 were analysed, some filmed by fighters wearing what appears to be RSF insignia, showing the events and their aftermath.
\n
\n
RSF fighter filming himself at the scene with visible shoulder-patch insignia. Bellingcat blurred the face and bodies in the background. Credit: X/Himma0099
\n
Al Fashir, the capital of North Darfur, was the last major stronghold of the Sudanese military in the region and a refuge for people fleeing nearby villages and towns. The city, considered a strategic military location near the southwestern border leading to Chad and South Sudan, fell last week after more than 500 days under RSF siege. The RSF, which has been fighting Sudan’s military for control over the country, killed at least 1,500 people during the takeover, according to the Sudan Doctors Network.
\n
A Mass Killing Outside the City
\n
Open source evidence reviewed by Bellingcat showed that people attempting to flee Al Fashir were killed by men, some of whom appeared to be wearing RSF uniforms. The RSF had earlier pledged to remain “firmly committed to protect the civilians” and “not tolerate any attempts to exploit their lives or tamper with their security.”
\n
Footage posted on social media showed a mass killing more than seven kilometres outside the city, where the fleeing convoy had encountered an earthen berm encircling Al Fashir. The berm and an accompanying trench were built by the RSF during the siege. Satellite imagery indicates that the section to the north-west of the city, where the mass killing took place, was constructed throughout August.
\n
\n
Satellite imagery of Oct. 27, showing the berm and trench dug around Al Fashir, checkpoints for roads leaving the city and the location of the mass killing. Satellite image credit: Planet Labs PBC
\n
Bellingcat analysed more than a dozen videos. Given the extreme violence visible in much of the footage, only material that was not explicitly graphic has been included in this report. Some of the videos filmed at this location have also been geolocated by the Centre for Information Resilience and others.
\n
Cross-referencing the different camera angles in the videos made it possible to map the exact location of the killings to a single location: 13.71340, 25.27854. In satellite imagery, clusters of dark shapes around the vehicles north of the berm correspond to the locations of bodies identified in verified videos from the scene.
\n
\n
Satellite imagery showing the convoy of burned vehicles and clusters of bodies by the berm, with corresponding screengrabs of videos geolocated to locations around the site. Satellite image credit: Vantor
\n
In total, the video footage showed more than 80 bodies lying beside damaged and burning vehicles along the berm, inside the adjacent trench and across the surrounding area. More than two dozen destroyed vehicles were also identified.
\n
\n
\n
Support Bellingcat
\n
Your donations directly contribute to our ability to publish groundbreaking investigations and uncover wrongdoing around the world.
In one video, filmed early on in the attack, fighters in what appear to be RSF uniforms are seen taking cover behind a vehicle as gunfire is heard. One fighter can be seen closing in on the burning convoy and aiming his weapon.
\n
However, in later videos filmed by the fighters showing the aftermath of the attack, none of the casualties visible appeared to be in possession of weapons; almost all appeared to be dressed in civilian clothing and several were women. Of the more than 80 casualties visible in various videos, a couple can be seen wearing what appears to be dark, military-style clothing, and one can be seen lying next to a belt of ammunition. Although, it is unclear who the belt belonged to.
\n
The arrangement of destroyed vehicles and groups of bodies in the videos matched what was visible in satellite imagery. Moreover, tyre tracks visible in the sand appeared to link the mass killing site with the checkpoints along the sand berm. Fresh marks appeared on satellite images after Oct. 27., the day of the RSF’s complete takeover of Al Fashir.
\n
\n
Panorama of video footage taken at the site. The burnt vehicles in the background could also be identified on satellite imagery. Video: x.com/hodajannat24
\n
A RSF commander known as Abu Lulu, identified in the videos by his distinctive white scarf, appeared in at least three videos from the scene. The commander has previously been featured in social media footage from the conflict. On Oct. 31, TikTok removed a profile believed to be linked to him.
\n
In one video, geolocated by Bellingcat to the far side of the berm, the commander can be seen and heard taunting a group of 10 male captives sitting lined up on the ground, forcing them to praise the RSF and denigrate the military before they are shot.
\n
\n
Screenshots of the video showing RSF commander Abu Lulu talking to the 10 male captives before they were shot. Distinct features in the background of the video allowed geolocation of the scene. Satellite image credit: Vantor Video credit: X/Sudan_tweet
\n
In another video filmed to the west of the convoy, Abu Lulu can be heard questioning a wounded man lying on the ground, threatening sexual violence before shooting him at point-blank range. In a third video filmed meters away, the same commander can be seen standing among a number of motionless bodies as he repeatedly fires into them.
\n
In several other videos, one fleeing vehicle appears to have crossed the berm before becoming stuck in the trench on the far side. In one of these videos, filmed from inside the trench, a fighter can be seen firing into a group of motionless bodies on the ground next to the car, while others film and celebrate.
\n
UN Investigation Points to a Pattern
\n
Morris Tidball-Binz, the UN Special Rapporteur on extrajudicial, summary or arbitrary executions, told Bellingcat that “if the images were genuine”, these were “unlawful killings under International Human Rights Law, International Humanitarian Law (IHL) and International Criminal Law.”
\n
“Under the former they correspond to extrajudicial or arbitrary executions; under IHL they correspond to war crimes and under the latter they may amount to crimes against humanity in view of the number of victims and the seemingly planned and systematic nature of the killings,” he said.
\n
\n
Subscribe to the Bellingcat newsletter
\n
Subscribe to our newsletter for first access to our published content and events that our staff and contributors are involved with, including interviews and training workshops.
\n\n
\n
Tidball-Binz added that he was not aware of any steps taken by RSF command or relevant authorities to investigate these incidents or hold those responsible accountable. “So far impunity has prevailed,” he stated via email.
\n
Footage from Oct. 30. shows Abu Lulu in RSF custody, being transported under heavy guard and later incarcerated in Shala Prison on the western edge of the city. The RSF announced an investigation into “violations” committed by fighters following the capture of Al Fashir.
\n
Other open source data also documented evidence of atrocities within the city itself. In a chapter of a detailed report entitled “evidence of mass killings”, satellite imagery analysis by Yale’s Humanitarian Research Lab from Oct. 27 found what it said were “objects consistent with the size of human bodies on the ground near RSF vehicles” and “instances of reddish earth discoloration”.
\n
The UN Security Council condemned the assault by the RSF on Al Fashir and “its devastating impact on the civilian population.”
\n
The UN Fact-Finding Mission on Sudan expressed “grave alarm at escalating atrocities” detailing “a devastating pattern of atrocities by both warring parties” that include “war crimes of violence to life and person”. The mission’s initial findings pointed to “a deliberate pattern of ethnically targeted executions of unarmed civilians, assaults, sexual violence, widespread looting and destruction of vital infrastructure, and mass forced displacement”.
Carlos Gonzales, Jake Godin and Logan Williams contributed to this report.
\n
Bellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so here. You can also subscribe to our Patreon channel here. Subscribe to our Newsletter and follow us on Bluesky here and Mastodon here.
\n
",
+ "content_text": "Warning: Contains written descriptions of extreme violence including murder. All graphic footage has been omitted.\nA Bellingcat investigation has identified and verified the exact location of a mass killing of more than 80 people that took place near Al Fashir during the recent siege and takeover of the Sudanese city by the paramilitary Rapid Support Forces (RSF). More than a dozen videos which have emerged online since Oct. 27 were analysed, some filmed by fighters wearing what appears to be RSF insignia, showing the events and their aftermath.\nRSF fighter filming himself at the scene with visible shoulder-patch insignia. Bellingcat blurred the face and bodies in the background. Credit: X/Himma0099\nAl Fashir, the capital of North Darfur, was the last major stronghold of the Sudanese military in the region and a refuge for people fleeing nearby villages and towns. The city, considered a strategic military location near the southwestern border leading to Chad and South Sudan, fell last week after more than\n500 days\nunder RSF siege. The RSF, which has been fighting Sudan’s military for control over the country, killed at least 1,500 people during the takeover, according to the\nSudan Doctors Network\n.\nA Mass Killing Outside the City\nOpen source evidence reviewed by Bellingcat showed that people attempting to flee Al Fashir were killed by men, some of whom appeared to be wearing RSF uniforms. The RSF had earlier\npledged\nto remain “firmly committed to protect the civilians” and “not tolerate any attempts to exploit their lives or tamper with their security.”\nFootage posted on social media showed a mass killing more than seven kilometres outside the city, where the fleeing convoy had encountered an earthen berm encircling Al Fashir. The berm and an accompanying trench were\nbuilt by\nthe RSF during the siege. Satellite imagery indicates that the section to the north-west of the city, where the mass killing took place, was constructed throughout August.\nSatellite imagery of Oct. 27, showing the berm and trench dug around Al Fashir, checkpoints for roads leaving the city and the location of the mass killing. Satellite image credit: Planet Labs PBC\nBellingcat analysed more than a dozen videos. Given the extreme violence visible in much of the footage, only material that was not explicitly graphic has been included in this report. Some of the videos filmed at this location have also been geolocated by the\nCentre for Information Resilience\nand others.\nCross-referencing the different camera angles in the videos made it possible to map the exact location of the killings to a single location:\n13.71340, 25.27854\n. In satellite imagery, clusters of dark shapes around the vehicles north of the berm correspond to the locations of bodies identified in verified videos from the scene.\nSatellite imagery showing the convoy of burned vehicles and clusters of bodies by the berm, with corresponding screengrabs of videos geolocated to locations around the site. Satellite image credit: Vantor\nIn total, the video footage showed more than 80 bodies lying beside damaged and burning vehicles along the berm, inside the adjacent trench and across the surrounding area. More than two dozen destroyed vehicles were also identified.\nSupport Bellingcat\nYour donations directly contribute to our ability to publish groundbreaking investigations and uncover wrongdoing around the world.\nDonate\nIn\none video\n, filmed early on in the attack, fighters in what appear to be RSF uniforms are seen taking cover behind a vehicle as gunfire is heard. One fighter can be seen closing in on the burning convoy and aiming his weapon.\nHowever, in later videos filmed by the fighters showing the aftermath of the attack, none of the casualties visible appeared to be in possession of weapons; almost all appeared to be dressed in civilian clothing and several were women. Of the more than 80 casualties visible in various videos, a couple can be seen wearing what appears to be dark, military-style clothing, and one can be seen lying next to a belt of ammunition. Although, it is unclear who the belt belonged to.\nThe arrangement of destroyed vehicles and groups of bodies in the videos matched what was visible in satellite imagery. Moreover, tyre tracks visible in the sand appeared to link the mass killing site with the checkpoints along the sand berm. Fresh marks appeared on satellite images after Oct. 27., the day of the RSF’s complete takeover of Al Fashir.\nPanorama of video footage taken at the site. The burnt vehicles in the background could also be identified on satellite imagery. Video: x.com/hodajannat24\nA RSF commander known as Abu Lulu, identified in the videos by his distinctive white scarf, appeared in at least three videos from the scene. The commander has previously been featured in social media footage from the conflict. On Oct. 31,\nTikTok removed\na profile believed to be linked to him.\nIn one video,\ngeolocated by Bellingcat\nto the far side of the berm, the commander can be seen and heard taunting a group of 10 male captives sitting lined up on the ground, forcing them to praise the RSF and denigrate the military before they are shot.\nScreenshots of the video showing RSF commander Abu Lulu talking to the 10 male captives before they were shot. Distinct features in the background of the video allowed geolocation of the scene. Satellite image credit: Vantor Video credit: X/Sudan_tweet\nIn another video filmed to the west of the convoy, Abu Lulu can be heard questioning a wounded man lying on the ground, threatening sexual violence before shooting him at point-blank range. In a third video filmed meters away, the same commander can be seen standing among a number of motionless bodies as he repeatedly fires into them.\nIn several other videos, one fleeing vehicle appears to have crossed the berm before becoming stuck in the trench on the far side. In one of these videos, filmed from inside the trench, a fighter can be seen firing into a group of motionless bodies on the ground next to the car, while others film and celebrate.\nUN Investigation Points to a Pattern\nMorris Tidball-Binz, the UN Special Rapporteur on extrajudicial, summary or arbitrary executions, told Bellingcat that “if the images were genuine”, these were “unlawful killings under International Human Rights Law, International Humanitarian Law (IHL) and International Criminal Law.”\n“Under the former they correspond to extrajudicial or arbitrary executions; under IHL they correspond to war crimes and under the latter they may amount to crimes against humanity in view of the number of victims and the seemingly planned and systematic nature of the killings,” he said.\nSubscribe to the Bellingcat newsletter\nSubscribe to our newsletter for first access to our published content and events that our staff and contributors are involved with, including interviews and training workshops.\nTidball-Binz added that he was not aware of any steps taken by RSF command or relevant authorities to investigate these incidents or hold those responsible accountable. “So far impunity has prevailed,” he stated via email.\nFootage\nfrom Oct. 30. shows Abu Lulu in RSF custody, being transported under heavy guard and later incarcerated in Shala Prison on the western edge of the city. The RSF\nannounced\nan investigation into “violations” committed by fighters following the capture of Al Fashir.\nOther open source data also documented evidence of atrocities within the city itself. In a chapter of a detailed report entitled “evidence of mass killings”, satellite imagery\nanalysis\nby Yale’s Humanitarian Research Lab from Oct. 27 found what it said were “objects consistent with the size of human bodies on the ground near RSF vehicles” and “instances of reddish earth discoloration”.\nThe UN Security Council\ncondemned\nthe assault by the RSF on Al Fashir and “its devastating impact on the civilian population.”\nThe UN Fact-Finding Mission on Sudan\nexpressed\n“grave alarm at escalating atrocities” detailing “a devastating pattern of atrocities by both warring parties” that include “war crimes of violence to life and person”. The mission’s initial findings pointed to “a deliberate pattern of ethnically targeted executions of unarmed civilians, assaults, sexual violence, widespread looting and destruction of vital infrastructure, and mass forced displacement”.\nThe RSF\nnow controls\nthe entire Darfur region.\nCarlos Gonzales, Jake Godin and Logan Williams contributed to this report.\nBellingcat is a non-profit and the ability to carry out our work is dependent on the kind support of individual donors. If you would like to support our work, you can do so\nhere\n. You can also subscribe to our Patreon channel\nhere\n. Subscribe to our\nNewsletter\nand follow us on Bluesky\nhere\nand Mastodon\nhere\n.",
+ "images": [
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/11/RSF-shoulder-patch.jpg",
+ "alt": "",
+ "caption": "RSF fighter filming himself at the scene with visible shoulder-patch insignia. Bellingcat blurred the face and bodies in the background. Credit: X/Himma0099"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/11/Map.jpg",
+ "alt": "",
+ "caption": "Satellite imagery of Oct. 27, showing the berm and trench dug around Al Fashir, checkpoints for roads leaving the city and the location of the mass killing. Satellite image credit: Planet Labs PBC"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/11/Map-with-images.jpg",
+ "alt": "",
+ "caption": "Satellite imagery showing the convoy of burned vehicles and clusters of bodies by the berm, with corresponding screengrabs of videos geolocated to locations around the site. Satellite image credit: Vantor"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/05/question-mark.png",
+ "alt": "",
+ "caption": ""
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/11/Panorama-1200x338.jpg",
+ "alt": "",
+ "caption": "Panorama of video footage taken at the site. The burnt vehicles in the background could also be identified on satellite imagery. Video: x.com/hodajannat24"
+ },
+ {
+ "url": "https://www.bellingcat.com/app/uploads/2025/11/Abu-Lulu.jpg",
+ "alt": "",
+ "caption": "Screenshots of the video showing RSF commander Abu Lulu talking to the 10 male captives before they were shot. Distinct features in the background of the video allowed geolocation of the scene. Satellite image credit: Vantor Video credit: X/Sudan_tweet"
+ }
+ ]
+}
diff --git a/bellingcat_geolocation_json/index.json b/bellingcat_geolocation_json/index.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad069d3565f81dbaf90d50c027cc3bed4c5f92c3
--- /dev/null
+++ b/bellingcat_geolocation_json/index.json
@@ -0,0 +1,423 @@
+{
+ "source": "https://www.bellingcat.com/",
+ "selection": "Long-form geolocation and chronolocation investigations and walkthroughs",
+ "article_count": 69,
+ "failed_count": 0,
+ "articles": [
+ {
+ "title": "Verification and Geolocation Tricks and Tips with Google Earth",
+ "date": "2014-07-09T14:04:35+00:00",
+ "url": "https://www.bellingcat.com/resources/how-tos/2014/07/09/verification-and-geolocation-tricks-and-tips-with-google-earth/",
+ "slug": "verification-and-geolocation-tricks-and-tips-with-google-earth"
+ },
+ {
+ "title": "A Beginner's Guide to Geolocating Videos",
+ "date": "2014-07-09T14:43:26+00:00",
+ "url": "https://www.bellingcat.com/resources/how-tos/2014/07/09/a-beginners-guide-to-geolocation/",
+ "slug": "a-beginners-guide-to-geolocation"
+ },
+ {
+ "title": "Geolocation Techniques – Mapping Landmarks",
+ "date": "2014-07-15T20:12:45+00:00",
+ "url": "https://www.bellingcat.com/resources/how-tos/2014/07/15/geolocation-techniques-mapping-landmarks/",
+ "slug": "geolocation-techniques-mapping-landmarks"
+ },
+ {
+ "title": "Geolocating the Missile Launcher Linked to the Downing of MH17",
+ "date": "2014-07-17T21:37:25+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/07/17/geolocating-the-missile-launcher-linked-to-the-downing-of-mh17/",
+ "slug": "geolocating-the-missile-launcher-linked-to-the-downing-of-mh17"
+ },
+ {
+ "title": "Identifying the Location of the MH17 Linked Missile Launcher From One Photograph",
+ "date": "2014-07-18T21:44:33+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/07/18/identifying-the-location-of-the-mh17-linked-missile-launcher-from-one-photograph/",
+ "slug": "identifying-the-location-of-the-mh17-linked-missile-launcher-from-one-photograph"
+ },
+ {
+ "title": "Locating the Rockets Used During the August 21st Sarin Attacks in Damascus",
+ "date": "2014-08-10T21:23:37+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/08/10/locating-the-rockets-used-during-the-august-21st-sarin-attacks-in-damascus/",
+ "slug": "locating-the-rockets-used-during-the-august-21st-sarin-attacks-in-damascus"
+ },
+ {
+ "title": "How to Locate a “Secret” Pro-Russian Training Camp",
+ "date": "2014-08-15T14:38:02+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/08/15/how-to-locate-a-secret-pro-russian-training-camp/",
+ "slug": "how-to-locate-a-secret-pro-russian-training-camp"
+ },
+ {
+ "title": "The Hills of Raqqa - Geolocating the James Foley Video",
+ "date": "2014-08-23T16:46:26+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/08/23/the-hills-of-raqqa-geolocating-the-james-foley-video/",
+ "slug": "the-hills-of-raqqa-geolocating-the-james-foley-video"
+ },
+ {
+ "title": "Geolocating the MH17 Buk Convoy in Russia",
+ "date": "2014-09-29T10:34:52+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/09/29/geolocating-the-mh17-buk-convoy-in-russia/",
+ "slug": "geolocating-the-mh17-buk-convoy-in-russia"
+ },
+ {
+ "title": "Geolocated June BUK convoy videos in Russia",
+ "date": "2014-11-07T21:11:57+00:00",
+ "url": "https://www.bellingcat.com/news/2014/11/07/geolocated-june-buk-convoy-videos-in-russia/",
+ "slug": "geolocated-june-buk-convoy-videos-in-russia"
+ },
+ {
+ "title": "Geolocated July BUK convoy videos in Russia",
+ "date": "2014-11-07T22:38:16+00:00",
+ "url": "https://www.bellingcat.com/news/europe/2014/11/07/geolocated-july-buk-convoy-videos-in-russia/",
+ "slug": "geolocated-july-buk-convoy-videos-in-russia"
+ },
+ {
+ "title": "Geolocating the June Russian Buk Convoy in Millerovo",
+ "date": "2014-11-08T23:52:44+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/11/08/geolocating-the-russian-buk-convoy-in-millerovo/",
+ "slug": "geolocating-the-russian-buk-convoy-in-millerovo"
+ },
+ {
+ "title": "Geolocating Tunisian Jihadists in Raqqa",
+ "date": "2014-12-19T14:49:37+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2014/12/19/geolocating-tunisian-jihadists-in-raqqa/",
+ "slug": "geolocating-tunisian-jihadists-in-raqqa"
+ },
+ {
+ "title": "Stunt Geolocation - Verifying the Unverifiable",
+ "date": "2015-03-03T09:43:07+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/03/03/stunt-geolocation-verifying-the-unverifiable/",
+ "slug": "stunt-geolocation-verifying-the-unverifiable"
+ },
+ {
+ "title": "Geolocating the Walter Scott Shooting",
+ "date": "2015-04-12T09:09:40+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/04/12/geolocating-the-walter-scott-shooting/",
+ "slug": "geolocating-the-walter-scott-shooting"
+ },
+ {
+ "title": "Air Rescue in Ramadi: Geolocating Made Easy with Head-Up Display",
+ "date": "2015-05-21T12:39:04+00:00",
+ "url": "https://www.bellingcat.com/news/2015/05/21/air-rescue-in-ramadi-geolocating-made-easy-with-heads-up-display/",
+ "slug": "air-rescue-in-ramadi-geolocating-made-easy-with-heads-up-display"
+ },
+ {
+ "title": "Confirmed: Active Russian Spetsnaz Serviceman Photographed in Luhansk",
+ "date": "2015-05-26T13:12:33+00:00",
+ "url": "https://www.bellingcat.com/news/europe/2015/05/26/mamai-geolocation/",
+ "slug": "mamai-geolocation"
+ },
+ {
+ "title": "Geolocating Stanislav Tarasov",
+ "date": "2015-05-28T19:43:52+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/05/28/geolocating-stanislav-tarasov/",
+ "slug": "geolocating-stanislav-tarasov"
+ },
+ {
+ "title": "Russia's Pantsir-S1s Geolocated in Ukraine",
+ "date": "2015-05-28T20:38:49+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/05/28/russias-pantsir-s1s-geolocated-in-ukraine/",
+ "slug": "russias-pantsir-s1s-geolocated-in-ukraine"
+ },
+ {
+ "title": "Confirming the Location of the Same Msta-S in Russia and Ukraine",
+ "date": "2015-05-29T13:56:28+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/05/29/confirming-the-location-of-the-same-msta-s-in-russia-and-ukraine/",
+ "slug": "confirming-the-location-of-the-same-msta-s-in-russia-and-ukraine"
+ },
+ {
+ "title": "Who to Trust, Google or the Russian MoD? A Guide to Verifying Google Earth Satellite Image Dates",
+ "date": "2015-06-05T06:54:53+00:00",
+ "url": "https://www.bellingcat.com/resources/how-tos/2015/06/05/google-earth-image-verification/",
+ "slug": "google-earth-image-verification"
+ },
+ {
+ "title": "Searching the Earth: Essential Geolocation Tools for Verification",
+ "date": "2015-07-25T07:30:15+00:00",
+ "url": "https://www.bellingcat.com/resources/how-tos/2015/07/25/searching-the-earth-essential-geolocation-tools-for-verification/",
+ "slug": "searching-the-earth-essential-geolocation-tools-for-verification"
+ },
+ {
+ "title": "Shadow of a Doubt: Crowdsourcing Time Verification of the MH17 Missile Launch Photo",
+ "date": "2015-08-07T12:53:32+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/08/07/shadow-of-a-doubt/",
+ "slug": "shadow-of-a-doubt"
+ },
+ {
+ "title": "Geolocating Russian Military Aircraft in Syria",
+ "date": "2015-09-22T12:01:14+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2015/09/22/geolocating-russian-military-aircraft-in-syria/",
+ "slug": "geolocating-russian-military-aircraft-in-syria"
+ },
+ {
+ "title": "Geolocation Once Again Disproves Russia's Targeting Claims in Syria",
+ "date": "2015-11-11T07:18:54+00:00",
+ "url": "https://www.bellingcat.com/news/2015/11/11/geolocation-once-again-disproves-russias-targeting-claims-in-syria/",
+ "slug": "geolocation-once-again-disproves-russias-targeting-claims-in-syria"
+ },
+ {
+ "title": "Geolocating Mexican Sicarios in Chihuahua",
+ "date": "2016-02-25T21:12:50+00:00",
+ "url": "https://www.bellingcat.com/news/americas/2016/02/25/geolocating-mexican-sicarios-in-chihuahua/",
+ "slug": "geolocating-mexican-sicarios-in-chihuahua"
+ },
+ {
+ "title": "Geolocation of Turkish Airstrikes on YPG Positions",
+ "date": "2016-10-21T12:52:25+00:00",
+ "url": "https://www.bellingcat.com/news/2016/10/21/geolocation-turkish-airstrikes-ypg-positions/",
+ "slug": "geolocation-turkish-airstrikes-ypg-positions"
+ },
+ {
+ "title": "Photographs of Ukrainian Buks Geolocated in Poltava Oblast, Nearly Three Years Later",
+ "date": "2017-04-04T18:41:32+00:00",
+ "url": "https://www.bellingcat.com/news/europe/2017/04/04/photographs-ukrainian-buks-geolocated-nearly-three-years-later/",
+ "slug": "photographs-ukrainian-buks-geolocated-nearly-three-years-later"
+ },
+ {
+ "title": "Crowdsourced Geolocation and Analysis of Coalition Airstrike Videos from Syria and Iraq",
+ "date": "2017-05-05T13:12:16+00:00",
+ "url": "https://www.bellingcat.com/news/middle-east/2017/05/05/geolocated-coalition-videos/",
+ "slug": "geolocated-coalition-videos"
+ },
+ {
+ "title": "Geolocating Libya's Social Media Executioner",
+ "date": "2017-09-04T13:40:30+00:00",
+ "url": "https://www.bellingcat.com/news/middle-east/2017/09/04/geolocating-libyas-social-media-executioner/",
+ "slug": "geolocating-libyas-social-media-executioner"
+ },
+ {
+ "title": "How a Werfalli Execution Site Was Geolocated",
+ "date": "2017-10-03T16:30:33+00:00",
+ "url": "https://www.bellingcat.com/news/middle-east/2017/10/03/how-an-execution-site-was-geolocated/",
+ "slug": "how-an-execution-site-was-geolocated"
+ },
+ {
+ "title": "New MH17 Photograph Geolocated to Donetsk",
+ "date": "2017-10-20T15:13:28+00:00",
+ "url": "https://www.bellingcat.com/news/europe/2017/10/20/new-mh17-photograph-geolocated-donetsk/",
+ "slug": "new-mh17-photograph-geolocated-donetsk"
+ },
+ {
+ "title": "How to Identify Burnt Villages by Satellite Imagery — Case-Studies from California, Nigeria and Myanmar",
+ "date": "2018-09-04T05:16:49+00:00",
+ "url": "https://www.bellingcat.com/resources/how-tos/2018/09/04/identify-burnt-villages-satellite-imagery -case-studies-california-nigeria-myanmar/",
+ "slug": "identify-burnt-villages-satellite-imagery -case-studies-california-nigeria-myanmar"
+ },
+ {
+ "title": "Europol’s Asian City Child Abuse Photographs Geolocated",
+ "date": "2018-11-08T09:21:33+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2018/11/08/europols-asian-city-child-abuse-photographs-geolocated/",
+ "slug": "europols-asian-city-child-abuse-photographs-geolocated"
+ },
+ {
+ "title": "Using Time-Lapse Satellite Imagery To Detect Infrastructure Changes: Case-Studies via Myanmar, Nigeria and the South China Sea",
+ "date": "2018-11-19T18:52:35+00:00",
+ "url": "https://www.bellingcat.com/resources/how-tos/2018/11/19/using-time-lapse-satellite-imagery-detect-infrastructure-changes-case-studies-via-myanmar-nigeria-south-china-sea/",
+ "slug": "using-time-lapse-satellite-imagery-detect-infrastructure-changes-case-studies-via-myanmar-nigeria-south-china-sea"
+ },
+ {
+ "title": "Geolocation of Infrastructure Destruction in Cameroon: A Case Study of Kumbo and Kumfutu",
+ "date": "2018-11-21T23:41:12+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2018/11/21/geolocation-infrastructure-destruction-cameroon-case-study-kumbo-kumfutu/",
+ "slug": "geolocation-infrastructure-destruction-cameroon-case-study-kumbo-kumfutu"
+ },
+ {
+ "title": "Where’s Babiš? Geolocating the Alleged Father-Son Kidnapping Mystery",
+ "date": "2019-02-07T11:00:54+00:00",
+ "url": "https://www.bellingcat.com/news/2019/02/07/wheres-babis-geolocating-the-alleged-father-son-kidnapping-mystery/",
+ "slug": "wheres-babis-geolocating-the-alleged-father-son-kidnapping-mystery"
+ },
+ {
+ "title": "Geolocation and a Philosopher's Stone in Kashmir",
+ "date": "2019-03-04T08:32:09+00:00",
+ "url": "https://www.bellingcat.com/resources/case-studies/2019/03/04/geolocation-and-a-philosophers-stone-in-kashmir/",
+ "slug": "geolocation-and-a-philosophers-stone-in-kashmir"
+ },
+ {
+ "title": "How To Use Google Earth's Three Dimensional View: Feat. Syria, Yemen, Sudan",
+ "date": "2019-03-05T16:31:01+00:00",
+ "url": "https://www.bellingcat.com/resources/how-tos/2019/03/05/how-to-use-google-earths-three-dimensional-view-feat-syria-yemen-sudan/",
+ "slug": "how-to-use-google-earths-three-dimensional-view-feat-syria-yemen-sudan"
+ },
+ {
+ "title": "Locating The Netherlands' Most Wanted Criminal By Scrutinising Instagram",
+ "date": "2019-03-19T23:22:31+00:00",
+ "url": "https://www.bellingcat.com/news/europe/2019/03/19/locating-the-netherlands-most-wanted-criminal-by-scrutinising-instagram/",
+ "slug": "locating-the-netherlands-most-wanted-criminal-by-scrutinising-instagram"
+ },
+ {
+ "title": "More Europol’s “Stop Child Abuse” Photographs Geolocated",
+ "date": "2019-07-02T14:39:16+00:00",
+ "url": "https://www.bellingcat.com/news/europe/2019/07/02/more-europols-stop-child-abuse-photographs-geolocated/",
+ "slug": "more-europols-stop-child-abuse-photographs-geolocated"
+ },
+ {
+ "title": "Europol’s Child Abuse Image Geolocated In Ukraine: A Forgotten Story Hidden Behind A Landscape",
+ "date": "2019-09-11T08:27:15+00:00",
+ "url": "https://www.bellingcat.com/news/2019/09/11/europols-child-abuse-image-geolocated-in-ukraine-a-forgotten-story-hidden-behind-a-landscape/",
+ "slug": "europols-child-abuse-image-geolocated-in-ukraine-a-forgotten-story-hidden-behind-a-landscape"
+ },
+ {
+ "title": "Two Europol StopChildAbuse Images Geolocated: Part I — Madagascar",
+ "date": "2019-12-05T20:23:59+00:00",
+ "url": "https://www.bellingcat.com/news/2019/12/05/two-europol-stopchildabuse-images-geolocated-part-i-madagascar/",
+ "slug": "two-europol-stopchildabuse-images-geolocated-part-i-madagascar"
+ },
+ {
+ "title": "Two Europol StopChildAbuse Images Geolocated: Part II — Cambodia",
+ "date": "2019-12-17T14:23:30+00:00",
+ "url": "https://www.bellingcat.com/news/2019/12/17/two-europol-stopchildabuse-images-geolocated-part-ii-cambodia/",
+ "slug": "two-europol-stopchildabuse-images-geolocated-part-ii-cambodia"
+ },
+ {
+ "title": "Video Apparently Showing Flight PS752 Missile Strike Geolocated to Iranian Suburb",
+ "date": "2020-01-09T21:05:51+00:00",
+ "url": "https://www.bellingcat.com/news/middle-east/2020/01/09/video-apparently-showing-flight-ps572-missile-strike-geolocated-to-iranian-suburb/",
+ "slug": "video-apparently-showing-flight-ps572-missile-strike-geolocated-to-iranian-suburb"
+ },
+ {
+ "title": "Geolocating Venezuelan Lawmakers In Europe",
+ "date": "2020-01-21T19:12:03+00:00",
+ "url": "https://www.bellingcat.com/news/2020/01/21/geolocating-venezuelan-lawmakers-in-europe/",
+ "slug": "geolocating-venezuelan-lawmakers-in-europe"
+ },
+ {
+ "title": "Using the Sun and the Shadows for Geolocation",
+ "date": "2020-12-03T15:02:44+00:00",
+ "url": "https://www.bellingcat.com/resources/2020/12/03/using-the-sun-and-the-shadows-for-geolocation/",
+ "slug": "using-the-sun-and-the-shadows-for-geolocation"
+ },
+ {
+ "title": "John Doe 29: Image From FBI Child Exploitation Case Geolocated to Turkey",
+ "date": "2021-04-30T11:40:46+00:00",
+ "url": "https://www.bellingcat.com/news/2021/04/30/john-doe-29-image-from-fbi-child-exploitation-case-geolocated-in-turkey/",
+ "slug": "john-doe-29-image-from-fbi-child-exploitation-case-geolocated-in-turkey"
+ },
+ {
+ "title": "Where in the World is Q? Clues from Image Metadata",
+ "date": "2021-05-10T18:47:26+00:00",
+ "url": "https://www.bellingcat.com/news/rest-of-world/2021/05/10/where-in-the-world-is-q-clues-from-image-metadata/",
+ "slug": "where-in-the-world-is-q-clues-from-image-metadata"
+ },
+ {
+ "title": "Unsure When a Video or Photo was Taken? How to Tell by Measuring the Length of Shadows",
+ "date": "2021-05-18T11:29:04+00:00",
+ "url": "https://www.bellingcat.com/resources/2021/05/18/unsure-when-a-video-or-photo-was-taken-how-to-tell-by-measuring-the-length-of-shadows/",
+ "slug": "unsure-when-a-video-or-photo-was-taken-how-to-tell-by-measuring-the-length-of-shadows"
+ },
+ {
+ "title": "How we Geolocated a Photo of a Russian Missile Programming Team",
+ "date": "2022-10-28T12:48:06+00:00",
+ "url": "https://www.bellingcat.com/resources/2022/10/28/how-we-geolocated-a-photo-of-a-russian-missile-programming-team/",
+ "slug": "how-we-geolocated-a-photo-of-a-russian-missile-programming-team"
+ },
+ {
+ "title": "Wanted by Interpol, Relaxing in Dubai: Geolocating Isabel dos Santos’ Life of Luxury",
+ "date": "2023-02-03T10:41:23+00:00",
+ "url": "https://www.bellingcat.com/news/2023/02/03/wanted-by-interpol-relaxing-in-dubai-geolocating-isabel-dos-santos-life-of-luxury/",
+ "slug": "wanted-by-interpol-relaxing-in-dubai-geolocating-isabel-dos-santos-life-of-luxury"
+ },
+ {
+ "title": "International Far-Right Fight Night Comes to Budapest",
+ "date": "2023-05-03T13:49:18+00:00",
+ "url": "https://www.bellingcat.com/news/2023/05/03/international-far-right-fight-night-comes-to-budapest/",
+ "slug": "international-far-right-fight-night-comes-to-budapest"
+ },
+ {
+ "title": "Chronolocation: Determining When a Photo was Taken Using Facebook, Google Street View and Assorted Tiny Details",
+ "date": "2023-05-08T13:20:03+00:00",
+ "url": "https://www.bellingcat.com/resources/2023/05/08/chronolocation-determining-when-a-photo-was-taken-using-facebook-google-street-view-and-assorted-tiny-details/",
+ "slug": "chronolocation-determining-when-a-photo-was-taken-using-facebook-google-street-view-and-assorted-tiny-details"
+ },
+ {
+ "title": "Finding Geolocation Leads with Bellingcat's OpenStreetMap Search Tool",
+ "date": "2023-05-08T16:42:29+00:00",
+ "url": "https://www.bellingcat.com/resources/how-tos/2023/05/08/finding-geolocation-leads-with-bellingcats-openstreetmap-search-tool/",
+ "slug": "finding-geolocation-leads-with-bellingcats-openstreetmap-search-tool"
+ },
+ {
+ "title": "Russia's Foreign Fighters: Geolocating the Nepalis Training in the Russian Army",
+ "date": "2023-06-23T18:40:01+00:00",
+ "url": "https://www.bellingcat.com/news/2023/06/23/russias-foreign-fighters-geolocating-the-nepalis-training-in-the-russian-army/",
+ "slug": "russias-foreign-fighters-geolocating-the-nepalis-training-in-the-russian-army"
+ },
+ {
+ "title": "More than Mountaineering: Using PeakVisor for Geolocation",
+ "date": "2023-07-13T13:11:50+00:00",
+ "url": "https://www.bellingcat.com/resources/2023/07/13/more-than-mountaineering-using-peakvisor-for-geolocation/",
+ "slug": "more-than-mountaineering-using-peakvisor-for-geolocation"
+ },
+ {
+ "title": "Can AI Chatbots Be Used for Geolocation?",
+ "date": "2023-07-14T12:13:06+00:00",
+ "url": "https://www.bellingcat.com/resources/2023/07/14/can-ai-chatbots-be-used-for-geolocation/",
+ "slug": "can-ai-chatbots-be-used-for-geolocation"
+ },
+ {
+ "title": "Geolocating Russia’s Disgraced General Surovikin",
+ "date": "2023-09-06T18:04:10+00:00",
+ "url": "https://www.bellingcat.com/news/2023/09/06/geolocating-russias-disgraced-general-surovikin/",
+ "slug": "geolocating-russias-disgraced-general-surovikin"
+ },
+ {
+ "title": "Geolocating Hamas-Led Attacks on Israeli Civilians",
+ "date": "2023-10-20T13:20:25+00:00",
+ "url": "https://www.bellingcat.com/news/2023/10/20/geolocating-hamas-led-attacks-on-israeli-civilians/",
+ "slug": "geolocating-hamas-led-attacks-on-israeli-civilians"
+ },
+ {
+ "title": "From Pixels to Punches: Geolocating a neo-Nazi and White Nationalist Combat Event in Los Angeles",
+ "date": "2023-11-16T18:05:33+00:00",
+ "url": "https://www.bellingcat.com/news/2023/11/16/from-pixels-to-punches-geolocating-a-neo-nazi-and-white-nationalist-combat-event-in-los-angeles/",
+ "slug": "from-pixels-to-punches-geolocating-a-neo-nazi-and-white-nationalist-combat-event-in-los-angeles"
+ },
+ {
+ "title": "Geolocating a US Far-Right Fight Night",
+ "date": "2024-03-15T17:01:17+00:00",
+ "url": "https://www.bellingcat.com/resources/2024/03/15/geolocating-a-us-far-right-fight-night/",
+ "slug": "geolocating-a-us-far-right-fight-night"
+ },
+ {
+ "title": "Fugitive US Militant Ammon Bundy Geolocated to Utah",
+ "date": "2024-03-27T22:24:25+00:00",
+ "url": "https://www.bellingcat.com/news/2024/03/27/fugitive-us-militant-ammon-bundy-geolocated-to-utah/",
+ "slug": "fugitive-us-militant-ammon-bundy-geolocated-to-utah"
+ },
+ {
+ "title": "Exploring the Skyline: How we Located an Alleged Cartel Member in Dubai",
+ "date": "2024-07-16T11:41:21+00:00",
+ "url": "https://www.bellingcat.com/resources/2024/07/16/dubai-uae-cartel-organised-crime-geolocation-open-source-guide-technique-tools/",
+ "slug": "dubai-uae-cartel-organised-crime-geolocation-open-source-guide-technique-tools"
+ },
+ {
+ "title": "Chasing Shadows: Geolocate Images with Bellingcat's Shadow Finder Tool",
+ "date": "2024-08-22T09:04:18+00:00",
+ "url": "https://www.bellingcat.com/resources/2024/08/22/shadow-geolocate-geolocation-locate-image-tool-open-source-bellingcat-measure/",
+ "slug": "shadow-geolocate-geolocation-locate-image-tool-open-source-bellingcat-measure"
+ },
+ {
+ "title": "Geolocating Site Where Hamas Leader Yahya Sinwar Was Killed",
+ "date": "2024-10-17T22:26:13+00:00",
+ "url": "https://www.bellingcat.com/news/2024/10/17/geolocating-site-where-hamas-leader-yahya-sinwar-was-killed/",
+ "slug": "geolocating-site-where-hamas-leader-yahya-sinwar-was-killed"
+ },
+ {
+ "title": "Have LLMs Finally Mastered Geolocation?",
+ "date": "2025-06-06T07:33:08+00:00",
+ "url": "https://www.bellingcat.com/resources/how-tos/2025/06/06/have-llms-finally-mastered-geolocation/",
+ "slug": "have-llms-finally-mastered-geolocation"
+ },
+ {
+ "title": "LLMs Vs. Geolocation: GPT-5 Performs Worse Than Other AI Models",
+ "date": "2025-08-14T10:42:44+00:00",
+ "url": "https://www.bellingcat.com/resources/2025/08/14/llms-vs-geolocation-gpt-5-performs-worse-than-other-ai-models/",
+ "slug": "llms-vs-geolocation-gpt-5-performs-worse-than-other-ai-models"
+ },
+ {
+ "title": "Geolocating Darfur Killings of Those Escaping Al Fashir",
+ "date": "2025-11-05T13:35:27+00:00",
+ "url": "https://www.bellingcat.com/news/2025/11/05/geolocating-darfur-killings-of-those-escaping-al-fashir/",
+ "slug": "geolocating-darfur-killings-of-those-escaping-al-fashir"
+ }
+ ],
+ "failures": []
+}
diff --git a/nixintel_geolocation_json/2019-06-26_tell-explain-describe-asking-the-right-questions-to-get-the-right-answers-quiztime-25th-june-2019.json b/nixintel_geolocation_json/2019-06-26_tell-explain-describe-asking-the-right-questions-to-get-the-right-answers-quiztime-25th-june-2019.json
new file mode 100644
index 0000000000000000000000000000000000000000..a9b7cdf15c8266b0f828c670129411c8acc4f41f
--- /dev/null
+++ b/nixintel_geolocation_json/2019-06-26_tell-explain-describe-asking-the-right-questions-to-get-the-right-answers-quiztime-25th-june-2019.json
@@ -0,0 +1,58 @@
+{
+ "slug": "tell-explain-describe-asking-the-right-questions-to-get-the-right-answers-quiztime-25th-june-2019",
+ "title": "Tell, Explain, Describe… Asking The Right Questions To Get The Right Answers – Quiztime 25th June 2019",
+ "url": "https://nixintel.info/osint/tell-explain-describe-asking-the-right-questions-to-get-the-right-answers-quiztime-25th-june-2019/",
+ "date": "2019-06-26T21:49:07+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
In the short time that I have been blogging, many of the OSINT tools I’ve referred to have already become obsolete or no longer exist. Facebook removed Graph Search, Twitter removed geotagged Tweets, IntelTechniques lost its tools, and Pipl went behind a paywall. Although these setbacks can be frustrating, it’s important to remember that the foundation of good OSINT investigation is a sound method, not just the right tools. Tools are transient, but the methods of good investigation stay the same.
\n
With that in mind, this article will look at how a sound investigative method can make even difficult OSINT challenges possible. As an example I’ll use Marc Krueger’s recent Quiztime challenge. How do you geolocate this image?
\n
\n
\n
At first this might seem a little bit daunting. Marc has edited the image in such a way that obscures a lot of the detail. This makes reverse image searches ineffective but this also adds an element of realism – most real-life OSINT investigations are not straightforward. So where to start? By asking the right questions and using gap analysis we can determine what information we already, and what information is needed to solve the puzzle.
\n
Gap Analysis
\n
I wrote a previous blog post about how Gap Analysis can be applied to OSINT investigations but I’ll summarise it again here. It is a simple investigative method that asks four key questions about the information gathered up to that point and then determines what enquiries are needed to take the investigation forward. The questions are:
\n
1) What do I know?
\n
2) What does this mean?
\n
3) What do I need to know?
\n
4) How do I find out?
\n
It’s a very effective way to bring a lot of clarity to a large collection of confusing data. Notice that questions about tools (“How do I find out?”) come last. Every time an OSINT enquiry gets muddled or comes to a dead end, reviewing the case with gap analysis can help identify new ways forward.
\n
Ask The Right Questions
\n
That’s all very well when you’ve already got some information in front of you. It’s easy to ask “What do I know?” when you’re already sat on collection of images, street signs, IP addresses, usernames, and e-mail addresses. But what if you’ve got nothing to start with? No one managed to solve Marc’s challenge in the first 24 hours and on the surface it appears very difficult, but by asking the right questions of the image it is possible to start gathering information that can be assessed with gap analysis and a solution can be found quite quickly.
\nDon’t worry Rae!\n
Asking questions of an image might seem unusual, but we do it subconsciously all the time. In traditional investigations such as those carried out by journalists or detectives, asking the right questions of witnesses is a central part of any case. Asking good questions produces a lot of information, but bad questions do not. Good questions are open and elicit a wide range of information, but closed questions only allow for a narrow range of answers. Even worse are leading questions, which already contain the answer. Political journalists love to ask leading questions, but it makes for terrible interviews.
Good questions begin with “Tell me…“, “Explain to me….“, “Describe to me….“. They can be followed up with more open questions like What? Why? When? Where? or How? to draw out further detail. These kind of questions are the foundation of good investigative interviews because they elicit a lot of information. Closed questions on the other hand shut down the flow of information. Questions that begin with “Is this place x?” only offer the possibility of a “yes” or “no” answer. Even worse are the leading questions so beloved of TV interviewers: “Isn’t it the case that the government said they would spend x but in fact they only spent y?” The anticipated answer is already in the question. These kinds of questions have their place, but they are useless for eliciting any kind of information.
\n
Let’s see how this could be applied to Marc’s challenge photo…
\n
\n
\n
It may appear at first that we do not have an awful lot of detail, but by asking open questions of the image, we can actually assemble quite a lot of information to begin applying gap analysis to take the challenge forward and finally solve it. Here’s a good open question to start:
\n
“Tell me…everything you can see in the image”
\n
An old building
\n
Graffiti
\n
An old kiosk type building covered in graffiti
\n
A tarmac road
\n
Sunny weather
\n
A tree
\n
Shuttered windows and door
\n
Two signs next to the door
\n
Two flagpoles above the door
\n
A sign next to the door that says “ESPASOS” or maybe “ESPANOL”
\n
Shutters on the kiosk
\n
So already we have eleven pieces of information to work with. It might not seem like much yet, but it is a start. Let’s ask more open questions to develop this a bit further…
\n
“Describe everything in the photo. What does it look like?”
\n
Looks old.
\n
Abandoned.
\n
Looks like southern Europe or Mediterranean.
\n
The road looks quite new but the buildings look old.
\n
A light above the door is broken.
\n
The graffiti says “hier” and lots of other things that aren’t all clear.
\n
So very quickly we have five additional pieces of information just by looking at the photo methodically. If we had just asked closed questions like “Is this in Italy or Spain?”, “Is it a house?” or “This is where Marc Krueger went on holiday, isn’t it?” we would not have a lot of information to work with. In normal daily conversation we ask closed and leading questions all the time, but with investigations it is necessary to make a conscious effort to ask open questions about the material you are working with.
\n
Applying Gap Analysis
\n
So now there is a little information to work with, it is possible to begin to apply gap analysis to determine what to do next. Let’s make a start:
\n
1) What do I know?
\n
This is an old abandoned building covered in graffiti.
\n
It is opposite a kiosk that has been shuttered up and which is also covered in graffiti.
\n
There are flagpoles but no flags.
\n
The sign next to the door says either “ESPASO” or “ESPANO”.
\n
There is a road between the kiosk and the building.
\n
The sunny weather and the architecture suggest the building is in southern Europe or somewhere around the Mediterranean.
\n
There is a tree with leaves on.
\n
2) What does this mean?
\n
The building is probably no longer in use.
\n
It once flew flags from the flagpole.
\n
The word “ESPASO” or “ESPANO” suggest a Latin language rather than a Germanic one. This also makes North Africa, the Balkans, or Greece unlikely locations despite the similar architecture.
\n
The buildings are probably not in use, but the road looks as though it may be (it is not covered by weeds, unlike the building)
\n
3) What do I need to know?
\n
What kind of locations could be so close to a road?
\n
What language does the word “ESPASO/ESPANO” belong to – this could at least give me the country, if not the exact location.
\n
What kind of buildings fly flags?
\n
4) How do I find out?
\n
Use Google Translate to investigate the “ESPASO / ESPANO” word.
\n
Make a list of all kinds of buildings that might fit the criteria of being next to a road, flying a flag, and being close to a kiosk.
\n
So after experimenting with Google Translate, I couldn’t find any suitable words in Spanish, Portugese, or French that had the root “ESPASO”. “ESPANO” was a much better fit. A simple Google search showed that “ESPANO” is the root of words that mean Spain or Spanish. My Spanish is almost non-existent so I had to trust Google at this point, but it was a vital clue. I now knew that the sign next to the door of the building probably said “Spain” or ” Spanish”.
\nIt’s one of the little buildings somewhere on this part of the map.\n
The second part of the “How Do I Find Out?” didn’t even really require the internet. What kind of building would be next to a road, opposite a kiosk, have flagpoles, and have “Spain” or “Spanish” next to the door? There are very few possibilities:
\n
1) An embassy
\n
2) A consulate
\n
3) An immigration centre or border checkpoint
\n
4) A Spanish restaurant. This is some kind of cruel joke by Marc.
\n
So having established roughly what kind of building it is, and what country it is probably associated to, it was possible to use another round of gap analysis before firing up the tools.
\n
1) What do I know?
\n
This building is either in Spain or is associated to Spain in some way.
\n
It is an abandoned building/kiosk covered in graffiti.
\n
The kind of building that would have these features as well as flagpoles could be an embassy, border checkpoint, or other diplomatic location.
\n
2) What does this mean?
\n
If it was in use, it probably isn’t now. The target building is probably an abandoned Spanish diplomatic building of some kind. For some reason this building is no longer needed or has been replaced.
\n
3) What do I need to know?
\n
I need to find out about abandoned Spanish diplomatic buildings. Are there any? Where are they? Do they look like the image in the challenge?
\n
4) How do I find out?
\n
Google, Google Maps, and Google Images will probably be enough.
\n
So I began researching Spanish Embassies. Fortunately Embassy Pages proved to be a helpful resource. It only took a few minutes to identify all Spanish embassies/consulates and other diplomatic missions in Europe and then check them against Google Maps / Images. There were no real matches and it struck me that the average embassy building is much bigger and more ornate than the one in the picture. I also couldn’t find anything that suggested Spain had any abandoned embassies. There was one exception in Morocco, but it didn’t match:
\nAbandoned Spanish consulate in Morocco.\n
Having discounted embassies I then turned to border checkpoints. A little Geography knowledge (or Google Maps) is required. Which countries have a border with Spain? Spain has a border with Portugal, Andorra, France, and Gibraltar – but also not forgetting its outposts of Ceuta and Melilla in North Africa.
\n
A Google Image search for “Spanish border” related photos brought up lots of pictures of the high fences at Melilla and Ceuta designed to keep out immigrants from Africa. The high security needed there means that these borders don’t fit with the old abandoned location in the challenge photo, so I discounted them. A little knowledge of European history means that because the Schengen Agreement, most internal EU borders were abolished in the 1990s. This was more fitting for the image I was trying to geolocate.
\n
Finally I conducted an image search for “old Spanish border post”. On the fourth page I found this familiar looking kiosk:
\n
\n
A click on the link leads to an article about the history of cross-border escapes between France and Spain in WW2. From it I learned that the kiosk in the picture was at the French-Spanish border at Portrou:
\n
\n
A further image search for the border post at Portrou brought up this. It is now easy to see that the sign on the other side of the door was for the French Police.
\n
\n
And finally a glimpse of the location on Streetview.
\n
\n
Having established the location, it is now possible to research Marc’s question about the local road number changes, but that’s beyond the purpose of this post. This was not an easy challenge on the surface, but the solution did not involve any tools more complicated that Google Images. The key to solving this challenge was really about looking closely at the image and asking the right questions, and then developing them with gap analysis to find the solution. These methods are always reliable to start a difficult investigation and they’ll never be turned off by social media providers!
",
+ "content_text": "In the short time that I have been blogging, many of the OSINT tools I’ve referred to have already become obsolete or no longer exist. Facebook removed Graph Search, Twitter removed geotagged Tweets, IntelTechniques lost its tools, and Pipl went behind a paywall. Although these setbacks can be frustrating, it’s important to remember that the foundation of good OSINT investigation is a sound method, not just the right tools. Tools are transient, but the methods of good investigation stay the same.\n\nWith that in mind, this article will look at how a sound investigative method can make even difficult OSINT challenges possible. As an example I’ll use Marc Krueger’s recent Quiztime challenge. How do you geolocate this image?\n\nAt first this might seem a little bit daunting. Marc has edited the image in such a way that obscures a lot of the detail. This makes reverse image searches ineffective but this also adds an element of realism – most real-life OSINT investigations are not straightforward. So where to start? By asking the right questions and using gap analysis we can determine what information we already, and what information is needed to solve the puzzle.\n\nGap Analysis\n\nI wrote a previous blog post about how Gap Analysis can be applied to OSINT investigations but I’ll summarise it again here. It is a simple investigative method that asks four key questions about the information gathered up to that point and then determines what enquiries are needed to take the investigation forward. The questions are:\n\n1) What do I know?\n\n2) What does this mean?\n\n3) What do I need to know?\n\n4) How do I find out?\n\nIt’s a very effective way to bring a lot of clarity to a large collection of confusing data. Notice that questions about tools (“How do I find out?”) come last. Every time an OSINT enquiry gets muddled or comes to a dead end, reviewing the case with gap analysis can help identify new ways forward.\n\nAsk The Right Questions\n\nThat’s all very well when you’ve already got some information in front of you. It’s easy to ask “What do I know?” when you’re already sat on collection of images, street signs, IP addresses, usernames, and e-mail addresses. But what if you’ve got nothing to start with? No one managed to solve Marc’s challenge in the first 24 hours and on the surface it appears very difficult, but by asking the right questions of the image it is possible to start gathering information that can be assessed with gap analysis and a solution can be found quite quickly.\n\nDon’t worry Rae!\n\nAsking questions of an image might seem unusual, but we do it subconsciously all the time. In traditional investigations such as those carried out by journalists or detectives, asking the right questions of witnesses is a central part of any case. Asking good questions produces a lot of information, but bad questions do not. Good questions are open and elicit a wide range of information, but closed questions only allow for a narrow range of answers. Even worse are leading questions, which already contain the answer. Political journalists love to ask leading questions, but it makes for terrible interviews.\n\nGood entertainment, ineffective interview technique.\n\nGood questions begin with “Tell me…“, “Explain to me….“, “Describe to me….“. They can be followed up with more open questions like What? Why? When? Where? or How? to draw out further detail. These kind of questions are the foundation of good investigative interviews because they elicit a lot of information. Closed questions on the other hand shut down the flow of information. Questions that begin with “Is this place x?” only offer the possibility of a “yes” or “no” answer. Even worse are the leading questions so beloved of TV interviewers: “Isn’t it the case that the government said they would spend x but in fact they only spent y?” The anticipated answer is already in the question. These kinds of questions have their place, but they are useless for eliciting any kind of information.\n\nLet’s see how this could be applied to Marc’s challenge photo…\n\nIt may appear at first that we do not have an awful lot of detail, but by asking open questions of the image, we can actually assemble quite a lot of information to begin applying gap analysis to take the challenge forward and finally solve it. Here’s a good open question to start:\n\n“Tell me…everything you can see in the image”\n\nAn old building\n\nGraffiti\n\nAn old kiosk type building covered in graffiti\n\nA tarmac road\n\nSunny weather\n\nA tree\n\nShuttered windows and door\n\nTwo signs next to the door\n\nTwo flagpoles above the door\n\nA sign next to the door that says “ESPASOS” or maybe “ESPANOL”\n\nShutters on the kiosk\n\nSo already we have eleven pieces of information to work with. It might not seem like much yet, but it is a start. Let’s ask more open questions to develop this a bit further…\n\n“Describe everything in the photo. What does it look like?”\n\nLooks old.\n\nAbandoned.\n\nLooks like southern Europe or Mediterranean.\n\nThe road looks quite new but the buildings look old.\n\nA light above the door is broken.\n\nThe graffiti says “hier” and lots of other things that aren’t all clear.\n\nSo very quickly we have five additional pieces of information just by looking at the photo methodically. If we had just asked closed questions like “Is this in Italy or Spain?”, “Is it a house?” or “This is where Marc Krueger went on holiday, isn’t it?” we would not have a lot of information to work with. In normal daily conversation we ask closed and leading questions all the time, but with investigations it is necessary to make a conscious effort to ask open questions about the material you are working with.\n\nApplying Gap Analysis\n\nSo now there is a little information to work with, it is possible to begin to apply gap analysis to determine what to do next. Let’s make a start:\n\n1) What do I know?\n\nThis is an old abandoned building covered in graffiti.\n\nIt is opposite a kiosk that has been shuttered up and which is also covered in graffiti.\n\nThere are flagpoles but no flags.\n\nThe sign next to the door says either “ESPASO” or “ESPANO”.\n\nThere is a road between the kiosk and the building.\n\nThe sunny weather and the architecture suggest the building is in southern Europe or somewhere around the Mediterranean.\n\nThere is a tree with leaves on.\n\n2) What does this mean?\n\nThe building is probably no longer in use.\n\nIt once flew flags from the flagpole.\n\nThe word “ESPASO” or “ESPANO” suggest a Latin language rather than a Germanic one. This also makes North Africa, the Balkans, or Greece unlikely locations despite the similar architecture.\n\nThe buildings are probably not in use, but the road looks as though it may be (it is not covered by weeds, unlike the building)\n\n3) What do I need to know?\n\nWhat kind of locations could be so close to a road?\n\nWhat language does the word “ESPASO/ESPANO” belong to – this could at least give me the country, if not the exact location.\n\nWhat kind of buildings fly flags?\n\n4) How do I find out?\n\nUse Google Translate to investigate the “ESPASO / ESPANO” word.\n\nMake a list of all kinds of buildings that might fit the criteria of being next to a road, flying a flag, and being close to a kiosk.\n\nSo after experimenting with Google Translate, I couldn’t find any suitable words in Spanish, Portugese, or French that had the root “ESPASO”. “ESPANO” was a much better fit. A simple Google search showed that “ESPANO” is the root of words that mean Spain or Spanish. My Spanish is almost non-existent so I had to trust Google at this point, but it was a vital clue. I now knew that the sign next to the door of the building probably said “Spain” or ” Spanish”.\n\nIt’s one of the little buildings somewhere on this part of the map.\n\nThe second part of the “How Do I Find Out?” didn’t even really require the internet. What kind of building would be next to a road, opposite a kiosk, have flagpoles, and have “Spain” or “Spanish” next to the door? There are very few possibilities:\n\n1) An embassy\n\n2) A consulate\n\n3) An immigration centre or border checkpoint\n\n4) A Spanish restaurant. This is some kind of cruel joke by Marc.\n\nSo having established roughly what kind of building it is, and what country it is probably associated to, it was possible to use another round of gap analysis before firing up the tools.\n\n1) What do I know?\n\nThis building is either in Spain or is associated to Spain in some way.\n\nIt is an abandoned building/kiosk covered in graffiti.\n\nThe kind of building that would have these features as well as flagpoles could be an embassy, border checkpoint, or other diplomatic location.\n\n2) What does this mean?\n\nIf it was in use, it probably isn’t now. The target building is probably an abandoned Spanish diplomatic building of some kind. For some reason this building is no longer needed or has been replaced.\n\n3) What do I need to know?\n\nI need to find out about abandoned Spanish diplomatic buildings. Are there any? Where are they? Do they look like the image in the challenge?\n\n4) How do I find out?\n\nGoogle, Google Maps, and Google Images will probably be enough.\n\nSo I began researching Spanish Embassies. Fortunately Embassy Pages proved to be a helpful resource. It only took a few minutes to identify all Spanish embassies/consulates and other diplomatic missions in Europe and then check them against Google Maps / Images. There were no real matches and it struck me that the average embassy building is much bigger and more ornate than the one in the picture. I also couldn’t find anything that suggested Spain had any abandoned embassies. There was one exception in Morocco, but it didn’t match:\n\nAbandoned Spanish consulate in Morocco.\n\nHaving discounted embassies I then turned to border checkpoints. A little Geography knowledge (or Google Maps) is required. Which countries have a border with Spain? Spain has a border with Portugal, Andorra, France, and Gibraltar – but also not forgetting its outposts of Ceuta and Melilla in North Africa.\n\nA Google Image search for “Spanish border” related photos brought up lots of pictures of the high fences at Melilla and Ceuta designed to keep out immigrants from Africa. The high security needed there means that these borders don’t fit with the old abandoned location in the challenge photo, so I discounted them. A little knowledge of European history means that because the Schengen Agreement, most internal EU borders were abolished in the 1990s. This was more fitting for the image I was trying to geolocate.\n\nFinally I conducted an image search for “old Spanish border post”. On the fourth page I found this familiar looking kiosk:\n\nA click on the link leads to an article about the history of cross-border escapes between France and Spain in WW2. From it I learned that the kiosk in the picture was at the French-Spanish border at Portrou:\n\nA further image search for the border post at Portrou brought up this. It is now easy to see that the sign on the other side of the door was for the French Police.\n\nAnd finally a glimpse of the location on Streetview.\n\nHaving established the location, it is now possible to research Marc’s question about the local road number changes, but that’s beyond the purpose of this post. This was not an easy challenge on the surface, but the solution did not involve any tools more complicated that Google Images. The key to solving this challenge was really about looking closely at the image and asking the right questions, and then developing them with gap analysis to find the solution. These methods are always reliable to start a difficult investigation and they’ll never be turned off by social media providers!\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/D97j7ZiW4AA7hfE.jpg-orig-1024x576.jpeg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Screenshot-2019-06-26-at-22.17.24-1024x421.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/55832920_boris1.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Screenshot-2019-06-26-at-22.43.28-1024x357.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/abandoned-spanish-consulate-building-sidi-450w-421852126.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/07_16grenze_portbou-1-1024x738.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Screenshot-2019-06-26-at-22.33.32-1024x463.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/french-spanish-border-crossing-above-portbou-in-catalonia-spain-sept-E8K29N-1024x752.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Screenshot-2019-06-26-at-22.32.40-1024x591.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2019-06-28_quiztime-28th-june-2019-planespotting-on-the-heide.json b/nixintel_geolocation_json/2019-06-28_quiztime-28th-june-2019-planespotting-on-the-heide.json
new file mode 100644
index 0000000000000000000000000000000000000000..fccb1d8abb6f439792d8727e738b34967105428e
--- /dev/null
+++ b/nixintel_geolocation_json/2019-06-28_quiztime-28th-june-2019-planespotting-on-the-heide.json
@@ -0,0 +1,53 @@
+{
+ "slug": "quiztime-28th-june-2019-planespotting-on-the-heide",
+ "title": "Quiztime 28th June 2019 – Planespotting On The Heide",
+ "url": "https://nixintel.info/osint/quiztime-28th-june-2019-planespotting-on-the-heide/",
+ "date": "2019-06-28T21:25:59+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
When I was a child I spent a lot of time living around various military establishments in Germany. In the 1980s the north German plain was packed with NATO soldiers ready to fight the Russians in case they ever rolled over the East German border (remember when that was a thing?). Nowadays the Russians prefer to use spearphishing and poisoned door handles instead of tanks, so there aren’t as many soldiers in Germany as there once were. Today’s Quiztime challenge was a nice revisit to the part of Germany I once lived in and brought back a lot of memories of the endless NATO exercises that were so exciting to watch when I was a kid.
Did I truly pass this on my way to Zurich? If so, whereabouts? And who is it? Get a clear sight and tell me how you solved the#Fridayquiz.
\n
\n
The photo is clearly taken from a passing vehicle. Fiete lives in Hamburg and so to travel to Zurich he’d have to go by train and travel North to South through the centre of Germany. The reflection in the window also suggests a train rather than a coach or car.
\n
The key detail to solve this challenge is in the middle of the picture. Here’s an enhanced version:
\n
\n
This close-up shows three helicopters. The one in the middle has the distinctive shape of a Chinook CH-47 helicopter, with rotors at the front and back. This is what one looks like close up:
\n
\n
So this must be an easy challenge to solve, right? All we have to do is find where the German Air Force bases its CH-47 Chinooks, confirm the base is next to a railway line, and then we will find the location. There is only one problem with this: Wikipedia tells us that the German military do not operate Chinook helicopters. So either Fiete was not in Germany, or it is necessary to do a little more digging.
\n
Planespotters To The Rescue
\n
Geolocating military aircraft can pose a lot of problems. The most obvious one being that they are movable objects, and can travel anywhere at any time. Just because one is photographed in one location does not mean it has always been there. Secondly while there are a lot of resources for tracking civil aviation, there are not really any comparable resources for tracking military flights. Or are there?
\n
One of the least recognised sources of OSINT for military aircraft is the dedicated community of amateur aircraft enthusiasts. Sites like scramble.nl and Fighter Control allow planespotters to post photos, serial numbers, and movement information about aircraft they’ve been to see. NOTAMs (flying area restrictions) can also provide insights into upcoming military air movements and large-scale exercises. This site has diagrams of most major combat aircraft to help with recognition.
I should point out at this stage that an alternative method could have been to check the list of all the air force bases in Germany, find one next to a railway line, and then work from there. This would have almost certainly have left Wunstorf Air Base as the only option, but it just doesn’t match the picture. The runway and perimeter track is too far from the railway line for starters.
\nWunstorf was not the right location.\n
I decided to use Scramble to learn about recent activity at German military airbases. There are actually not very many active military airfields in Germany these days so it didn’t take too long to go through them. In the thread about the base at Celle, I found this snippet of information:
\n
\n
Bennie tells us that a support flight number NAF21 is flying to Celle from Eindhoven to support the Dutch helicopters that are there. Another quick piece of Wikipedia research also shows that the Dutch Air Force also fly CH-47 Chinooks just like the ones in Fiete’s original photo. This means that there is a high possibility that Dutch Air Force CH-47s were at Celle last month, but a little extra detail was needed before I could be certain.
\n
If I’d tried the method of just looking for German Air Force bases on Google Maps to try and find a match, I’d never have come across this kind of information. This is also partly due to the fact that Celle was not a German Air Force base. It is a base for the Germany Army Aviation Corps, which is separate from the Air Force. A small detail perhaps, but easily overlooked.
\n
But back to the Dutch Air Force link. Was there anything else to learn? An initial Google search for “dutch ch47 + germany” with time parameters set to the last 12 months brought back this article about Exercise Green Griffin, a joint Dutch-German military exercise that took place in Lower Saxony. Celle is in Lower Saxony, so that helped to confirm the theory that the helicopters in Fiete’s picture could be at Celle.
\n
I focused on searches for Exercise Green Griffin and found lots of really useful information. This article from a local news site provides a lot of extra detail. It even says that there were 10 NH90s, 6 Tigers, 6 AH-64 Apaches, 5 CH-47 Chinooks, and a single C-130 Hercules present at Celle. The article also includes this picture of the helicopters taking off:
\nApocalypse Niedersachsen\n
For more confirmation overkill, there’s even a full-length video about the training exercise. The description states it took place in the Celle-Bergen-Nienburg area, which once again provides additional confirmation about the location. The footage starts with the helicopters taking off and landing.
\n\n
There is so much information in this case that it is easy to overlook more traditional means of confirmation. All the aircraft information indicates Celle was the location, but there was one final way to confirm with a quick Google Earth comparison. Below is a picture of the railway line running along the edge of the base at Celle. The red box corresponds roughly with the field of view in Fiete’s original photo.
\n
\n
\n
\n
Note the building just appearing on the left, the junction on the perimeter track and the position of the trees. This final wider view looks towards where the helicopters were parked:
",
+ "content_text": "When I was a child I spent a lot of time living around various military establishments in Germany. In the 1980s the north German plain was packed with NATO soldiers ready to fight the Russians in case they ever rolled over the East German border (remember when that was a thing?). Nowadays the Russians prefer to use spearphishing and poisoned door handles instead of tanks, so there aren’t as many soldiers in Germany as there once were. Today’s Quiztime challenge was a nice revisit to the part of Germany I once lived in and brought back a lot of memories of the endless NATO exercises that were so exciting to watch when I was a kid.\n\nThe Challenge\n\nFiete Stegers posted this challenge photo and asked:\n\nDid I truly pass this on my way to Zurich? If so, whereabouts? And who is it? Get a clear sight and tell me how you solved the #Fridayquiz.\n\nThe photo is clearly taken from a passing vehicle. Fiete lives in Hamburg and so to travel to Zurich he’d have to go by train and travel North to South through the centre of Germany. The reflection in the window also suggests a train rather than a coach or car.\n\nThe key detail to solve this challenge is in the middle of the picture. Here’s an enhanced version:\n\nThis close-up shows three helicopters. The one in the middle has the distinctive shape of a Chinook CH-47 helicopter, with rotors at the front and back. This is what one looks like close up:\n\nSo this must be an easy challenge to solve, right? All we have to do is find where the German Air Force bases its CH-47 Chinooks, confirm the base is next to a railway line, and then we will find the location. There is only one problem with this: Wikipedia tells us that the German military do not operate Chinook helicopters. So either Fiete was not in Germany, or it is necessary to do a little more digging.\n\nPlanespotters To The Rescue\n\nGeolocating military aircraft can pose a lot of problems. The most obvious one being that they are movable objects, and can travel anywhere at any time. Just because one is photographed in one location does not mean it has always been there. Secondly while there are a lot of resources for tracking civil aviation, there are not really any comparable resources for tracking military flights. Or are there?\n\nOne of the least recognised sources of OSINT for military aircraft is the dedicated community of amateur aircraft enthusiasts. Sites like scramble.nl and Fighter Control allow planespotters to post photos, serial numbers, and movement information about aircraft they’ve been to see. NOTAMs (flying area restrictions) can also provide insights into upcoming military air movements and large-scale exercises. This site has diagrams of most major combat aircraft to help with recognition.\n\nScramble has a forum dedicated to spotters who record the movements of military aircraft in Germany. Each airfield has its own thread, so it is possible to see all the popular base locations and records of aircraft movements in one place. Not every single flight is recorded of course, but it offers a lot useful insights.\n\nI should point out at this stage that an alternative method could have been to check the list of all the air force bases in Germany, find one next to a railway line, and then work from there. This would have almost certainly have left Wunstorf Air Base as the only option, but it just doesn’t match the picture. The runway and perimeter track is too far from the railway line for starters.\n\nWunstorf was not the right location.\n\nI decided to use Scramble to learn about recent activity at German military airbases. There are actually not very many active military airfields in Germany these days so it didn’t take too long to go through them. In the thread about the base at Celle, I found this snippet of information:\n\nBennie tells us that a support flight number NAF21 is flying to Celle from Eindhoven to support the Dutch helicopters that are there. Another quick piece of Wikipedia research also shows that the Dutch Air Force also fly CH-47 Chinooks just like the ones in Fiete’s original photo. This means that there is a high possibility that Dutch Air Force CH-47s were at Celle last month, but a little extra detail was needed before I could be certain.\n\nIf I’d tried the method of just looking for German Air Force bases on Google Maps to try and find a match, I’d never have come across this kind of information. This is also partly due to the fact that Celle was not a German Air Force base. It is a base for the Germany Army Aviation Corps, which is separate from the Air Force. A small detail perhaps, but easily overlooked.\n\nBut back to the Dutch Air Force link. Was there anything else to learn? An initial Google search for “dutch ch47 + germany” with time parameters set to the last 12 months brought back this article about Exercise Green Griffin, a joint Dutch-German military exercise that took place in Lower Saxony. Celle is in Lower Saxony, so that helped to confirm the theory that the helicopters in Fiete’s picture could be at Celle.\n\nI focused on searches for Exercise Green Griffin and found lots of really useful information. This article from a local news site provides a lot of extra detail. It even says that there were 10 NH90s, 6 Tigers, 6 AH-64 Apaches, 5 CH-47 Chinooks, and a single C-130 Hercules present at Celle. The article also includes this picture of the helicopters taking off:\n\nApocalypse Niedersachsen\n\nFor more confirmation overkill, there’s even a full-length video about the training exercise. The description states it took place in the Celle-Bergen-Nienburg area, which once again provides additional confirmation about the location. The footage starts with the helicopters taking off and landing.\n\nThere is so much information in this case that it is easy to overlook more traditional means of confirmation. All the aircraft information indicates Celle was the location, but there was one final way to confirm with a quick Google Earth comparison. Below is a picture of the railway line running along the edge of the base at Celle. The red box corresponds roughly with the field of view in Fiete’s original photo.\n\nNote the building just appearing on the left, the junction on the perimeter track and the position of the trees. This final wider view looks towards where the helicopters were parked:\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Selection_001-2-1024x743.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Selection_003-3.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Selection_004-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Selection_005.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Selection_006-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Selection_007-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Selection_008-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/06/Selection_009-1.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2019-07-05_quiztime-5th-july-lifeboats-and-fast-reverse-image-searching.json b/nixintel_geolocation_json/2019-07-05_quiztime-5th-july-lifeboats-and-fast-reverse-image-searching.json
new file mode 100644
index 0000000000000000000000000000000000000000..bf4fc6aa5f266bf271bd90ac8490f033a82010b9
--- /dev/null
+++ b/nixintel_geolocation_json/2019-07-05_quiztime-5th-july-lifeboats-and-fast-reverse-image-searching.json
@@ -0,0 +1,83 @@
+{
+ "slug": "quiztime-5th-july-lifeboats-and-fast-reverse-image-searching",
+ "title": "Quiztime 5th July – Lifeboats and Fast Reverse Image Searching",
+ "url": "https://nixintel.info/osint/quiztime-5th-july-lifeboats-and-fast-reverse-image-searching/",
+ "date": "2019-07-05T19:50:49+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
Rescue boats are a common theme in this week’s Quiztime. On Tuesday Lars Wienand asked us to locate a rescue boat and today Philipp Dudek asks another Lifeboat related question. This was quite a fun quiz to solve but after solving it once I realised there was another solution which I think would allow you to solve the puzzle in about ten seconds. The first solution relies on researching the subject matter in a bit more depth, whereas the quick-fire solution relies on a very precise reverse image search. I’ll post both solutions here.
\n
Which Boat Is This?
\n
Philipp asks us to identify the boat in this picture:
\n
\n
The picture is from a publication about the social media activities of the DGzRS, the German Maritime Rescue Service. It’s just about possible make out the URL seenotretter.de.
\n
There is not enough detail visible to make out anything as distinctive as the boat’s name (as if we would be that lucky). The first place I looked at was the DGzRS official website. It contains a very handy guide to all the boats currently in the fleet.
\n
\n
Unless you already know a lot about boats most of them all look pretty similar and it’s hard to tell them apart. However with a little attention to detail it is possible to get enough information from the original photograph to positively identify the right kind of boat.
\n
\n
We can tell the boat’s bridge has three windows at the front, with another single window on each side. It has a flagpole on the prow, a white mast behind the bridge, and an orange-brown trim along the edge of the deck. It was possible to go through the different types of boat used by the DGzRS and look for these matching features. This boat looked very similar, but attention to detail shows it cannot be right:
\n
\n
\n
The boat we are looking for is unusual among the DGzRS fleet because it doesn’t have any orange paint across the front of the bridge. I was puzzled by this at first as every boat in the fleet appeared to be in the same colour scheme. Was the boat in the image in an older colour scheme, in which case we might be searching for an old out of service boat? To make sure, a quick check of the DGzRS social media accounts referred to in the original image helped to date the article posted by Philipp. The article referenced the organisation’s YouTube, Twitter, and Instagram accounts. A quick check showed that the Twitter and YouTube accounts were created in 2009, but the first Instagram post was not until May 2015. This means that the original article must have been posted on or after that date, so the paint scheme on the boat can’t be that old after all.
\n
So I continued to look through the pictures of the fleet and found a possible match:
\n
\n
\n
This boat matches the shape and features of the boat we are looking for and unlike all the other similar boats in the fleet, it has no orange paint on the front part of the bridge. There’s only one slight problem though, the website says “Anzahl:6” , which means there are six ships of this kind in service. How to find the exact one?
\n
The method to do this was fairly simple. Each boat is allocated to its own port, so by using Google Maps and/or Google Earth I hoped to look at each location and verify it was the correct one by matching it to the original image.
\n
\n
I opened up Google Maps and searched for DGzRS Nordstrand. The location didn’t obviously match, so I moved on. When I searched for DGzRS Busum, a familiar looking image appeared along with the map:
\n
\n
This cannot be the same boat because of the paint scheme, but the wharf it is tied to is identical to the one in the original image. This means that Busum is the correct location, and so the correct boat must be Theodor Storm because it is the only boat of that type based there. A quick image search confirms this is the case:
\n
\n
Same paint scheme, same boat, same location.
\n
\n
An aerial view of the location shows the background features also match, i.e. the original photo has a harbour in the background and not the open sea.
\n
\n
So the correct answer is that the boat is the Theodor Storm.
\n
The 10 Second Method
\n
There was a super fast method to solve this that took all of about ten seconds. I’ve described it here to show the technique but in real-life OSINT speed is not everything. It’s rare to be really fast and really accurate, and accuracy is more important. Better to be slow and right then fast and wrong. Whatever real-life OSINT question you’re asking, your answer probably needs to be more than just a the name of a boat and nothing else. Being methodical and researching widely will always give better and more detailed answers than instant hits. That said, sometimes solving things almost instantly can be fun too.
\n
I’ve mentioned Search By Image in a previous blog post, and I’ve also made a short video showing it in action. It’s ideal for a task like this because it lets you focus on one particular part of an image and just search on that particular detail. Here’s the ten-second method to identify the boat.
\n
1. Use the capture feature of Search By Image to select just the boat from the original image.
\n
\n
\n
2. Choose “All Search Engines” and click search. Go to Yandex first and see a familiar looking boat in a familiar location:
\n
\n
\n
3. Click on it. Yandex tells you this is a picture of the Theodor Storm. Easy-peasy!
\n
\n
\n
If only this Quiztime challenge was quite so straightforward! (I’ll post my solution if I ever solve it…)
",
+ "content_text": "Rescue boats are a common theme in this week’s Quiztime. On Tuesday Lars Wienand asked us to locate a rescue boat and today Philipp Dudek asks another Lifeboat related question. This was quite a fun quiz to solve but after solving it once I realised there was another solution which I think would allow you to solve the puzzle in about ten seconds. The first solution relies on researching the subject matter in a bit more depth, whereas the quick-fire solution relies on a very precise reverse image search. I’ll post both solutions here.\n\nWhich Boat Is This?\n\nPhilipp asks us to identify the boat in this picture:\n\nThe picture is from a publication about the social media activities of the DGzRS, the German Maritime Rescue Service. It’s just about possible make out the URL seenotretter.de.\n\nThere is not enough detail visible to make out anything as distinctive as the boat’s name (as if we would be that lucky). The first place I looked at was the DGzRS official website. It contains a very handy guide to all the boats currently in the fleet.\n\nUnless you already know a lot about boats most of them all look pretty similar and it’s hard to tell them apart. However with a little attention to detail it is possible to get enough information from the original photograph to positively identify the right kind of boat.\n\nWe can tell the boat’s bridge has three windows at the front, with another single window on each side. It has a flagpole on the prow, a white mast behind the bridge, and an orange-brown trim along the edge of the deck. It was possible to go through the different types of boat used by the DGzRS and look for these matching features. This boat looked very similar, but attention to detail shows it cannot be right:\n\nThe boat we are looking for is unusual among the DGzRS fleet because it doesn’t have any orange paint across the front of the bridge. I was puzzled by this at first as every boat in the fleet appeared to be in the same colour scheme. Was the boat in the image in an older colour scheme, in which case we might be searching for an old out of service boat? To make sure, a quick check of the DGzRS social media accounts referred to in the original image helped to date the article posted by Philipp. The article referenced the organisation’s YouTube, Twitter, and Instagram accounts. A quick check showed that the Twitter and YouTube accounts were created in 2009, but the first Instagram post was not until May 2015. This means that the original article must have been posted on or after that date, so the paint scheme on the boat can’t be that old after all.\n\nSo I continued to look through the pictures of the fleet and found a possible match:\n\nThis boat matches the shape and features of the boat we are looking for and unlike all the other similar boats in the fleet, it has no orange paint on the front part of the bridge. There’s only one slight problem though, the website says “Anzahl:6” , which means there are six ships of this kind in service. How to find the exact one?\n\nThe method to do this was fairly simple. Each boat is allocated to its own port, so by using Google Maps and/or Google Earth I hoped to look at each location and verify it was the correct one by matching it to the original image.\n\nI opened up Google Maps and searched for DGzRS Nordstrand. The location didn’t obviously match, so I moved on. When I searched for DGzRS Busum, a familiar looking image appeared along with the map:\n\nThis cannot be the same boat because of the paint scheme, but the wharf it is tied to is identical to the one in the original image. This means that Busum is the correct location, and so the correct boat must be Theodor Storm because it is the only boat of that type based there. A quick image search confirms this is the case:\n\nSame paint scheme, same boat, same location.\n\nAn aerial view of the location shows the background features also match, i.e. the original photo has a harbour in the background and not the open sea.\n\nSo the correct answer is that the boat is the Theodor Storm.\n\nThe 10 Second Method\n\nThere was a super fast method to solve this that took all of about ten seconds. I’ve described it here to show the technique but in real-life OSINT speed is not everything. It’s rare to be really fast and really accurate, and accuracy is more important. Better to be slow and right then fast and wrong. Whatever real-life OSINT question you’re asking, your answer probably needs to be more than just a the name of a boat and nothing else. Being methodical and researching widely will always give better and more detailed answers than instant hits. That said, sometimes solving things almost instantly can be fun too.\n\nI’ve mentioned Search By Image in a previous blog post, and I’ve also made a short video showing it in action. It’s ideal for a task like this because it lets you focus on one particular part of an image and just search on that particular detail. Here’s the ten-second method to identify the boat.\n\n1. Use the capture feature of Search By Image to select just the boat from the original image.\n\n2. Choose “All Search Engines” and click search. Go to Yandex first and see a familiar looking boat in a familiar location:\n\n3. Click on it. Yandex tells you this is a picture of the Theodor Storm. Easy-peasy!\n\nIf only this Quiztime challenge was quite so straightforward! (I’ll post my solution if I ever solve it…)\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/brochure.jpglarge-768x1024.jpeg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_003-1024x604.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_010.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_010_box.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_006_box.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_004-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Screenshot-2019-07-05-at-19.49.47.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Screenshot-2019-07-05-at-19.54.49-1024x572.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/D-uItj5XkAEw1C5.jpglarge.jpeg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_005-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_002.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/D-t32d6XkAADqd1.jpglarge-JPEG-Image-1536-×-2048-pixels-Mozilla-Firefox_007-1024x562.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_008-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Yandex.Images-search-for-similar-images-Mozilla-Firefox_009-1024x562.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2019-07-17_quiztime-16th-july-start-big-get-small-how-to-be-an-osint-sherlock.json b/nixintel_geolocation_json/2019-07-17_quiztime-16th-july-start-big-get-small-how-to-be-an-osint-sherlock.json
new file mode 100644
index 0000000000000000000000000000000000000000..ce436381fe0dde8cb07970aad70809982210adea
--- /dev/null
+++ b/nixintel_geolocation_json/2019-07-17_quiztime-16th-july-start-big-get-small-how-to-be-an-osint-sherlock.json
@@ -0,0 +1,63 @@
+{
+ "slug": "quiztime-16th-july-start-big-get-small-how-to-be-an-osint-sherlock",
+ "title": "Quiztime 16th July – Start Big, Get Small – How To Be An OSINT Sherlock",
+ "url": "https://nixintel.info/osint/quiztime-16th-july-start-big-get-small-how-to-be-an-osint-sherlock/",
+ "date": "2019-07-17T20:53:41+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
“Once you eliminate the impossible, whatever remains, no matter how improbable, must be the truth.”
\n
Sherlock Holmes
\n
\nI’m not usually a big fan of Sherlock Holmes. Like many other fictional detectives, I don’t think he’s a particularly effective investigator. Holmes, Columbo, Morse, and countless others usually rely on evidence that is highly circumstantial and even then they usually need the villain to incriminate himself by making a full confession anyway. \n
\nCombined total number of Quiztime challenges solved: zero.\n
\nThat aside, Sherlock makes a good point about eliminating the impossible that can easily be applied to OSINT investigations. Even small-scale OSINT enquiries can generate a lot of information that has to be carefully sifted and eliminated to arrive at the correct answer. By choosing the right parameters and applying them to the data, it is possible to use logic to eliminate the incorrect answers and arrive at the truth.
\n
\n
\n
\nA really good recent example is this notorious Quiztime challenge (pictured above) that Sector035 managed to solve yesterday. The task was to identify which school the two girls in the picture attended. The car registrations showed that the location was in Brazil, but trying to find a single school in Brazil is nearly impossible since there are thousands. However by using just two parameters to filter the data, Sector035 was able to reduce the number of possible locations to a much smaller number and eventually he found the right answer.
\n
\nBy assuming that the location was probably in Recife (based on @y_vdw’s posting history) and that it was on a four-lane highway, it was possible to find the right location by searching all the four lane highways in Recife for a visual match. This was still a relatively time-consuming exercise, but by starting from the whole of Brazil and eliminating all locations that were not Recife, and all roads that were not four lane highways, it was possible to find the exact spot. Start big, get small, and find the right location.
\n \n \n1) Can you name the four cruise ships I saw on the day I took the photos? \n2) One cruise ship can be seen there for several months. Why? \n \nMost of the quiz could actually be solved with just the first photo. A cruise ship is visible in the background, but the train provides the most useful information to begin working with:
\n
\n1) The DB (Deutsches Bundesbahn) logo shows this is in Germany. \n2) The train is at a station next to a location where cruise ships dock.
\nSo basically just here somewhere…\n
\nGermany only has a small coastline relative to its size, but there are still a lot of possible locations where large ships dock, and almost all of them have railway stations. Looking at a map the possible locations were: Emden, Norden, Wilhelmshaven, Bremerhaven, Bremen, Cuxhaven, Flensburg, Hamburg, Kiel, Luebeck, Wismar, Rostock, and Stralsund. There were also a few other ports that I might have missed. In theory it would have been possible to just look at Google Earth to see if any of the port locations resemble the one in the photograph, but this would be quite a slow and ineffective method, not to mention quite a boring way to spend an evening. The pool of potential locations was quite big, so how could I do some Sherlock Holmes-style elimination to find the correct location?
\n
\nThe answer was to revisit the photograph and extract a little more information to set some new parameters. I couldn’t read the destination on the front of the train (as if Lars would make it that easy…) so instead I searched for the train’s serial number 442338 to find out more about it. Searching with Google for “zug + 442338” (zug = train) brought up this snippet of information: \n
\n
\n
\nThis tells us that train number 443 338 was introduced to run services between Rostock and Gustow. Now I knew if this was correct, it would mean that I could eliminate all the other locations immediately. To check I looked at Rostock with Google Earth: \n
\n
\n
\n
This a match for the location in the photograph. The train platforms are at the correct angle in relation to the water so that the ships would appear as they do in Lars’ photos, and a check with Google Maps showed that the large building next to the water is a cruise ship terminal. So just applying a single parameter (the train number) to the data meant that all the incorrect locations could be eliminated, and the big pool of possible answers was narrowed down to just one. The correct location is Rostock-Warnemunde station.
\n
Which Ships Were At Rostock?
\n
\n
So now to answer the rest of Lars’ questions. There were a few ways this could have been solved. One method could have been to try and identify the ships in the photographs and then follow their movements. For example the ship in the second photo clearly belongs to the Aida cruise company. In theory I could have looked at their fleet of ships and then worked out the identity of the vessel in the picture by looking at funnel positions and counting windows, and then looking at when the boat was last in Rostock. In some cases this would be appropriate technique but it would also be time-consuming and there was a quicker way to do this in the circumstances. \n
\n
No need to spend too much time looking at these. This is fortunate because at least four of the ships are physically identical. \n \nThere’s also the small matter of when the photo was taken. Lars does not specify the date, but it is necessary to solve the quiz. After all, how can we know which ships Lars saw if we do not even know when he was there? Sherlock Holmes can help again here. In theory there were an unlimited number of days on which Lars could have gone to Rostock, but by applying a few parameters it was possible to eliminate all the incorrect dates and find the correct one. Start big, get small. \n \nThe Port of Rostock publishes a list of all the past and future arrival dates of cruise ships here. Lars told us that he saw four cruise ships on the day he visited. This meant all the incorrect dates could be eliminated by discounting all dates when fewer than four cruise ships were present. Working through the list, it was clear that Friday 12th July is the only date so far this year when four or more cruise ships came to Rostock, so it must be the correct date (assuming that Lars visited this year). Here are the ships that were present on that date: \n
\n
\n
\nThis also means it was possible to be certain about the ships that were present. So the answer to question 1 was that the ships Lars could have seen were AidAmar, Costa Favolosa, Marina, Regal Princess, and Viking Sea. Logically it is not possible that Lars could have visited on any other day and seen the same number of cruise ships.
\n
\nVesselFinder also confirmed that the ships had all been present in Rostock on 12th July.\n
\nA search of all the ship names in Google News also brought up this article from the local press, which confirmed the visit of five ships, the names of the ships, and the fact that this is an unusual occurrence for Rostock. There’s even a YouTube video here.
\n
\n
Which Ship Hasn’t Moved For Months?
\n
\n
The second part of the quiz was a little more tricky, but was solvable with a bit of effort. Firstly I decided to make use VesselFinder to see if there were still any cruise ships there: \n
\n
\n
\nThere weren’t any, which meant that any cruise ships that had been there for months had either left since Lars was there on Friday, or they were not transmitting their location. If the ship had left, then surely it would have shown up on the Port of Rostock website? The more important question to ask was – why would a cruise ship be in a port for months and not go anywhere? Cruise ships are expensive assets and need to be working to make money. The only reasons I could think of that a cruise ship wouldn’t move for months would be because a) it is in need of repair, b) it is being scrapped, or c) it hasn’t been completed yet.
",
+ "content_text": "“Once you eliminate the impossible, whatever remains, no matter how improbable, must be the truth.”\n\nSherlock Holmes\n\nI’m not usually a big fan of Sherlock Holmes. Like many other fictional detectives, I don’t think he’s a particularly effective investigator. Holmes, Columbo, Morse, and countless others usually rely on evidence that is highly circumstantial and even then they usually need the villain to incriminate himself by making a full confession anyway.\n\nCombined total number of Quiztime challenges solved: zero.\n\nThat aside, Sherlock makes a good point about eliminating the impossible that can easily be applied to OSINT investigations. Even small-scale OSINT enquiries can generate a lot of information that has to be carefully sifted and eliminated to arrive at the correct answer. By choosing the right parameters and applying them to the data, it is possible to use logic to eliminate the incorrect answers and arrive at the truth.\n\nA really good recent example is this notorious Quiztime challenge (pictured above) that Sector035 managed to solve yesterday. The task was to identify which school the two girls in the picture attended. The car registrations showed that the location was in Brazil, but trying to find a single school in Brazil is nearly impossible since there are thousands. However by using just two parameters to filter the data, Sector035 was able to reduce the number of possible locations to a much smaller number and eventually he found the right answer.\n\nBy assuming that the location was probably in Recife (based on @y_vdw’s posting history) and that it was on a four-lane highway, it was possible to find the right location by searching all the four lane highways in Recife for a visual match. This was still a relatively time-consuming exercise, but by starting from the whole of Brazil and eliminating all locations that were not Recife, and all roads that were not four lane highways, it was possible to find the exact spot. Start big, get small, and find the right location.\n\nWhere Was Lars Wienand?\n\nIt was possible to use the same Sherlock Holmes method to solve yesterday’s Quiztime challenge. Lars Wienand posted these two images and asked:\n\n1) Can you name the four cruise ships I saw on the day I took the photos?\n\n2) One cruise ship can be seen there for several months. Why?\n\nMost of the quiz could actually be solved with just the first photo. A cruise ship is visible in the background, but the train provides the most useful information to begin working with:\n\n1) The DB (Deutsches Bundesbahn) logo shows this is in Germany.\n\n2) The train is at a station next to a location where cruise ships dock.\n\nSo basically just here somewhere…\n\nGermany only has a small coastline relative to its size, but there are still a lot of possible locations where large ships dock, and almost all of them have railway stations. Looking at a map the possible locations were: Emden, Norden, Wilhelmshaven, Bremerhaven, Bremen, Cuxhaven, Flensburg, Hamburg, Kiel, Luebeck, Wismar, Rostock, and Stralsund. There were also a few other ports that I might have missed. In theory it would have been possible to just look at Google Earth to see if any of the port locations resemble the one in the photograph, but this would be quite a slow and ineffective method, not to mention quite a boring way to spend an evening. The pool of potential locations was quite big, so how could I do some Sherlock Holmes-style elimination to find the correct location?\n\nThe answer was to revisit the photograph and extract a little more information to set some new parameters. I couldn’t read the destination on the front of the train (as if Lars would make it that easy…) so instead I searched for the train’s serial number 442338 to find out more about it. Searching with Google for “zug + 442338” (zug = train) brought up this snippet of information:\n\nThis tells us that train number 443 338 was introduced to run services between Rostock and Gustow. Now I knew if this was correct, it would mean that I could eliminate all the other locations immediately. To check I looked at Rostock with Google Earth:\n\nThis a match for the location in the photograph. The train platforms are at the correct angle in relation to the water so that the ships would appear as they do in Lars’ photos, and a check with Google Maps showed that the large building next to the water is a cruise ship terminal. So just applying a single parameter (the train number) to the data meant that all the incorrect locations could be eliminated, and the big pool of possible answers was narrowed down to just one. The correct location is Rostock-Warnemunde station.\n\nWhich Ships Were At Rostock?\n\nSo now to answer the rest of Lars’ questions. There were a few ways this could have been solved. One method could have been to try and identify the ships in the photographs and then follow their movements. For example the ship in the second photo clearly belongs to the Aida cruise company. In theory I could have looked at their fleet of ships and then worked out the identity of the vessel in the picture by looking at funnel positions and counting windows, and then looking at when the boat was last in Rostock. In some cases this would be appropriate technique but it would also be time-consuming and there was a quicker way to do this in the circumstances.\n\nNo need to spend too much time looking at these. This is fortunate because at least four of the ships are physically identical.\n\nThere’s also the small matter of when the photo was taken. Lars does not specify the date, but it is necessary to solve the quiz. After all, how can we know which ships Lars saw if we do not even know when he was there? Sherlock Holmes can help again here. In theory there were an unlimited number of days on which Lars could have gone to Rostock, but by applying a few parameters it was possible to eliminate all the incorrect dates and find the correct one. Start big, get small.\n\nThe Port of Rostock publishes a list of all the past and future arrival dates of cruise ships here. Lars told us that he saw four cruise ships on the day he visited. This meant all the incorrect dates could be eliminated by discounting all dates when fewer than four cruise ships were present. Working through the list, it was clear that Friday 12th July is the only date so far this year when four or more cruise ships came to Rostock, so it must be the correct date (assuming that Lars visited this year). Here are the ships that were present on that date:\n\nThis also means it was possible to be certain about the ships that were present. So the answer to question 1 was that the ships Lars could have seen were AidAmar, Costa Favolosa, Marina, Regal Princess, and Viking Sea. Logically it is not possible that Lars could have visited on any other day and seen the same number of cruise ships.\n\nVesselFinder also confirmed that the ships had all been present in Rostock on 12th July.\n\nA search of all the ship names in Google News also brought up this article from the local press, which confirmed the visit of five ships, the names of the ships, and the fact that this is an unusual occurrence for Rostock. There’s even a YouTube video here.\n\nWhich Ship Hasn’t Moved For Months?\n\nThe second part of the quiz was a little more tricky, but was solvable with a bit of effort. Firstly I decided to make use VesselFinder to see if there were still any cruise ships there:\n\nThere weren’t any, which meant that any cruise ships that had been there for months had either left since Lars was there on Friday, or they were not transmitting their location. If the ship had left, then surely it would have shown up on the Port of Rostock website? The more important question to ask was – why would a cruise ship be in a port for months and not go anywhere? Cruise ships are expensive assets and need to be working to make money. The only reasons I could think of that a cruise ship wouldn’t move for months would be because a) it is in need of repair, b) it is being scrapped, or c) it hasn’t been completed yet.\n\nGoogling for ship repair at Rostock brought up the company website of MV Werften, who have a large shipbuilding facility in the port. Their website shows that they are currently building a new cruise ship called “Global One”. Work started there in 2018 and the ship will not leave until 2020 – which explains how a cruise ship can be at Rostock for so long without moving!\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/holmes.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/D-koskVWsAcCcDt.jpglarge-1024x768.jpeg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Screenshot-2019-07-17-at-19.44.39.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_031-1024x290.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_027a.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_030.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_027-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_026.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_029.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/07/Selection_028.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2019-08-11_two-different-approaches-to-photo-geolocation.json b/nixintel_geolocation_json/2019-08-11_two-different-approaches-to-photo-geolocation.json
new file mode 100644
index 0000000000000000000000000000000000000000..39c1d639cea85615bb91612347544cfa548931d4
--- /dev/null
+++ b/nixintel_geolocation_json/2019-08-11_two-different-approaches-to-photo-geolocation.json
@@ -0,0 +1,48 @@
+{
+ "slug": "two-different-approaches-to-photo-geolocation",
+ "title": "Two Different Approaches To Photo-Geolocation",
+ "url": "https://nixintel.info/osint/two-different-approaches-to-photo-geolocation/",
+ "date": "2019-08-11T10:39:17+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
Holidays and work commitments mean that I haven’t had much time to tackle the daily Quiztime challenges recently, but participants in two of this week’s challenges from Sector035 and Fiete Stegers highlighted a couple of different approaches to photo geolocation that I wanted to write about in this blogpost. The two photos were very different in terms of the kind of information that they contained, and required two different methods to successfully geolocate the images. What worked for one image did not work for the other, and only participants who applied the right method to the right image (whether consciously or not) were successful while those who applied the wrong method were frustrated and unsuccessful. This post looks at two different methods for making quick progress with a photogeolocation challenge, and how to identify when to apply the right one.
There are two basic ways to approach this image. You can either search by visual similarity, or by searching based on objects in the photograph. Searching by visual similarity essentially involves using reverse image search tools to find matches for all or part of the original image. The dominance of the Metro sign in the centre of the picture makes it hard for reverse image search tools to pick out anything else, so you just end up with a lot of matches that look like this:
\nFrom Yandex reverse image search\n
It’s not a bad match, and it’s enough to tell you that this was taken in Paris (although you could figure that out from the original photo anyway), but there isn’t enough to give the location by a visual matching.
\nNot even close Google…\n
The successful particpants didn’t seem to rely on this method. Object-based searching is a much better way to solve this challenge. The reason is that there is a wealth of searchable information such as shop and street signs in the photo that are enough to provide the correct location in a way that searches that rely on visual similarity cannot.
\n
Consider the number of identifiable objects in this image:
\n
1) Metro sign
\n
2) Royal *lly bar
\n
3) Sign that says “Boulanger / Patissier”
\n
4) Loto signs x2
\n
5) Partly obscured street sign that says *tgallet
\n
Even if you don’t know Paris, 3) tells you the location is French-speaking, 1) tells you this likely Paris (there are five other French cities that have a Metro, but googling their logos shows that only Paris has Metro signs like the one in the image.) Searching for “rue paris” + gallet with Google lead to the autocomplete suggesting “Montgallet” as the location.
\n
\n
\n
A quick check of the location with Streetview confirms that this Montgallet Metro Station, Paris:
\n
\n
Where an image provides so many specific and indentifiable objects to search with, object-based searching is a far more effective technique to geolocate an image quickly when compared to searching for visual similarities.
\n
Finding Fiete In Slovenia – Visual Similarity
\n
Fiete Steger’s Friday photo challenge was a completely different type of image, but it serves as a good example of why the same method that works for one photo will not work for another. Here’s the image in question:
\n
\n
With Sector035’s challenge, there were so many identifiable objects in the image that object-based searching was a very effective way to find where the picture was taken. Could the same method be used to find where Fiete took this photo? Several participants tried but it didn’t work, and there is a good reason why. There are very few features in the photo that are specific and searchable, and so searching for items that feature in the image will produce a list of results that is simply too big to be of any use. Consider the identifiable features in the photo:
\n
1) River
\n
2) Hills
\n
3) Red-roofed houses
\n
4) Salt pans
\n
5) Estuary
\n
There is nothing specific or distinctive enough about these objects to make a search based on this method successful, as participants who chose this technique found. While Sector035’s photo was rich in specific searchable detail, Fiete’s was the complete opposite. In this case the lack of specific objects meant a visual search was much more effective.
\n
Using Search By Image, I cropped a central area of the photo, removing some of the sky and foreground detail:
\n
\n
And then a quick reverse image search with Yandex brought up the location match near Piranski in Slovenia. Searching by visual similarity is much more effective for this kind of geolocation.
\n
For most types of geolocation problems it will be necessary to use a combination of techniques to find the answer, and it is not a question of either/or when it comes to the two methods described in this post. What is important to remember is that automatically using the same technique for every image will not always work, and each image needs to be assessed based on its own content – what am I actually looking at? How much detail can I actually see? Once you answer this question, it is possible to make a better decision about what kind og approach to take.
",
+ "content_text": "Holidays and work commitments mean that I haven’t had much time to tackle the daily Quiztime challenges recently, but participants in two of this week’s challenges from Sector035 and Fiete Stegers highlighted a couple of different approaches to photo geolocation that I wanted to write about in this blogpost. The two photos were very different in terms of the kind of information that they contained, and required two different methods to successfully geolocate the images. What worked for one image did not work for the other, and only participants who applied the right method to the right image (whether consciously or not) were successful while those who applied the wrong method were frustrated and unsuccessful. This post looks at two different methods for making quick progress with a photogeolocation challenge, and how to identify when to apply the right one.\n\nObject Search v Image Similarity\n\nLet’s start with Sector035’s photo first:\n\nThere are two basic ways to approach this image. You can either search by visual similarity, or by searching based on objects in the photograph. Searching by visual similarity essentially involves using reverse image search tools to find matches for all or part of the original image. The dominance of the Metro sign in the centre of the picture makes it hard for reverse image search tools to pick out anything else, so you just end up with a lot of matches that look like this:\n\nFrom Yandex reverse image search\n\nIt’s not a bad match, and it’s enough to tell you that this was taken in Paris (although you could figure that out from the original photo anyway), but there isn’t enough to give the location by a visual matching.\n\nNot even close Google…\n\nThe successful particpants didn’t seem to rely on this method. Object-based searching is a much better way to solve this challenge. The reason is that there is a wealth of searchable information such as shop and street signs in the photo that are enough to provide the correct location in a way that searches that rely on visual similarity cannot.\n\nConsider the number of identifiable objects in this image:\n\n1) Metro sign\n\n2) Royal *lly bar\n\n3) Sign that says “Boulanger / Patissier”\n\n4) Loto signs x2\n\n5) Partly obscured street sign that says *tgallet\n\nEven if you don’t know Paris, 3) tells you the location is French-speaking, 1) tells you this likely Paris (there are five other French cities that have a Metro, but googling their logos shows that only Paris has Metro signs like the one in the image.) Searching for “rue paris” + gallet with Google lead to the autocomplete suggesting “Montgallet” as the location.\n\nA quick check of the location with Streetview confirms that this Montgallet Metro Station, Paris:\n\nWhere an image provides so many specific and indentifiable objects to search with, object-based searching is a far more effective technique to geolocate an image quickly when compared to searching for visual similarities.\n\nFinding Fiete In Slovenia – Visual Similarity\n\nFiete Steger’s Friday photo challenge was a completely different type of image, but it serves as a good example of why the same method that works for one photo will not work for another. Here’s the image in question:\n\nWith Sector035’s challenge, there were so many identifiable objects in the image that object-based searching was a very effective way to find where the picture was taken. Could the same method be used to find where Fiete took this photo? Several participants tried but it didn’t work, and there is a good reason why. There are very few features in the photo that are specific and searchable, and so searching for items that feature in the image will produce a list of results that is simply too big to be of any use. Consider the identifiable features in the photo:\n\n1) River\n\n2) Hills\n\n3) Red-roofed houses\n\n4) Salt pans\n\n5) Estuary\n\nThere is nothing specific or distinctive enough about these objects to make a search based on this method successful, as participants who chose this technique found. While Sector035’s photo was rich in specific searchable detail, Fiete’s was the complete opposite. In this case the lack of specific objects meant a visual search was much more effective.\n\nUsing Search By Image, I cropped a central area of the photo, removing some of the sky and foreground detail:\n\nAnd then a quick reverse image search with Yandex brought up the location match near Piranski in Slovenia. Searching by visual similarity is much more effective for this kind of geolocation.\n\nFor most types of geolocation problems it will be necessary to use a combination of techniques to find the answer, and it is not a question of either/or when it comes to the two methods described in this post. What is important to remember is that automatically using the same technique for every image will not always work, and each image needs to be assessed based on its own content – what am I actually looking at? How much detail can I actually see? Once you answer this question, it is possible to make a better decision about what kind og approach to take.\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/parismetro-859x1024.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Metro-red-682x1024.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Screenshot_2019-08-11-Google-Search-1024x743.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_030.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_028-1-1024x488.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/slovenia-1024x311.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Screenshot-2019-08-11-edited-at-11.11.32-1024x602.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2019-08-27_quiztime-25th-august-2019-military-vehicles-geolocation.json b/nixintel_geolocation_json/2019-08-27_quiztime-25th-august-2019-military-vehicles-geolocation.json
new file mode 100644
index 0000000000000000000000000000000000000000..b40ace4532821b90b4b8062fe6e569e108867486
--- /dev/null
+++ b/nixintel_geolocation_json/2019-08-27_quiztime-25th-august-2019-military-vehicles-geolocation.json
@@ -0,0 +1,118 @@
+{
+ "slug": "quiztime-25th-august-2019-military-vehicles-geolocation",
+ "title": "Quiztime 25th August 2019 – Military Vehicles & Geolocation",
+ "url": "https://nixintel.info/osint/quiztime-25th-august-2019-military-vehicles-geolocation/",
+ "date": "2019-08-27T20:35:05+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
This Sunday we were treated to an extra Quiztime image location challenge which involved an unidentified military vehicle. Doxsor posted this image and asked where this military vehicle was headed to, and which barracks was closest to the location:
\n
\n
\n
This was a good opportunity to practise military vehicle identification and explore how military units are organised and structured. OSINT is a very wide and varied discipline, but if your work focuses on conflict zones like Libya, Syria, or Ukraine, then being able to identify military vehicles and those that use them can be an integral part of an investigation. Bellingcat’s recent work on MH17 is a prime example of how essential this skill can be.
\n“You ain’t seen me, right…”\n
\n
The rest of this post will look at how I identified the vehicle and unit, before finding the correct location.
\n
Tanks & Armoured Fighting Vehicles
\n
There were a few ways to start of this puzzle, but I began with the number plate on the trailer. 6 digits in black print on a yellow background could be several different countries, but two strong candidates were the Netherlands and also Denmark. (World Licence Plates is a great resource for this kind of thing). So I thought if I could find images of vehicles used by the Dutch and Danish armies, it would be possible to confirm the vehicle type and then work from there to identify the units that used this vehicle, before finally identifying where these units were based.
\n
But how to identify the vehicle? To most people it looks like a tank – but this is not strictly correct. Even if you’re not a military buff, there are a few features in this photo that help to identify the general vehicle type. The vehicle has tracks and a turret on top – but it is not a tank. Tanks are generally much larger, have bigger guns, are more heavily armoured, and do not have the ability to transport additional soldiers internally.
\nA Leopard 2 tank as used by the Danish Army\n
The vehicle in the quiz image has a door at the rear like the one below, which indicates that it is designed to accommodate additional soldiers. This means that the vehicle is not a tank, but a kind of armoured personnel carrier, also known as an armoured fighting vehicle. This is important because it means the unit that uses this vehicle will most likely be a reconnaissance / mechanised / armoured infantry unit, and not a tank or cavalry unit.
\nThe door at the rear indicates this vehicle (CV90) is intended to transport troops, so it is an armoured fighting vehicle or armoured personnel carrier, not a main battle tank. The shape and fearures of the rear of the vehicle also matches the one in the photo. \n
So first of all I looked at what kind of tracked vehicles the Dutch army uses. The only vehicle they use that fits the criteria (tracks, turret, rear door) is this CV90:
\nDutch CV90 from Wikipedia \n
So next I looked at the vehicles used by the Danish army, and again the only possible match was also the CV90:
\nDanish CV90 from Wikipedia\n
It turns out that the CV90 is used by lots of different countries. So how to determine who the CV90 in the original picture belonged to? I noticed that the registration plate on the Danish CV90 pictured above was identical in style to the one in the Quiztime image. The yellow circle emblem followed by the five digit serial was very similar:
\n
\n
\n
The identical symbol and similar number sequence (75 1xx) made me fairly sure this CV90 belonged to the Danish army and not to any other country.
\n
\n
Unit and Barracks
\n
So if the CV90 was Danish, was it possible find out where it was based and where it might have been headed to when the picture was taken? Fortunately the Danish Army is very small, and its main combat units are based at only two different locations. There are some additional bases for support units, but a vehicle like the CV90 belongs to a front line combat unit. This Wikipedia page gives some insight into how the army is structured, and this unofficial Danish Army site states that the CV90 was only issued to two units, the Life Guards and the Guards Hussar Regiment.
\n
\n
Taking all these things into account, it seems that there were only two possible locations, either 1st Brigade at Holstebro, or 2nd Brigade at Slagelse. While there are a few other army bases around Denmark, only these two garrisons seem to house combat units of the sort that use the CV90. Helpfully Wikipedia also provides a structural overview of the Danish Army that made things a little clearer:
\n
\n
Somewhat unhelpfully this shows that the units equipped with the CV90 (Life Guards and Guards Hussars Regiment) are split between Slagelse and Holstebro, which didn’t actually help at all with the location challenge since they are a long way apart and could have been based at either location. Was there any clue that could help point in the right direction? There was one additional detail on the back of the CV90 that indicated a possible link to Slagelse. This symbol from the back of the CV90 bore a strong resemblance to the official badge of the 2nd Brigade, based at Slagelse:
\nEmblem on rear of CV90\n
\nSymbol of Danish 2nd Brigade\n
Based on this I thought Slagelse was a more likely location for the vehicle’s home base rather than Holstebro. Having said all that, military vehicles move frequently, and units can be deployed almost anywhere at any time, so it is important to keep an open mind about possible locations.
\n
\n
Finding the CV90
\n
So I was confident about the type of vehicle it was, and which unit it probably belonged to. I knew it was Danish, but that didn’t actually help locate the vehicle. Apart from anything else, the vehicle was moving and could be anywhere. Doxsor said it was near to a military barracks – but the two most likely candidates (Slagelse and Holstebro) were at opposite ends of the country, so where to start?
\n
First of all I decided to focus on the road sign that was just visible on the left of the original image. This was a mistake, but I didn’t know that at the time.
\n
\n
The text isn’t too clear, but I though the top line said something –derup. My knowledge of Denmark’s geography is not good, so I did a Google search for *derup + Denmark in order to try and find the place name. The clearest match was for the town of Jyderup, which is not all that far from Slagelse (about 35 minutes’ drive). The only main road near Jyderup is highway 23, but no matter where I looked, I couldn’t find any locations that looked right. The 23 is mostly a single-track road, but the road in the picture looked like a multi-lane motorway. The signs were also very different. This is the sign for the exit from the 23 heading towards Jyderup:
\n
\n
\n
The sign was red text on a white background, but the kind of sign I was looking for was a standard motorway sign with white text on a blue background like this one:
\n
\n
\n
This meant that the 23 was definitely not a motorway, only a large trunk road. More importantly this meant that not only could Jyderup not be the correct location, but the real location must be on a motorway. Up until this point I had tried to do this quiz on my laptop with a 15″ screen. As soon as I switched to my PC with a much larger monitor it became clear I had been mistaken and that the sign in the quiz image could not possibly say “Jyderup”. Apart from anything else there was clearly another letter before the d in drup/derup and it wasn’t a y.
\n
So Jyderup was not the right place, but through this diversion I learned that in order to find the location it would only be necessary to focus on motorways and not any other roads, since the blue signs with white text seemed to be exclusive to motorways.
\n
\n
What Does The Picture Actually Show?
\n
After the Jyderup diversion I went back to the original image to see what other information it held. I’ve posted before about asking questions of an image, and it can be very helpful to help articulate precisely what you can see in front of you. I had already got a lot of detail about the CV90, but was there anything in the image that could be used to find the location? Actually there was quite a bit once I looked at it again:
\n
\n
1) Red Box – this is the shadow of the vehicle – so what does it show? Looking at the trees, the image must have been taken in summer. Denmark is in the northern hemisphere, so the sun is always to the south (i.e. it never shines from the north). So the fact that the vehicle’s shadow is on the right of the picture must mean that the sun is to the left. This means that the vehicle must be heading roughly west, and the orientation of the motorway must be roughly West-East, it is unlikely to be North-South. In theory it could be dawn or dusk with the sun shining from the East/West, but the relatively short shadow length made me think this was less likely.
\n
2) Green box – this shows the top of an electrical mast like the type used by electric trains. It appears to be a railway running parallel to the motorway. I found a full picture of an electrical railway set up to confirm the type used in Denmark:
\n
\n
3) Blue box – The sign shows an exit for a place ending with -drup, and a second place ending with -ov, or -on. There also appears to be something else behind the sign that is grey, perhaps road or a bridge.
\n
So putting everything together, it was already possible to be sure of the following:
\n
1) The location is on a motorway that most likely runs West-East / East-West
\n
2) The motorway is parallel to an electric railway (thanks to @JTillerup for pointing this out)
\n
3) The railway must be on the northern side of the motorway, not the southern side.
\n
4) The location is next to a motorway exit for a place ending in -drup / -derup and another place ending in -ov/-on.
\n
5) It must be quite close to one of the military barracks at either Holstebro or Slagelse.
\n
So by strictly applying these criteria, it is possible to narrow down the search area very quickly. Look at the map of Denmark below, which shows all the main motorways:
\n
\n
It might seem like a huge search area, but by applying the five criteria above it is possible to find the right part of Denmark quite quickly. The E39, E45, E47 and E55 can be immediately discounted because they run North-South and not West-East. This really just leaves the E20, which runs West-East across Denmark. The E20 is also nowhere near Holstebro, which means Slagelse is the most likely of the two barrack locations. For this reason I also decided not to focus on the western section of the E20 between Esbjerg and Kolding because it was a long way from either of the military bases. I decided to work from Copenhagen towards Odense first. The aim would be to see which section of the motorway had an electric railway running parallel to it on the northern side and then focus on that.
\n
The advantage of this approach is that it was possible to work with Google Maps in satellite mode from quite a high altitude. At this stage I was only looking for a railway next to a motorway, so there was no need for street-level detail. You could drive along the E20 in streetview mode of course, but it would take a long time. Working from this kind of height also means it is possible to cover a large area in a very short time:
\nRailway on the north side of a West-East motorway. Perfect!\n
After about ten minutes of checking, the only part of the E20 that had a railway alongside AND ran from West-East was the section between Olby and Fjaellebro, (there is another tiny section near Svenstrup, but a quick check with Streetview showed it didn’t match). The Olby-Fjaellebro section is only approximately 20 kilometres of road so it was only quite a small area to focus on.
\n
The target location had to be near to an exit junction, of which there are only 4-5 in this section. Secondly, the exit had to be for a place name that ended with -drup and a second place ending with -or/-ov. A very quick street view check focusing only on the points of exit led me here, Exit 33 for Vemmedrup and Bjaerskov.
\n
\n
By applying the criteria that the correct location must possess, this was the only possible location where Doxsor could have taken the picture, but there was only one problem – it does not look anything like the photograph! In the original photo there is an embankment and some bushes next to the motorway, but here there was no embankment and no bushes.
\n
Right Place, Wrong Perspective
\n
I was slightly puzzled at this point, and I think Sector035 had reached a similar stumbling block. Surely there was some contradiction? Logically this must be the correct place, but the Street View image did not match. I like to think that there is no such thing as a contradiction, and that if you think you have found one, then it means that one of your initial assumptions must be wrong. This turned out to be the case, although I was not very far off (only by a few metres in the end).
\n
After rechecking how I got to the location, I tried to play around with Street View a little bit to try and recreate the original photo, because I was still sure that logically this must be correct.
\n
In the original picture Doxsor was much further away from the road sign, so I backed up a little bit with Streetview and came to this position:
\n
\n
It looks almost right, but not quite. The sign also seems correct when you compare Street View to the original:
\n
\n
But in the original there is also no crash barrier. However after a few minutes of trying to replicate the original image as much as possible, I noticed that the Google Street View car was in Lane 1 of the motorway (inside lane), but Doxsor must have taken his photo from Lane 2 (outside lane), so this explained the slight differences of angle between Street View and the original picture. This along with the position of the truck in the original image also explains why the crash barrier is not quite visible.
\n
So the correct location was exist 33 of the E20 near Vemmedrup, and the nearest military barracks that the CV90 could have been heading to was most likely Antvorskov Barracks at Slagelse.
",
+ "content_text": "This Sunday we were treated to an extra Quiztime image location challenge which involved an unidentified military vehicle. Doxsor posted this image and asked where this military vehicle was headed to, and which barracks was closest to the location:\n\nThis was a good opportunity to practise military vehicle identification and explore how military units are organised and structured. OSINT is a very wide and varied discipline, but if your work focuses on conflict zones like Libya, Syria, or Ukraine, then being able to identify military vehicles and those that use them can be an integral part of an investigation. Bellingcat’s recent work on MH17 is a prime example of how essential this skill can be.\n\n“You ain’t seen me, right…”\n\nThe rest of this post will look at how I identified the vehicle and unit, before finding the correct location.\n\nTanks & Armoured Fighting Vehicles\n\nThere were a few ways to start of this puzzle, but I began with the number plate on the trailer. 6 digits in black print on a yellow background could be several different countries, but two strong candidates were the Netherlands and also Denmark. (World Licence Plates is a great resource for this kind of thing). So I thought if I could find images of vehicles used by the Dutch and Danish armies, it would be possible to confirm the vehicle type and then work from there to identify the units that used this vehicle, before finally identifying where these units were based.\n\nBut how to identify the vehicle? To most people it looks like a tank – but this is not strictly correct. Even if you’re not a military buff, there are a few features in this photo that help to identify the general vehicle type. The vehicle has tracks and a turret on top – but it is not a tank. Tanks are generally much larger, have bigger guns, are more heavily armoured, and do not have the ability to transport additional soldiers internally.\n\nA Leopard 2 tank as used by the Danish Army\n\nThe vehicle in the quiz image has a door at the rear like the one below, which indicates that it is designed to accommodate additional soldiers. This means that the vehicle is not a tank, but a kind of armoured personnel carrier, also known as an armoured fighting vehicle. This is important because it means the unit that uses this vehicle will most likely be a reconnaissance / mechanised / armoured infantry unit, and not a tank or cavalry unit.\n\nThe door at the rear indicates this vehicle (CV90) is intended to transport troops, so it is an armoured fighting vehicle or armoured personnel carrier, not a main battle tank. The shape and fearures of the rear of the vehicle also matches the one in the photo.\n\nSo first of all I looked at what kind of tracked vehicles the Dutch army uses. The only vehicle they use that fits the criteria (tracks, turret, rear door) is this CV90:\n\nDutch CV90 from Wikipedia\n\nSo next I looked at the vehicles used by the Danish army, and again the only possible match was also the CV90:\n\nDanish CV90 from Wikipedia\n\nIt turns out that the CV90 is used by lots of different countries. So how to determine who the CV90 in the original picture belonged to? I noticed that the registration plate on the Danish CV90 pictured above was identical in style to the one in the Quiztime image. The yellow circle emblem followed by the five digit serial was very similar:\n\nThe identical symbol and similar number sequence (75 1xx) made me fairly sure this CV90 belonged to the Danish army and not to any other country.\n\nUnit and Barracks\n\nSo if the CV90 was Danish, was it possible find out where it was based and where it might have been headed to when the picture was taken? Fortunately the Danish Army is very small, and its main combat units are based at only two different locations. There are some additional bases for support units, but a vehicle like the CV90 belongs to a front line combat unit. This Wikipedia page gives some insight into how the army is structured, and this unofficial Danish Army site states that the CV90 was only issued to two units, the Life Guards and the Guards Hussar Regiment.\n\nTaking all these things into account, it seems that there were only two possible locations, either 1st Brigade at Holstebro, or 2nd Brigade at Slagelse. While there are a few other army bases around Denmark, only these two garrisons seem to house combat units of the sort that use the CV90. Helpfully Wikipedia also provides a structural overview of the Danish Army that made things a little clearer:\n\nSomewhat unhelpfully this shows that the units equipped with the CV90 (Life Guards and Guards Hussars Regiment) are split between Slagelse and Holstebro, which didn’t actually help at all with the location challenge since they are a long way apart and could have been based at either location. Was there any clue that could help point in the right direction? There was one additional detail on the back of the CV90 that indicated a possible link to Slagelse. This symbol from the back of the CV90 bore a strong resemblance to the official badge of the 2nd Brigade, based at Slagelse:\n\nEmblem on rear of CV90\n\nSymbol of Danish 2nd Brigade\n\nBased on this I thought Slagelse was a more likely location for the vehicle’s home base rather than Holstebro. Having said all that, military vehicles move frequently, and units can be deployed almost anywhere at any time, so it is important to keep an open mind about possible locations.\n\nFinding the CV90\n\nSo I was confident about the type of vehicle it was, and which unit it probably belonged to. I knew it was Danish, but that didn’t actually help locate the vehicle. Apart from anything else, the vehicle was moving and could be anywhere. Doxsor said it was near to a military barracks – but the two most likely candidates (Slagelse and Holstebro) were at opposite ends of the country, so where to start?\n\nFirst of all I decided to focus on the road sign that was just visible on the left of the original image. This was a mistake, but I didn’t know that at the time.\n\nThe text isn’t too clear, but I though the top line said something –derup. My knowledge of Denmark’s geography is not good, so I did a Google search for *derup + Denmark in order to try and find the place name. The clearest match was for the town of Jyderup, which is not all that far from Slagelse (about 35 minutes’ drive). The only main road near Jyderup is highway 23, but no matter where I looked, I couldn’t find any locations that looked right. The 23 is mostly a single-track road, but the road in the picture looked like a multi-lane motorway. The signs were also very different. This is the sign for the exit from the 23 heading towards Jyderup:\n\nThe sign was red text on a white background, but the kind of sign I was looking for was a standard motorway sign with white text on a blue background like this one:\n\nThis meant that the 23 was definitely not a motorway, only a large trunk road. More importantly this meant that not only could Jyderup not be the correct location, but the real location must be on a motorway. Up until this point I had tried to do this quiz on my laptop with a 15″ screen. As soon as I switched to my PC with a much larger monitor it became clear I had been mistaken and that the sign in the quiz image could not possibly say “Jyderup”. Apart from anything else there was clearly another letter before the d in drup/derup and it wasn’t a y.\n\nSo Jyderup was not the right place, but through this diversion I learned that in order to find the location it would only be necessary to focus on motorways and not any other roads, since the blue signs with white text seemed to be exclusive to motorways.\n\nWhat Does The Picture Actually Show?\n\nAfter the Jyderup diversion I went back to the original image to see what other information it held. I’ve posted before about asking questions of an image, and it can be very helpful to help articulate precisely what you can see in front of you. I had already got a lot of detail about the CV90, but was there anything in the image that could be used to find the location? Actually there was quite a bit once I looked at it again:\n\n1) Red Box – this is the shadow of the vehicle – so what does it show? Looking at the trees, the image must have been taken in summer. Denmark is in the northern hemisphere, so the sun is always to the south (i.e. it never shines from the north). So the fact that the vehicle’s shadow is on the right of the picture must mean that the sun is to the left. This means that the vehicle must be heading roughly west, and the orientation of the motorway must be roughly West-East, it is unlikely to be North-South. In theory it could be dawn or dusk with the sun shining from the East/West, but the relatively short shadow length made me think this was less likely.\n\n2) Green box – this shows the top of an electrical mast like the type used by electric trains. It appears to be a railway running parallel to the motorway. I found a full picture of an electrical railway set up to confirm the type used in Denmark:\n\n3) Blue box – The sign shows an exit for a place ending with -drup, and a second place ending with -ov, or -on. There also appears to be something else behind the sign that is grey, perhaps road or a bridge.\n\nSo putting everything together, it was already possible to be sure of the following:\n\n1) The location is on a motorway that most likely runs West-East / East-West\n\n2) The motorway is parallel to an electric railway (thanks to @JTillerup for pointing this out)\n\n3) The railway must be on the northern side of the motorway, not the southern side.\n\n4) The location is next to a motorway exit for a place ending in -drup / -derup and another place ending in -ov/-on.\n\n5) It must be quite close to one of the military barracks at either Holstebro or Slagelse.\n\nSo by strictly applying these criteria, it is possible to narrow down the search area very quickly. Look at the map of Denmark below, which shows all the main motorways:\n\nIt might seem like a huge search area, but by applying the five criteria above it is possible to find the right part of Denmark quite quickly. The E39, E45, E47 and E55 can be immediately discounted because they run North-South and not West-East. This really just leaves the E20, which runs West-East across Denmark. The E20 is also nowhere near Holstebro, which means Slagelse is the most likely of the two barrack locations. For this reason I also decided not to focus on the western section of the E20 between Esbjerg and Kolding because it was a long way from either of the military bases. I decided to work from Copenhagen towards Odense first. The aim would be to see which section of the motorway had an electric railway running parallel to it on the northern side and then focus on that.\n\nThe advantage of this approach is that it was possible to work with Google Maps in satellite mode from quite a high altitude. At this stage I was only looking for a railway next to a motorway, so there was no need for street-level detail. You could drive along the E20 in streetview mode of course, but it would take a long time. Working from this kind of height also means it is possible to cover a large area in a very short time:\n\nRailway on the north side of a West-East motorway. Perfect!\n\nAfter about ten minutes of checking, the only part of the E20 that had a railway alongside AND ran from West-East was the section between Olby and Fjaellebro, (there is another tiny section near Svenstrup, but a quick check with Streetview showed it didn’t match). The Olby-Fjaellebro section is only approximately 20 kilometres of road so it was only quite a small area to focus on.\n\nThe target location had to be near to an exit junction, of which there are only 4-5 in this section. Secondly, the exit had to be for a place name that ended with -drup and a second place ending with -or/-ov. A very quick street view check focusing only on the points of exit led me here, Exit 33 for Vemmedrup and Bjaerskov.\n\nBy applying the criteria that the correct location must possess, this was the only possible location where Doxsor could have taken the picture, but there was only one problem – it does not look anything like the photograph! In the original photo there is an embankment and some bushes next to the motorway, but here there was no embankment and no bushes.\n\nRight Place, Wrong Perspective\n\nI was slightly puzzled at this point, and I think Sector035 had reached a similar stumbling block. Surely there was some contradiction? Logically this must be the correct place, but the Street View image did not match. I like to think that there is no such thing as a contradiction, and that if you think you have found one, then it means that one of your initial assumptions must be wrong. This turned out to be the case, although I was not very far off (only by a few metres in the end).\n\nAfter rechecking how I got to the location, I tried to play around with Street View a little bit to try and recreate the original photo, because I was still sure that logically this must be correct.\n\nIn the original picture Doxsor was much further away from the road sign, so I backed up a little bit with Streetview and came to this position:\n\nIt looks almost right, but not quite. The sign also seems correct when you compare Street View to the original:\n\nBut in the original there is also no crash barrier. However after a few minutes of trying to replicate the original image as much as possible, I noticed that the Google Street View car was in Lane 1 of the motorway (inside lane), but Doxsor must have taken his photo from Lane 2 (outside lane), so this explained the slight differences of angle between Street View and the original picture. This along with the position of the truck in the original image also explains why the crash barrier is not quite visible.\n\nSo the correct location was exist 33 of the E20 near Vemmedrup, and the nearest military barracks that the CV90 could have been heading to was most likely Antvorskov Barracks at Slagelse.\n\nFor more Quiztime-related posts click here.\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/EC1QuFiXoAArc2z-768x1024.jpeg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/original.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/leopard2a4.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_027-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/1280px-CV90_photo-003-1024x768.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/1280px-Combined_Resolve_III_141106-A-SG416-002-1024x723.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_029-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_028-2.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_030-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/1280px-Royal_Danish_Army_2019-1024x718.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_001-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/DIB_armpatch.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_002-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_023-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_018-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/EC1QuFiXoAArc2z.-highlights-768x1024.jpeg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_004-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Denmark-map-motorways.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_005-2.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_024-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2019/08/Selection_025-2-1024x575.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2020-01-31_seeking-sector035-quiztime-30th-january-2020.json b/nixintel_geolocation_json/2020-01-31_seeking-sector035-quiztime-30th-january-2020.json
new file mode 100644
index 0000000000000000000000000000000000000000..f26db3130b22863c6e9b0af3edaed73eceb6ad06
--- /dev/null
+++ b/nixintel_geolocation_json/2020-01-31_seeking-sector035-quiztime-30th-january-2020.json
@@ -0,0 +1,58 @@
+{
+ "slug": "seeking-sector035-quiztime-30th-january-2020",
+ "title": "Seeking Sector035 – Quiztime 30th January 2020",
+ "url": "https://nixintel.info/osint/seeking-sector035-quiztime-30th-january-2020/",
+ "date": "2020-01-31T21:11:22+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
January has been a busy month both personally and professionally so I haven’t been able to take part in as many Quiztime challenges as I’d like to. Fortunately I had time to try Sector035’s Thursday Quiz. Sector035 posted not one but two photos. All we had to do was establish where exactly they were taken. Here are the two images:
\n
\n
\n
So where to start? Sometimes a nice quick reverse image search will help to identify a location quickly, but no such luck in this case. Reverse image searching is an essential tool for geolocation work, but it still has an awful lot of limitations. Besides where’s the fun in finding everything with a reverse image search?
\n
One useful method for finding photo locations is to try and extract as much information as possible from a photo and then use that to set some more detailed search parameters to home in on the location. It’s almost like a form of Boolean logic. If I can see that conditions X and Y are true in a photo, then searching for places where X and Y are true will help me find the location.
\n
This sounds very simple in theory, but in practice it can be difficult because searching for some “true” information will still leave a very large pool of possible locations. If you saw the first photo and searched for “Burger King” and then “Germany” because of the writing on the van in the second photo, you would find hundreds of possible matches. This is because there are more than 700 Burger King restaurants in Germany, so it doesn’t really help you to determine where this could be.
\n
\n
Narrowing Down The Possibilities
\n
The more “true” conditions you can identify in an image, the more specific your search will be. So for the first photo we’ve already seen that there is a Burger King, but we can also see that the napkin is from a Coffee Fellows outlet. The second photo also reveals a lot of “true” conditions that we can combine with these two and turn into search parameters:
\n
1) The location appears to be some kind service area, based on the parked vehicles.
\n
2) There is a van with German text on the side. Vehicles move, and Germany has borders with ten other countries, so by itself a German vehicle in a service area is not absolute proof that this is in Germany. It does make Germany a strong probability however.
\n
\n
3) There’s a blue sign that says “Hannover 2”. Even if you have never visited Germany (and you really should), a quick Google will tell you that Hannover is in Germany. Moreover there are really useful sites like the Motorway Signs database which will help to show that this style of sign is common to Germany motorways. This means the sign is indicating the direction to Hannover along Autobahn 2 (A2).
\nThere really is a resource for everything.\n
4) There are yellow bins and also a yellow bench at the location.
\n
With a little more data extracted from the photos, searching will be much easier and we will be able to find the location pretty quickly. All we have to do is put all our “true” conditions together and begin to search for them. We need to find a motorway service area on the A2 that has a Burger King, a Coffee Fellows, and is in the right direction of travel for Hannover. We also know that there will be yellow bins and a yellow bench at the location. These can be helpful for confirming the final position.
\n
\n
So how to begin finding the right service area? You could load up Google Maps out and start tracing a route along the A2, looking for a Burger King. The A2 stretches 473km across Germany and life is too short to do things this way. Instead we can use some of the details we identified and combine them into Boolean Google searches to help find the location. The more “true” conditions we can add, the more refined the results will be. For example:
A little better. We are getting some specific service areas mentioned now. One of them will be correct, but we can still do more to refine the search by adding another true condition to our Boolean search:
Just two service areas stand out from these results. One is at Gütersloh, and the other at Buckautal. We can quickly check these locations out with Google Maps to see if our theory is right. There’s no Street View coverage at Gütersloh, but there are quite a lot of images of the area uploaded to Google Maps. I can’t see any yellow bins though:
\n
\n
We established at the start that “yellow bin” must be a true condition of the right location. There are no yellow bins here like the one in the challenge photo, so unless we’ve missed something then logically this cannot be the location. So about Buckautal?
\n
\n
\n
It has the exact same kind of yellow bins as those in Sector035’s picture. There’s no Street View coverage here, but there is a panoramic photo that we can use to try and verify if this is the location:
\n
\n
Perfect! We can compare this to the original and see that Sector035 took his challenge photo just a few steps away from where this picture was taken in 2017.
",
+ "content_text": "January has been a busy month both personally and professionally so I haven’t been able to take part in as many Quiztime challenges as I’d like to. Fortunately I had time to try Sector035’s Thursday Quiz. Sector035 posted not one but two photos. All we had to do was establish where exactly they were taken. Here are the two images:\n\nSo where to start? Sometimes a nice quick reverse image search will help to identify a location quickly, but no such luck in this case. Reverse image searching is an essential tool for geolocation work, but it still has an awful lot of limitations. Besides where’s the fun in finding everything with a reverse image search?\n\nOne useful method for finding photo locations is to try and extract as much information as possible from a photo and then use that to set some more detailed search parameters to home in on the location. It’s almost like a form of Boolean logic. If I can see that conditions X and Y are true in a photo, then searching for places where X and Y are true will help me find the location.\n\nThis sounds very simple in theory, but in practice it can be difficult because searching for some “true” information will still leave a very large pool of possible locations. If you saw the first photo and searched for “Burger King” and then “Germany” because of the writing on the van in the second photo, you would find hundreds of possible matches. This is because there are more than 700 Burger King restaurants in Germany, so it doesn’t really help you to determine where this could be.\n\nNarrowing Down The Possibilities\n\nThe more “true” conditions you can identify in an image, the more specific your search will be. So for the first photo we’ve already seen that there is a Burger King, but we can also see that the napkin is from a Coffee Fellows outlet. The second photo also reveals a lot of “true” conditions that we can combine with these two and turn into search parameters:\n\n1) The location appears to be some kind service area, based on the parked vehicles.\n\n2) There is a van with German text on the side. Vehicles move, and Germany has borders with ten other countries, so by itself a German vehicle in a service area is not absolute proof that this is in Germany. It does make Germany a strong probability however.\n\n3) There’s a blue sign that says “Hannover 2”. Even if you have never visited Germany (and you really should), a quick Google will tell you that Hannover is in Germany. Moreover there are really useful sites like the Motorway Signs database which will help to show that this style of sign is common to Germany motorways. This means the sign is indicating the direction to Hannover along Autobahn 2 (A2).\n\nThere really is a resource for everything.\n\n4) There are yellow bins and also a yellow bench at the location.\n\nWith a little more data extracted from the photos, searching will be much easier and we will be able to find the location pretty quickly. All we have to do is put all our “true” conditions together and begin to search for them. We need to find a motorway service area on the A2 that has a Burger King, a Coffee Fellows, and is in the right direction of travel for Hannover. We also know that there will be yellow bins and a yellow bench at the location. These can be helpful for confirming the final position.\n\nSo how to begin finding the right service area? You could load up Google Maps out and start tracing a route along the A2, looking for a Burger King. The A2 stretches 473km across Germany and life is too short to do things this way. Instead we can use some of the details we identified and combine them into Boolean Google searches to help find the location. The more “true” conditions we can add, the more refined the results will be. For example:\n\n“Burger King” + A2\n\nThis brings back too many results. Lots of other countries also have a motorway called the A2.\n\n“Burger King” + A2 + Hannover\n\nA little better. We are getting some specific service areas mentioned now. One of them will be correct, but we can still do more to refine the search by adding another true condition to our Boolean search:\n\n“Coffee Fellows” + “Burger King” + a2 + Hannover\n\nJust two service areas stand out from these results. One is at Gütersloh, and the other at Buckautal. We can quickly check these locations out with Google Maps to see if our theory is right. There’s no Street View coverage at Gütersloh, but there are quite a lot of images of the area uploaded to Google Maps. I can’t see any yellow bins though:\n\nWe established at the start that “yellow bin” must be a true condition of the right location. There are no yellow bins here like the one in the challenge photo, so unless we’ve missed something then logically this cannot be the location. So about Buckautal?\n\nIt has the exact same kind of yellow bins as those in Sector035’s picture. There’s no Street View coverage here, but there is a panoramic photo that we can use to try and verify if this is the location:\n\nPerfect! We can compare this to the original and see that Sector035 took his challenge photo just a few steps away from where this picture was taken in 2017.\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/01/EPjMUYUWAAEy7vJ-1536x2048.jpeg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/01/EPjMUZKXUAU2vRr-2048x1536.jpeg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/01/Selection_003-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/01/Selection_004-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/01/Selection_005-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/01/Selection_007-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/01/Selection_008-2-2048x1041.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/01/Selection_009-2.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/01/Selection_010-1.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2020-02-08_where-was-this-taken-church-spires-and-steering-wheels.json b/nixintel_geolocation_json/2020-02-08_where-was-this-taken-church-spires-and-steering-wheels.json
new file mode 100644
index 0000000000000000000000000000000000000000..db74d2653ecff62573c8c60cc6a8e187db08d776
--- /dev/null
+++ b/nixintel_geolocation_json/2020-02-08_where-was-this-taken-church-spires-and-steering-wheels.json
@@ -0,0 +1,58 @@
+{
+ "slug": "where-was-this-taken-church-spires-and-steering-wheels",
+ "title": "Where Was This Taken? Church Spires and Steering Wheels",
+ "url": "https://nixintel.info/osint/where-was-this-taken-church-spires-and-steering-wheels/",
+ "date": "2020-02-08T15:35:25+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
\n
\n
In OSINT as in sport, there is definitely such a thing as home fixture advantage. We are always going to be able to identify the important little clues in an environment we are familiar with rather than one we aren’t. So when I saw Paul Fennell post this geolocation challenge I was pleased because my initial gut reaction was that this must be in the UK. The design of the church, the red brick terrace houses and even the car park look like they must be in the UK somewhere.
\n
There’s only one problem though – Paul tells us that this is not in the UK. There’s also something else staring us in the face that indicates this is not in the UK – the car registration plate. In the UK car registration plates are white at the front, but yellow at the back. The car clearly has white plates at the rear – so this is not a UK-registered vehicle. It seems there will be no home fixture advantage after all.
\n
\n
So what can we do to find out where this picture was taken? The image is quite low resolution so picking out street names and signs isn’t possible. Cropping out and reverse-image searching the houses and the church didn’t work either, the images are just not of sufficiently high quality to be useful. There is not going to be any quick fix to solve this puzzle so we need to extract a little more detail to move forward.
\n
The plan I came up with to find this location went something like this:
\n
1) Find a country that looks like the UK, but isn’t.
\n
2) Focus on the church and try to find church spires in that country which match the one in the photo.
\n
3) Use Google Maps 3D and/or Streetview to check the area around any matching churches to see if it is the correct location.
\n
So far the car has indicated that despite my initial reaction, this is not in the UK. There’s something else we can learn from the car that will help find the correct country. Even though the image is of poor quality, if we zoom in on the car there’s a clue in the window.
\n
\n
\n
If we look closely at the dashboard, what can we see? Nothing of note – and this is exactly what is so helpful. If this car were in a country that drives on the right, we would be able to see the steering wheel through the front window. Instead there’s just a plain old dashboard. This means that the car’s steering wheel must be on the right hand side, so the car must be a from a country that drives on the left.
\n
\n
If we have a look at a list of countries that drive on the left, there are very few that could physically resemble the location in Paul’s picture. It clearly isn’t taken in southern Africa, India, or Japan. We know it’s not the UK, so the most realistic possibility is Ireland. Irish cars also have white number plates at the rear, which helps to support the hypothesis.
\n
Church Spire Hell
\n
So how to find the right location in Ireland? I chose to focus on the church because it is by far the most unique and distinctive feature in the original image. This method is not without its own problems however. Ireland has a lot of churches. I spent some time browsing through image search results for Irish church spires but there are just too many. When searching for this kind of information it’s also worth remembering that search engines do not treat everything equally, and more popular locations will feature more prominently. If you Google for “church spires Ireland” you’ll find plenty of matches for some of the more well-known churches (and some that are not even in Ireland), but the search term is not specific enough to give the granularity we need.
\nSo many churches….\n
\n
This means we need to pause and refine our church spire searching a little. How about adding in town names to the search term? Instead of searching at country-level, searching for church spire + [town name] will give a much better level of detail. We still need to be careful not to make it too difficult though. If you’re an absolute masochist you could work through this list, but it would take a very long time to search through every last town and village. We need to be a little bit smarter and look at the photo again for a little more information.
\nJust no.\n
What sort of a place do we think this is? It isn’t a tiny little rural village, but we can’t see the high rise buildings of a large city either. The church itself is not a massive cathedral, but neither is it a tiny little village chapel. This probably means that the picture was taken in a medium sized town, so we’ll just focus on those. We can always change the search parameters if this approach doesn’t work.
\n
I decided to use this list of Irish towns and cities, sorted by size. It’s still a fairly long list, but it’s important to remember that towns in Northern Ireland can be excluded from our search since they are still classed as part of the UK. We can open up Google images and work through the list pretty quickly e.g.
\n
church spire limerick
\n
church spire galway
\n
etc.
\n
It’s usually immediately obvious whether there is a match or not, so it only takes a few seconds and a quick scan of the returned images to see if we’re on the right track. It’s slightly repetitive but it’s a fairly comprehensive way to find what we need quickly.
\n
Sure enough after about 20 negative searches we search for church spire athlone and see a familiar looking building:
\n
\n
This photo is from a different angle to the one in the original image but all the features match: a single spire that is the correct shape, and also located at the end of the main building. Now we can use Google Maps to check our hypothesis. Here’s St Mary’s Church in Athlone, as returned by the targeted Google image search:
\n
\n
\n
It’s near to a car park too, so let’s orientate Google Maps to match the original photo to see if we’re right:
\n
\n
Looks good. Even the terraced houses on the left seem to match. We can use Street View for a final check to see if we’re right:
\n
\n
A very familiar match! We can even see the same car in Street View, which of course means that this must be the source of Paul’s original challenge photo. So the correct location is Grace Road, Athlone, Ireland.
\n
For more photo challenges like this you can follow Paul on Twitter @Digit4lBytes or of course take part in the daily verification challenge at @Quiztime. For more of my geolocation posts, click here.
",
+ "content_text": "In OSINT as in sport, there is definitely such a thing as home fixture advantage. We are always going to be able to identify the important little clues in an environment we are familiar with rather than one we aren’t. So when I saw Paul Fennell post this geolocation challenge I was pleased because my initial gut reaction was that this must be in the UK. The design of the church, the red brick terrace houses and even the car park look like they must be in the UK somewhere.\n\nThere’s only one problem though – Paul tells us that this is not in the UK. There’s also something else staring us in the face that indicates this is not in the UK – the car registration plate. In the UK car registration plates are white at the front, but yellow at the back. The car clearly has white plates at the rear – so this is not a UK-registered vehicle. It seems there will be no home fixture advantage after all.\n\nSo what can we do to find out where this picture was taken? The image is quite low resolution so picking out street names and signs isn’t possible. Cropping out and reverse-image searching the houses and the church didn’t work either, the images are just not of sufficiently high quality to be useful. There is not going to be any quick fix to solve this puzzle so we need to extract a little more detail to move forward.\n\nThe plan I came up with to find this location went something like this:\n\n1) Find a country that looks like the UK, but isn’t.\n\n2) Focus on the church and try to find church spires in that country which match the one in the photo.\n\n3) Use Google Maps 3D and/or Streetview to check the area around any matching churches to see if it is the correct location.\n\nSo far the car has indicated that despite my initial reaction, this is not in the UK. There’s something else we can learn from the car that will help find the correct country. Even though the image is of poor quality, if we zoom in on the car there’s a clue in the window.\n\nIf we look closely at the dashboard, what can we see? Nothing of note – and this is exactly what is so helpful. If this car were in a country that drives on the right, we would be able to see the steering wheel through the front window. Instead there’s just a plain old dashboard. This means that the car’s steering wheel must be on the right hand side, so the car must be a from a country that drives on the left.\n\nIf we have a look at a list of countries that drive on the left, there are very few that could physically resemble the location in Paul’s picture. It clearly isn’t taken in southern Africa, India, or Japan. We know it’s not the UK, so the most realistic possibility is Ireland. Irish cars also have white number plates at the rear, which helps to support the hypothesis.\n\nChurch Spire Hell\n\nSo how to find the right location in Ireland? I chose to focus on the church because it is by far the most unique and distinctive feature in the original image. This method is not without its own problems however. Ireland has a lot of churches. I spent some time browsing through image search results for Irish church spires but there are just too many. When searching for this kind of information it’s also worth remembering that search engines do not treat everything equally, and more popular locations will feature more prominently. If you Google for “church spires Ireland” you’ll find plenty of matches for some of the more well-known churches (and some that are not even in Ireland), but the search term is not specific enough to give the granularity we need.\n\nSo many churches….\n\nThis means we need to pause and refine our church spire searching a little. How about adding in town names to the search term? Instead of searching at country-level, searching for church spire + [town name] will give a much better level of detail. We still need to be careful not to make it too difficult though. If you’re an absolute masochist you could work through this list, but it would take a very long time to search through every last town and village. We need to be a little bit smarter and look at the photo again for a little more information.\n\nJust no.\n\nWhat sort of a place do we think this is? It isn’t a tiny little rural village, but we can’t see the high rise buildings of a large city either. The church itself is not a massive cathedral, but neither is it a tiny little village chapel. This probably means that the picture was taken in a medium sized town, so we’ll just focus on those. We can always change the search parameters if this approach doesn’t work.\n\nI decided to use this list of Irish towns and cities, sorted by size. It’s still a fairly long list, but it’s important to remember that towns in Northern Ireland can be excluded from our search since they are still classed as part of the UK. We can open up Google images and work through the list pretty quickly e.g.\n\nchurch spire limerick\n\nchurch spire galway\n\netc.\n\nIt’s usually immediately obvious whether there is a match or not, so it only takes a few seconds and a quick scan of the returned images to see if we’re on the right track. It’s slightly repetitive but it’s a fairly comprehensive way to find what we need quickly.\n\nSure enough after about 20 negative searches we search for church spire athlone and see a familiar looking building:\n\nThis photo is from a different angle to the one in the original image but all the features match: a single spire that is the correct shape, and also located at the end of the main building. Now we can use Google Maps to check our hypothesis. Here’s St Mary’s Church in Athlone, as returned by the targeted Google image search:\n\nIt’s near to a car park too, so let’s orientate Google Maps to match the original photo to see if we’re right:\n\nLooks good. Even the terraced houses on the left seem to match. We can use Street View for a final check to see if we’re right:\n\nA very familiar match! We can even see the same car in Street View, which of course means that this must be the source of Paul’s original challenge photo. So the correct location is Grace Road, Athlone, Ireland.\n\nFor more photo challenges like this you can follow Paul on Twitter @Digit4lBytes or of course take part in the daily verification challenge at @Quiztime. For more of my geolocation posts, click here.\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_001.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_002.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_003.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_004-2048x410.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_005.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/800px-Athlone_-_St_Marys_Church_-_20180921174650.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Screenshot-2020-02-08-at-12.17.42.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Screenshot-2020-02-08-at-12.20.05.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Screenshot-2020-02-08-at-12.22.02.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2020-02-27_geolocation-more-info-more-accuracy.json b/nixintel_geolocation_json/2020-02-27_geolocation-more-info-more-accuracy.json
new file mode 100644
index 0000000000000000000000000000000000000000..24d7883d4971776a902d5fbb6a4f689858bf55a7
--- /dev/null
+++ b/nixintel_geolocation_json/2020-02-27_geolocation-more-info-more-accuracy.json
@@ -0,0 +1,108 @@
+{
+ "slug": "geolocation-more-info-more-accuracy",
+ "title": "Geolocation: More Info, More Accuracy",
+ "url": "https://nixintel.info/osint/geolocation-more-info-more-accuracy/",
+ "date": "2020-02-27T21:13:27+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
How can you become more accurate with photo geolocation? Finding a country is good, finding a town is better, finding a street in a town is better still – but is it possible to get even more accurate than that? It depends on the needs of your case, but if you want to geolocate and investigate something as specific as an airstrike on a mosque or where exactly a murder took place then being precise is really important. Unless you’re lucky enough to come across an image on social media that still contains EXIF data, then one of best ways to add extra granularity to a geolocation search is by increasing the amount of data available to use as a reference. This is easier to say than to do of course, and it is hardly controversial to say that multiple sources are better than one, but the type and quality of these sources can really make a difference to how accurate geolocation can be.
\n
To illustrate the point I’m going to use Ruaridh Arrow’s recent photo challenge from 10th February this year. Here’s the original tweet:
\n
\n
So where is the broken down Apache? Ruaridh already tells us that it’s in Lincolnshire, but this is quite a large area and we want to narrow this down to a few square metres if possible. The photo doesn’t help us all that much though. Lincolnshire is a large, flat area and there are very few distinctive features in the photo that will add any more precision to our search at this stage. This means we need more detail, so we need to find additional sources of information. The good news is that an Apache landing in a field is a very unusual event so hopefully there should be lots of other information out there, we just need to find it.
\n
(I should point out that this incident took place more than two weeks prior to when I wrote this post, so the amount of available information has increased significantly since then, but for the purposes of this blog I’ve only used sources that were available at the time of the incident.)
So the Apache landed in Helpringham, not far from Sleaford:
\n
\n
Case closed. We’ve found the helicopter!
\n
Well not really of course, we can do much, much better than this. We want to know exactly where the helicopter was if possible. This might only be a relatively unimportant incident in the Lincolnshire countryside, but the principle of refining searches and always seeking to be as accurate as possible is a good habit to cultivate. Let’s see what else the article can tell us. Here’s another picture:
\nThe locals fled in terror from the “big giant bird on wheels without horses.”\n
This photo is taken from a similar position to the original one. It doesn’t add much more location detail – although the position of the sun might at least help with the orientation.
\n
\n
The next image (above) was taken just a few metres from the previous one, but it still doesn’t add much more detail. The next image is more promising though:
\n
\n
That’s much better! The image contains a wealth of details that will help to enhance the accuracy of the geolocation search. There’s a church spire, a phone mast, houses, solar panels, hedgerows and other details that will help us be very specific in locating this helicopter. By expanding the range of sources just a little we have added a lot more helpful detail. Next we need to use this new information to search around Helpringham in more detail, but we need to see if we can improve the quality of these photo sources a little first. The image above is only 700 pixels wide – this isn’t a very good resolution. The width of the image is actually specified in the URL:
Images used by media websites are often served in such a way that the size of the image to be displayed is specified in the URL. In the example above the image is set to w_700 (width of 700 pixels) but by adjusting the URL we can request a much larger version of the image from the server. Let’s get a version that’s 2400 pixels across instead by tweaking the URL as follows:
If you want to feel like a real pro, you can say “computer, enhance” as you hit Enter.
\n
There’s also another image on the page that will help because it shows the position of the Apache relative to the hedge:
\n
\n
And this one will help determine how far it is from the road:
\n
\n
We’re already in a much better position to geolocate the helicopter than when we first started. Next we will highlight some of the details in the extra images we found to home in on the exact spot:
\n
\n
The most obvious place to start is the church spire (yellow box). Next we can follow more or less from the spire to the phone mast (red box) to get the approximate location of the helicopter. The blue line is a hedgerow next to a road we can also use as a point of reference. The three purple lines indicate different fields, so know the Apache is in the third field from the position of the mobile phone mast as it appears in the photo. The houses and solar panels in the other boxes will help as additional reference points.
\n
Here’s Helpringham, with the pin centred on the church:
\n
\n
We need to use the details we found in the extra images to locate the exact spot. We’re looking for a phone mast next to a road. There aren’t many roads out of Helpringham, so it doesn’t take too long to find:
\n
\n
\n
The phone mast is highlighted in red. The hedgerow is the blue line, so we know the Apache must be close by somewhere.
\n
Streetview would be a great tool to help with an ultra-precise geolocation, but the age of the images might pose a problem. This is what the road looks like on Streetview:
\n
\n
It’s an mage taken eleven years ago and in a different season, so direct comparisons to the Apache photos are difficult, although not impossible. We could just stop at this point and say that the Apache was located somewhere off North Drove, Helpringham. This was my initial guess and it’s definitely an improvement on just “Helpringham”, but we want to be as accurate as possible so we will need to extract even more detail from the source photographs.
\n
The landscape surrounding the Apache landing site is flat and fairly featureless, but by picking out even the smallest of details we can help to confirm the exact landing spot. This means that some details that were initially too generic to be helpful are now more useful. For example the two pylons and a row of tall narrow trees visible in the distance of the source photograph:
\n
\n
And on the other side of the road, we can see a telegraph pole in the field:
\n
\n
We can use Streetview to follow the road and identify where these features are. Here are some telegraph poles on the opposite side of the road:
\n
\n
\n
\n
And turning to the left, we can see the two pylons and the row of tall thin trees in the distance:
\n
\n
Now we can add all the features that we’ve identified so far onto an aerial photo of the whole area:
\n
\n
With all these reference points identified we can return to the original photo that Ruaridh asked us about and work out that the photographer was stood here at 52.942661, -0.282871.
\n
\n
A final check with the reference points in the original image:
\n
\n
This means that the exact location of the Apache is here at 52.942558, -0.283232.
\n
\n
Admittedly “52.942558, -0.283232” doesn’t quite trip off the tongue in the same way as “just outside Helpringham”, but it does convey a great deal more accuracy, and sometimes accuracy really matters. It all depends on what you’re trying to use geolocation for of course and sometimes a town or a street is enough, but sometimes you might want to get within a few metres to prove the point that you’re trying to make. Expanding the range of sources makes gathering more information possible, and the more available information you have, the more precise it is possible to be.
",
+ "content_text": "How can you become more accurate with photo geolocation? Finding a country is good, finding a town is better, finding a street in a town is better still – but is it possible to get even more accurate than that? It depends on the needs of your case, but if you want to geolocate and investigate something as specific as an airstrike on a mosque or where exactly a murder took place then being precise is really important. Unless you’re lucky enough to come across an image on social media that still contains EXIF data, then one of best ways to add extra granularity to a geolocation search is by increasing the amount of data available to use as a reference. This is easier to say than to do of course, and it is hardly controversial to say that multiple sources are better than one, but the type and quality of these sources can really make a difference to how accurate geolocation can be.\n\nTo illustrate the point I’m going to use Ruaridh Arrow’s recent photo challenge from 10th February this year. Here’s the original tweet:\n\nSo where is the broken down Apache? Ruaridh already tells us that it’s in Lincolnshire, but this is quite a large area and we want to narrow this down to a few square metres if possible. The photo doesn’t help us all that much though. Lincolnshire is a large, flat area and there are very few distinctive features in the photo that will add any more precision to our search at this stage. This means we need more detail, so we need to find additional sources of information. The good news is that an Apache landing in a field is a very unusual event so hopefully there should be lots of other information out there, we just need to find it.\n\n(I should point out that this incident took place more than two weeks prior to when I wrote this post, so the amount of available information has increased significantly since then, but for the purposes of this blog I’ve only used sources that were available at the time of the incident.)\n\nThe local newspaper provides us with several eyewitness accounts and some important photos. The first paragraph gives more information about where the helicopter was:\n\nSo the Apache landed in Helpringham, not far from Sleaford:\n\nCase closed. We’ve found the helicopter!\n\nWell not really of course, we can do much, much better than this. We want to know exactly where the helicopter was if possible. This might only be a relatively unimportant incident in the Lincolnshire countryside, but the principle of refining searches and always seeking to be as accurate as possible is a good habit to cultivate. Let’s see what else the article can tell us. Here’s another picture:\n\nThe locals fled in terror from the “big giant bird on wheels without horses.”\n\nThis photo is taken from a similar position to the original one. It doesn’t add much more location detail – although the position of the sun might at least help with the orientation.\n\nThe next image (above) was taken just a few metres from the previous one, but it still doesn’t add much more detail. The next image is more promising though:\n\nThat’s much better! The image contains a wealth of details that will help to enhance the accuracy of the geolocation search. There’s a church spire, a phone mast, houses, solar panels, hedgerows and other details that will help us be very specific in locating this helicopter. By expanding the range of sources just a little we have added a lot more helpful detail. Next we need to use this new information to search around Helpringham in more detail, but we need to see if we can improve the quality of these photo sources a little first. The image above is only 700 pixels wide – this isn’t a very good resolution. The width of the image is actually specified in the URL:\n\nhttps://images-o.jpimedia.uk/imagefetch/w_700,f_auto,ar_3:2,q_auto:low,c_fill/if_h_lte_200,c_mfit,h_201/https://www.sleafordstandard.co.uk/webimage/1.9229122.1581340510!/image/image.jpg\n\nImages used by media websites are often served in such a way that the size of the image to be displayed is specified in the URL. In the example above the image is set to w_700 (width of 700 pixels) but by adjusting the URL we can request a much larger version of the image from the server. Let’s get a version that’s 2400 pixels across instead by tweaking the URL as follows:\n\nhttps://images-o.jpimedia.uk/imagefetch/w_2400,f_auto,ar_3:2,q_auto:low,c_fill/if_h_lte_200,c_mfit,h_201/https://www.sleafordstandard.co.uk/webimage/1.9229122.1581340510!/image/image.jpg\n\nIf you want to feel like a real pro, you can say “computer, enhance” as you hit Enter.\n\nThere’s also another image on the page that will help because it shows the position of the Apache relative to the hedge:\n\nAnd this one will help determine how far it is from the road:\n\nWe’re already in a much better position to geolocate the helicopter than when we first started. Next we will highlight some of the details in the extra images we found to home in on the exact spot:\n\nThe most obvious place to start is the church spire (yellow box). Next we can follow more or less from the spire to the phone mast (red box) to get the approximate location of the helicopter. The blue line is a hedgerow next to a road we can also use as a point of reference. The three purple lines indicate different fields, so know the Apache is in the third field from the position of the mobile phone mast as it appears in the photo. The houses and solar panels in the other boxes will help as additional reference points.\n\nHere’s Helpringham, with the pin centred on the church:\n\nWe need to use the details we found in the extra images to locate the exact spot. We’re looking for a phone mast next to a road. There aren’t many roads out of Helpringham, so it doesn’t take too long to find:\n\nThe phone mast is highlighted in red. The hedgerow is the blue line, so we know the Apache must be close by somewhere.\n\nStreetview would be a great tool to help with an ultra-precise geolocation, but the age of the images might pose a problem. This is what the road looks like on Streetview:\n\nIt’s an mage taken eleven years ago and in a different season, so direct comparisons to the Apache photos are difficult, although not impossible. We could just stop at this point and say that the Apache was located somewhere off North Drove, Helpringham. This was my initial guess and it’s definitely an improvement on just “Helpringham”, but we want to be as accurate as possible so we will need to extract even more detail from the source photographs.\n\nThe landscape surrounding the Apache landing site is flat and fairly featureless, but by picking out even the smallest of details we can help to confirm the exact landing spot. This means that some details that were initially too generic to be helpful are now more useful. For example the two pylons and a row of tall narrow trees visible in the distance of the source photograph:\n\nAnd on the other side of the road, we can see a telegraph pole in the field:\n\nWe can use Streetview to follow the road and identify where these features are. Here are some telegraph poles on the opposite side of the road:\n\nAnd turning to the left, we can see the two pylons and the row of tall thin trees in the distance:\n\nNow we can add all the features that we’ve identified so far onto an aerial photo of the whole area:\n\nWith all these reference points identified we can return to the original photo that Ruaridh asked us about and work out that the photographer was stood here at 52.942661, -0.282871.\n\nA final check with the reference points in the original image:\n\nThis means that the exact location of the Apache is here at 52.942558, -0.283232.\n\nAdmittedly “52.942558, -0.283232” doesn’t quite trip off the tongue in the same way as “just outside Helpringham”, but it does convey a great deal more accuracy, and sometimes accuracy really matters. It all depends on what you’re trying to use geolocation for of course and sometimes a town or a street is enough, but sometimes you might want to get within a few metres to prove the point that you’re trying to make. Expanding the range of sources makes gathering more information possible, and the more available information you have, the more precise it is possible to be.\n\nFor more posts about geolocation click here.\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_001-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_002-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_003-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/apache1.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/apache2.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/apache3.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/apache4.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/apache6.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/apache3lg-notes-2048x1365.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_004-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_005_notes-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_012.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/EQQ3g-1WsAMIZpb.jpeg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/apache7-2-2048x1365.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_019.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_014.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_015-2.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_016.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/02/Selection_018-1.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2020-03-15_using-gap-analysis-for-smarter-osint-quiztime-4th-march-2020.json b/nixintel_geolocation_json/2020-03-15_using-gap-analysis-for-smarter-osint-quiztime-4th-march-2020.json
new file mode 100644
index 0000000000000000000000000000000000000000..1124e1cc91ee1ad12a7566c89247d5d4a1a9fae4
--- /dev/null
+++ b/nixintel_geolocation_json/2020-03-15_using-gap-analysis-for-smarter-osint-quiztime-4th-march-2020.json
@@ -0,0 +1,123 @@
+{
+ "slug": "using-gap-analysis-for-smarter-osint-quiztime-4th-march-2020",
+ "title": "Using Gap Analysis For Smarter OSINT – Quiztime 4th March 2020",
+ "url": "https://nixintel.info/osint/using-gap-analysis-for-smarter-osint-quiztime-4th-march-2020/",
+ "date": "2020-03-15T19:22:22+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
In OSINT we make use of the wealth of information that is now available to us, but so often it becomes the case that there is too much information for us to process effectively. It’s easy to drown in a sea of data and you can quickly find yourself with twenty browser tabs open having completely forgotten what it was that you originally set out to do. Going down some rabbit holes is inevitable because often you can only know that you were on the right track with the benefit of hindsight and you have to get things wrong before you get them right. Nevertheless it is still important that OSINT research is structured and methodical. Good OSINT reporting is clear and precise, and it should be possible for anyone who reviews your findings to understand how you came to your conclusions. This means structure and planning have to be built into your research at the outset. Not only will this ensure that you don’t go off on too many tangents, but it will also ensure that your final product is much more able to withstand scrutiny.
\n
This might sound very serious for a post that is just a write-up for one of this week’s Quiztime challenges. Quiztime is for fun of course, but it is such a useful tool for practice because some of the harder challenges require the same methodical approach that more complex “real life” investigations require. It’s fun to solve a Quiztime puzzle with a first-time reverse image search, but good luck only takes you so far! This is why I always recommend people who want to learn more about OSINT try to solve Quiztime puzzles to improve their skills.
\n
With that in mind I’m going to hopefully show how a more methodical approach made it possible to solve a recent Quiztime challenge from Tilman Wagner. Here’s the original tweet:
\n
\n
Something is out of place, so this is going to be a where question. It’s also a what question – what exactly do Swapfiets do, and why did this advert cause Tilman to think something was amiss? The advert is for a bike hire and repair service called Swapfiets, and it proudly announces the roll out of their service to Bonn, Germany – but how can we be certain that the photograph is actually in Bonn, and if it isn’t in Bonn, then where is it? Ultimately we will be able to show that not only is this photo not in Bonn, but we will be able to pinpoint the exact location where the photo was taken to prove our point. Let’s start at the beginning…
\n
Gap Analysis
\n
I’ve written about using gap analysis in many previous blog posts and I make no apology for writing about it again. It’s a very easy way of bringing structure and order to research by making you identify what it is you’re actually trying to achieve and how you will actually do it. It’s tempting to try and take shortcuts and try your luck by hopping from tool to tool in the hope of a quick fix but in the long run this won’t necessarily serve you well. Gap analysis takes stock of the initial information that you have and then applies four simple questions to identify what to do next. There are plenty of other ways to plan OSINT research, but I like this one for its simplicity and scalability. The four questions are:
\n
1) What do I know? \n2) What does this mean? \n3) (So) What do I need to know? \n4) How do I find out?
\n
So we can apply this to the very limited information we start off with and soon we will be able to identify what to do next without falling down too many rabbit holes on the way. Let’s give it a try:
\n
1) What do we know?
\n
This is an advert for a bike company that purports to be in Bonn, but Tilman is a little sceptical about this.
\n
2) What does this mean?
\n
We need to know where this photo was taken so we can either confirm or eliminate whether or not Bonn is the right location.
\n
3) So what do I need to know?
\n
Where was this photograph taken.
\n
4) How do I find out?
\n
We need to gather enough information to accurately geolocate this photograph. We can start by trying to find the original source.
\n
Finding the original source of the image is not too hard. We can visit the Swapfiets Facebook Page as a starting point. Look at what we find when we get there:
\n
\n
A very familiar looking image – but unlike the one that Tilman saw, this is much larger and has not been cropped. This means that it might contain important extra details that we would have missed otherwise. We also learn that Swapfiets have a .nl website, so they are likely a Dutch company:
\n
\n
We’ll come back to the website shortly. For now we can get the full size version of the image that we need to geolocate by clicking on the banner at the top of the page:
\n
\n
\n
Much better! There’s a lot of detail here, and ultimately it will make it possible to identify the exact spot where this image was taken from. Now we can gather as much information from the photo as possible before applying another round of gap analysis to decide what we need to do after that. There’s no magic trick to extracting information from a photograph – just use your eyes! If you’re struggling then using a technique like 20 Questions can force you to pull out more detail than you might otherwise. We can start by adding just a few observations from the photo, these will then be added to the “What do I know?” part of the next round of gap analysis. Here’s a few things from the photo to get started:
\n
1) It’s a city with tall buildings.
\n
2) There are tramlines in the street.
\n
3) The car registration is partly visible, looks like PJ-620-something.
\n
4) The lamppost has black and white stripes.
\n
5) There are some unusual buildings on the left hand side of the road – one has two distinct metal columns, while the other has some kind of rectangular column sticking into the street.
\n
That’s only five points, but in this case it will probably be enough when we combine it with the information from the Swapfiets Facebook page. Time to feed what we’ve learned into another round of gap analysis:
\n
1) What Do I know?
\n
The location is a city with tall, modern, buildings, the lampposts have black and white stripes, and there are tram tracks in the street. The registration plate of the car is partly visible. It is a location where Swapfiets offer their services. Swapfiets are a Dutch company (and Fiets is a Dutch word).
\n
2) What Does This Mean?
\n
The location will be a modern city with tall buildings that has trams running in the street. Swapfiets will have a service there. We might be able to use the street furniture and the car registration plate to home in a bit further.
\n
3) So What Do I Need To Know?
\n
Which cities do Swapfiets operate in, and which of these cities still run trams in the street?
\n
4) How Do I Find Out?
\n
The Swapfiets website would be a good place to start. Wikipedia might also be a good place to find out about trams.
\n
I haven’t used every single piece of information to take this mini-investigation forward. I’ve focused on the information that will identify the country and city first. The information about street furniture will be used for fine tuning later on.
\n
Country & City
\n
The Swapfiets.nl website contains a list of all the locations where they run their services. They provide bikes for hire in the Netherlands, Belgium, Denmark, and Germany in dozens of different cities. This might seem a little daunting, but we can be confident in our investigation plan and move on without having to check all of these cities individually for matching features. Why? Because we have already identified some features that the correct location will have, and we can apply these same parameters to all the Swapfiets locations to confirm or eliminate them. One way to do this is to identify the country before moving on to a specific city.
\n
\n
\n
\n
We’ve already narrowed it down to four possible countries by using the Swapfiets site. Can we use any other information to find the correct one? The car registration plate is the most obvious place to start:
\n
\n
\n
We can use WorldLicensePlates to compare this registration plate to those of the four possible countries. These are some sample Dutch number plates to begin with:
\n
\n
\n
And some German ones:
\n
\n
And the Danish:
\n
\n
And finally the Belgian plates:
\n
\n
Of the four countries, which has the closest match to the one in the picture? Only the Dutch plates really come close, although white number plates seem to be quite rare in the Netherlands. If we’re wrong on this point it doesn’t matter too much. Because we have a structured method in place, we can always retrace our steps and start again if we go in the wrong direction. This is much better than firing random search terms into Google or loading endless images into Yandex in the hope we get lucky.
\n
So if we work on the basis that the photo was taken in the Netherlands, we can now add in the second part of our “What Do I Need To Know?” question: Which cities in the Netherlands have trams that run in the street, and how do we find out? Let’s start with Wikipedia.
\n
Trams
\n
There really is a Wikipedia page for everything, and the Dutch tram system is no exception. This page tells us all about the tram systems all across the Netherlands. From this we learn that only five towns in the Netherlands still have tram systems. They are: Delft, Utrecht, Den Haag, Amsterdam, and Rotterdam. It’s still too early to open up Google Maps and start looking at these cities for matching locations though. With the new information about cities we can conduct another quick round of gap analysis to decide what we need to do next before getting the maps out.
\n
1) What Do I Know?
\n
The photo was taken in either Delft, Utrecht, Den Haag, Amsterdam, or Rotterdam.
\n
The location is on a street with some tall buildings. Two of them have some very distinctive columns:
\nTwo of the buildings on the left have very distinctive columns.\n
\n
2) What Does This Mean?
\n
If we can find these distinctive buildings in one of the five cities, we will be able to find the location. We’ll be able to use the tramlines and striped lamppost to find the precise spot.
\n
3) What Do I Need To Know?
\n
Of the five possible cities, which one has buildings with these distinctive features?
\n
4) How Do I Find Out?
\n
We’re going to use Phorio and Skyscraper Page to get a quick overview of the most prominent buildings in each of the five cities.
\n
Phorio & Skyscraper Page
\n
Phorio and Skyscraper Page are brilliant tools for geolocation challenges. Phorio contains all kinds of data relating to large buildings in cities all across the world, including height, map, and photo information. It also usually features skyline photos and high quality pictures of the key buildings in a particular city. Skyscraper page contains sketches of all the key buildings in a city and arranges them by height, so it’s great for identifying specific buildings and skyline features. To solve Tilman’s quiz we can go through the Phorio and Skyscraper Page sites for each of the five possible cities we have identified to see which one matches the buildings in the photo.
\n
Let’s use Delft as an example. Here’s the Phorio page:
\n
\n
Complete with pictures of all the most prominent buildings in the city:
I can’t see any matches for the buildings in the picture. Something else is not quite right here – Delft may have trams but it does not really seem to have the big commercial buildings like those in the picture. Its largest buildings are either the university or residential blocks. I suspect Delft is too small to be the city in question. Utrecht is a little bigger and has a few larger commercial buildings, but there are still no obvious matches. It must be either Amsterdam, Den Haag, or Rotterdam.
\n
These cities are much larger and so it takes slightly longer to check their Phorio profiles, but I still find this much quicker than using Google Maps. For more refined searching, Phorio allows the selection of different filters such as building type or district. These are the available filters for Rotterdam:
\n
\n
Of all the different building types (or “channels”), it seems most likely that our location is dominated by office buildings, so we will just look the those for now. It gives a nice selection of buildings to focus on. This is a very efficient way to check a large number of buildings at one time. Looking through the list, there is one specific building that stands out.
\n
\n
The Unilever Building has the same distinctive metal columns as the building in the original photo. We need to check if it’s the right one.
\n
\n
\n
The building has its own Phorio and Skyscraperpage entries, so we can find out its address too. It’s located at Weena 455, Rotterdam. This is what it looks like in Google Maps 3D:
\n
\n
\n
A final look with Streetview just to check – it looks pretty good!
\n
\n
All the features we identified throughout are present here: the tramlines, the building with metal columns, all the same tall tower blocks, the striped lampposts and all the other street details all match. This means that the photo was taken looking down the Delfste Poort in Rotterdam. This explains why Tilman wasn’t entirely convinced by the original picture:
\n
\n
This was by no means the only way to have solved this quiz. There is usually more than one solution but if you get a little stuck or aren’t sure what to do next then gap analysis is an excellent way of making things a little more clear and structured. If you enjoyed reading this then check out some of my other Quiztime write-ups and don’t forget to follow @Quiztime to take part in the daily quiz.
",
+ "content_text": "In OSINT we make use of the wealth of information that is now available to us, but so often it becomes the case that there is too much information for us to process effectively. It’s easy to drown in a sea of data and you can quickly find yourself with twenty browser tabs open having completely forgotten what it was that you originally set out to do. Going down some rabbit holes is inevitable because often you can only know that you were on the right track with the benefit of hindsight and you have to get things wrong before you get them right. Nevertheless it is still important that OSINT research is structured and methodical. Good OSINT reporting is clear and precise, and it should be possible for anyone who reviews your findings to understand how you came to your conclusions. This means structure and planning have to be built into your research at the outset. Not only will this ensure that you don’t go off on too many tangents, but it will also ensure that your final product is much more able to withstand scrutiny.\n\nThis might sound very serious for a post that is just a write-up for one of this week’s Quiztime challenges. Quiztime is for fun of course, but it is such a useful tool for practice because some of the harder challenges require the same methodical approach that more complex “real life” investigations require. It’s fun to solve a Quiztime puzzle with a first-time reverse image search, but good luck only takes you so far! This is why I always recommend people who want to learn more about OSINT try to solve Quiztime puzzles to improve their skills.\n\nWith that in mind I’m going to hopefully show how a more methodical approach made it possible to solve a recent Quiztime challenge from Tilman Wagner. Here’s the original tweet:\n\nSomething is out of place, so this is going to be a where question. It’s also a what question – what exactly do Swapfiets do, and why did this advert cause Tilman to think something was amiss? The advert is for a bike hire and repair service called Swapfiets, and it proudly announces the roll out of their service to Bonn, Germany – but how can we be certain that the photograph is actually in Bonn, and if it isn’t in Bonn, then where is it? Ultimately we will be able to show that not only is this photo not in Bonn, but we will be able to pinpoint the exact location where the photo was taken to prove our point. Let’s start at the beginning…\n\nGap Analysis\n\nI’ve written about using gap analysis in many previous blog posts and I make no apology for writing about it again. It’s a very easy way of bringing structure and order to research by making you identify what it is you’re actually trying to achieve and how you will actually do it. It’s tempting to try and take shortcuts and try your luck by hopping from tool to tool in the hope of a quick fix but in the long run this won’t necessarily serve you well. Gap analysis takes stock of the initial information that you have and then applies four simple questions to identify what to do next. There are plenty of other ways to plan OSINT research, but I like this one for its simplicity and scalability. The four questions are:\n\n1) What do I know?\n\n2) What does this mean?\n\n3) (So) What do I need to know?\n\n4) How do I find out?\n\nSo we can apply this to the very limited information we start off with and soon we will be able to identify what to do next without falling down too many rabbit holes on the way. Let’s give it a try:\n\n1) What do we know?\n\nThis is an advert for a bike company that purports to be in Bonn, but Tilman is a little sceptical about this.\n\n2) What does this mean?\n\nWe need to know where this photo was taken so we can either confirm or eliminate whether or not Bonn is the right location.\n\n3) So what do I need to know?\n\nWhere was this photograph taken.\n\n4) How do I find out?\n\nWe need to gather enough information to accurately geolocate this photograph. We can start by trying to find the original source.\n\nFinding the original source of the image is not too hard. We can visit the Swapfiets Facebook Page as a starting point. Look at what we find when we get there:\n\nA very familiar looking image – but unlike the one that Tilman saw, this is much larger and has not been cropped. This means that it might contain important extra details that we would have missed otherwise. We also learn that Swapfiets have a .nl website, so they are likely a Dutch company:\n\nWe’ll come back to the website shortly. For now we can get the full size version of the image that we need to geolocate by clicking on the banner at the top of the page:\n\nMuch better! There’s a lot of detail here, and ultimately it will make it possible to identify the exact spot where this image was taken from. Now we can gather as much information from the photo as possible before applying another round of gap analysis to decide what we need to do after that. There’s no magic trick to extracting information from a photograph – just use your eyes! If you’re struggling then using a technique like 20 Questions can force you to pull out more detail than you might otherwise. We can start by adding just a few observations from the photo, these will then be added to the “What do I know?” part of the next round of gap analysis. Here’s a few things from the photo to get started:\n\n1) It’s a city with tall buildings.\n\n2) There are tramlines in the street.\n\n3) The car registration is partly visible, looks like PJ-620-something.\n\n4) The lamppost has black and white stripes.\n\n5) There are some unusual buildings on the left hand side of the road – one has two distinct metal columns, while the other has some kind of rectangular column sticking into the street.\n\nThat’s only five points, but in this case it will probably be enough when we combine it with the information from the Swapfiets Facebook page. Time to feed what we’ve learned into another round of gap analysis:\n\n1) What Do I know?\n\nThe location is a city with tall, modern, buildings, the lampposts have black and white stripes, and there are tram tracks in the street. The registration plate of the car is partly visible. It is a location where Swapfiets offer their services. Swapfiets are a Dutch company (and Fiets is a Dutch word).\n\n2) What Does This Mean?\n\nThe location will be a modern city with tall buildings that has trams running in the street. Swapfiets will have a service there. We might be able to use the street furniture and the car registration plate to home in a bit further.\n\n3) So What Do I Need To Know?\n\nWhich cities do Swapfiets operate in, and which of these cities still run trams in the street?\n\n4) How Do I Find Out?\n\nThe Swapfiets website would be a good place to start. Wikipedia might also be a good place to find out about trams.\n\nI haven’t used every single piece of information to take this mini-investigation forward. I’ve focused on the information that will identify the country and city first. The information about street furniture will be used for fine tuning later on.\n\nCountry & City\n\nThe Swapfiets.nl website contains a list of all the locations where they run their services. They provide bikes for hire in the Netherlands, Belgium, Denmark, and Germany in dozens of different cities. This might seem a little daunting, but we can be confident in our investigation plan and move on without having to check all of these cities individually for matching features. Why? Because we have already identified some features that the correct location will have, and we can apply these same parameters to all the Swapfiets locations to confirm or eliminate them. One way to do this is to identify the country before moving on to a specific city.\n\nWe’ve already narrowed it down to four possible countries by using the Swapfiets site. Can we use any other information to find the correct one? The car registration plate is the most obvious place to start:\n\nWe can use WorldLicensePlates to compare this registration plate to those of the four possible countries. These are some sample Dutch number plates to begin with:\n\nAnd some German ones:\n\nAnd the Danish:\n\nAnd finally the Belgian plates:\n\nOf the four countries, which has the closest match to the one in the picture? Only the Dutch plates really come close, although white number plates seem to be quite rare in the Netherlands. If we’re wrong on this point it doesn’t matter too much. Because we have a structured method in place, we can always retrace our steps and start again if we go in the wrong direction. This is much better than firing random search terms into Google or loading endless images into Yandex in the hope we get lucky.\n\nSo if we work on the basis that the photo was taken in the Netherlands, we can now add in the second part of our “What Do I Need To Know?” question: Which cities in the Netherlands have trams that run in the street, and how do we find out? Let’s start with Wikipedia.\n\nTrams\n\nThere really is a Wikipedia page for everything, and the Dutch tram system is no exception. This page tells us all about the tram systems all across the Netherlands. From this we learn that only five towns in the Netherlands still have tram systems. They are: Delft, Utrecht, Den Haag, Amsterdam, and Rotterdam. It’s still too early to open up Google Maps and start looking at these cities for matching locations though. With the new information about cities we can conduct another quick round of gap analysis to decide what we need to do next before getting the maps out.\n\n1) What Do I Know?\n\nThe photo was taken in either Delft, Utrecht, Den Haag, Amsterdam, or Rotterdam.\n\nThe location is on a street with some tall buildings. Two of them have some very distinctive columns:\n\nTwo of the buildings on the left have very distinctive columns.\n\n2) What Does This Mean?\n\nIf we can find these distinctive buildings in one of the five cities, we will be able to find the location. We’ll be able to use the tramlines and striped lamppost to find the precise spot.\n\n3) What Do I Need To Know?\n\nOf the five possible cities, which one has buildings with these distinctive features?\n\n4) How Do I Find Out?\n\nWe’re going to use Phorio and Skyscraper Page to get a quick overview of the most prominent buildings in each of the five cities.\n\nPhorio & Skyscraper Page\n\nPhorio and Skyscraper Page are brilliant tools for geolocation challenges. Phorio contains all kinds of data relating to large buildings in cities all across the world, including height, map, and photo information. It also usually features skyline photos and high quality pictures of the key buildings in a particular city. Skyscraper page contains sketches of all the key buildings in a city and arranges them by height, so it’s great for identifying specific buildings and skyline features. To solve Tilman’s quiz we can go through the Phorio and Skyscraper Page sites for each of the five possible cities we have identified to see which one matches the buildings in the photo.\n\nLet’s use Delft as an example. Here’s the Phorio page:\n\nComplete with pictures of all the most prominent buildings in the city:\n\nThere are no immediately obvious matches here. This is what Delft looks like on Skyscraper Page:\n\nI can’t see any matches for the buildings in the picture. Something else is not quite right here – Delft may have trams but it does not really seem to have the big commercial buildings like those in the picture. Its largest buildings are either the university or residential blocks. I suspect Delft is too small to be the city in question. Utrecht is a little bigger and has a few larger commercial buildings, but there are still no obvious matches. It must be either Amsterdam, Den Haag, or Rotterdam.\n\nThese cities are much larger and so it takes slightly longer to check their Phorio profiles, but I still find this much quicker than using Google Maps. For more refined searching, Phorio allows the selection of different filters such as building type or district. These are the available filters for Rotterdam:\n\nOf all the different building types (or “channels”), it seems most likely that our location is dominated by office buildings, so we will just look the those for now. It gives a nice selection of buildings to focus on. This is a very efficient way to check a large number of buildings at one time. Looking through the list, there is one specific building that stands out.\n\nThe Unilever Building has the same distinctive metal columns as the building in the original photo. We need to check if it’s the right one.\n\nThe building has its own Phorio and Skyscraperpage entries, so we can find out its address too. It’s located at Weena 455, Rotterdam. This is what it looks like in Google Maps 3D:\n\nA final look with Streetview just to check – it looks pretty good!\n\nAll the features we identified throughout are present here: the tramlines, the building with metal columns, all the same tall tower blocks, the striped lampposts and all the other street details all match. This means that the photo was taken looking down the Delfste Poort in Rotterdam. This explains why Tilman wasn’t entirely convinced by the original picture:\n\nThis was by no means the only way to have solved this quiz. There is usually more than one solution but if you get a little stuck or aren’t sure what to do next then gap analysis is an excellent way of making things a little more clear and structured. If you enjoyed reading this then check out some of my other Quiztime write-ups and don’t forget to follow @Quiztime to take part in the daily quiz.\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_001.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_002.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_003.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/72551738_2538535489593992_7739630611849019392_o.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_064.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_065.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_005-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_009-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_010-4.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_006-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_007.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_008-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/72551738_2538535489593992_7739630611849019392_o-annotated.jpg",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_011-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_012-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_013-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_014.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_015-1-2048x807.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_017.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_019-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_018.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/03/Selection_020-1.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2020-05-12_chronolocation-clues-quiztime-11th-may-2020.json b/nixintel_geolocation_json/2020-05-12_chronolocation-clues-quiztime-11th-may-2020.json
new file mode 100644
index 0000000000000000000000000000000000000000..f6b2347109f4a1a85916a92a9f2e754958ed95c7
--- /dev/null
+++ b/nixintel_geolocation_json/2020-05-12_chronolocation-clues-quiztime-11th-may-2020.json
@@ -0,0 +1,118 @@
+{
+ "slug": "chronolocation-clues-quiztime-11th-may-2020",
+ "title": "Chronolocation Clues – Quiztime 11th May 2020",
+ "url": "https://nixintel.info/osint/chronolocation-clues-quiztime-11th-may-2020/",
+ "date": "2020-05-12T18:58:18+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
The question we need to answer is whether or not this photo was taken before or after lockdown. To do this we need to know where the image was taken so we can research the relevant country’s lockdown start and end dates. Fortunately in this case the geolocation is quite easy; we can see a sign for the Alexanderplatz U-Bahn station so we know this was taken in Berlin, from approximately here:
\n
\n
\n
The lockdown in Berlin was implemented in different sectors from about 19th March 2020 onwards. The lockdown is being gradually lifted there as of 6th May 2020, so we need to look for information in Julia’s photo that will help to decide if this image was taken before 19th March or after 6th May. To do this we will extract as much detail as we can from the image, see what these details can tell us about times and dates, and then assess whether these details mean it was more likely that the photo was taken before or after lockdown. Let’s look at the first clue:
\n
1. The Trees
\n
\n
There are two trees visible on the left side of the image. They have no leaves on them so it might be better to assume the photo was taken before lockdown, and not afterwards. Case closed, too easy – right? Not necessarily. Assuming is always dangerous and with more complex OSINT puzzles it will catch you out if you take things for granted without checking them first. It’s much more important to be able to find proof to support an argument rather than just assume it and hope for the best. So how can we do that? There are a few ways, but one of the most direct is to access a live feed of one of the many webcams that broadcast live from Alexanderplatz. This feed transmits from the Park Inn Hotel (which is just out of shot in Julia’s original photo) and also contains an archive of time lapse footage (accessed from the pop-up menu on the right of the screen). This means we can check what the trees in Alexanderplatz looked like on 6th May, when the Berlin lockdown began to be lifted:
\n
\n
We can show that on the day lockdown ended the trees were green and full of leaves, so now we know (rather than just assume) that the photo was taken before the lockdown started.
\n
2. Cold Weather Clothing
\n
\n
What are people in the picture wearing? Most of them are wearing thick coats and a few have hats and scarves. Once again it would be easy to automatically assume that this means the photo was taken in the colder weather before lockdown started. It wise to be cautious with weather though; last weekend I got sunburn while out in my garden but three days later and it’s freezing cold, so trying to judge a date from a single snapshot in time could be risky. It’s much better to try and find a reliable source of information to back up our hypothesis. Using Wolfram Alpha’s historic weather database we can see that the weather in Berlin has been fairly mild since lockdown ended:
\n
\n
Now compare it to the week leading up to lockdown when it was much colder:
\n
\n
Although we don’t know that this image was taken in the week before lockdown, we can show objectively that it was consistently colder in the time before lockdown than during the time since it was lifted, so the coats and hats indicate the image was most likely taken before lockdown. This might seem like an excessive way to prove a simple point, but the aim is to cultivate good verification habits. Most real-life OSINT puzzles are rarely so obvious, and it is always necessary to provide evidence to support each part of your case. Arguments made without evidence can be dismissed without evidence, so it is important to be able to prove even the small things.
\n
3. The C&A Sale
\n
The trees and seasonal clothing might be the easiest way to chronolocate this image, but what other clues are there? On the upper left of the photo we can see the Alexanderplatz branch of C&A. It appears to be having a sale, and a quick check of the C&A website shows that they’re having a sale now too – so could this be of any use? It’s tricky because retailers have sales all the time, so the fact there is a sale does not prove whether this photo was taken before or after lockdown. We need to dig a little deeper first.. The C&A website has some current information about which of its stores have reopened since the Coronavirus lockdown began to end:
\n
\n
None of their German stores are open yet, so it seems unlikely that this image was taken after the lockdown ended. Researching the C&A hashtag #BestDeal also isn’t too helpful. It’s been one of their marketing slogans for several years so it by itself it doesn’t help to narrow down when the image was taken. We could at least set some time parameters for the earliest possible date this image was taken. It’s beyond the remit of the quiz, but it would be useful to see if we can say that the image was take no earlier/later than a specific date. The marketing material on display in the C&A window can help a little with this. Here it is as it appears in Julia’s photo:
\n
\n
We can check images that people have uploaded from the area in front of the C&A store by visiting the location on Google Maps and see if the same material appears. From this image we can see that these same adverts were in place as far back as August 2019:
\n
\n
I couldn’t find any matching images earlier than that date, so it must have been present since at least then.
\n
4. The Photographer’s Position?
\n
From the angle of the picture we can work out that Julia took the photo from Alexanderplatz 3, which houses the Saturn electronics store:
So this means it is likely that the image was taken in early 2020 before the lockdown began when Galeria Kaufhof were heavily promoting their new loyalty card. I’d be slightly more cautious about using this ad to answer the pre/post lockdown question though. The Galeria Kaufhof app is still available and was updated as recently as May 4th, so the fact that it was launched before lockdown does not automatically mean that the ad would disappear when lockdown ended. By itself this advert cannot be conclusive, but when we put it together with the other details it tends to support the idea that the image was taken before lockdown.
The water has been turned back on, so again this tends to suggest that the original image was more likely to be taken before the lockdown rather than afterwards.
\n
7. The Bus
\n
\n
There’s an orange and green bus reflected in the window as Julia took the photo. A little Google image searching for orange and green bus Berlin quickly leads to the travel company Flixbus:
\n
\n
Their website tells us that they suspended their operations on March 17th this year:
\n
\n
As of today, it seems that Flixbus is still suspended. Although they have made a few journeys to help return people to their home countries during lockdown they still haven’t resumed operations so the presence of a Flixbus in central Berlin is a strong indicator the image was taken before lockdown started.
Contrast this with the behaviour of the people in the photograph:
\n
\n
\n
Not a single face mask to be seen, and not much social distancing either. While there is a minority of people in Germany who are opposed to the lockdown regulations, such widespread and casual disregard for rules and social customs would be very un-German, wouldn’t it? The only sensible conclusion is that this photo was taken before the Coronavirus restrictions were put in place and masks and distancing were not a concern.
",
+ "content_text": "Not all image-based OSINT work is about geolocation. It is always helpful to know where a photo was taken, but knowing when an image was taken can be just as important when trying to verify it. Is this a COVID-hit Italian hospital, or the aftermath of an earthquake in Croatia? Are these food packages left for Coronavirus survivors in India or are they really from a Turkish earthquake that took place before the pandemic? If it is possible to establish when a photo was taken then the easier it will be to determine if it really shows what it claims to. Yesterday’s Quiztime challenge from Julia Bayer was a good opportunity to practise some of the skills that help to identify when a photo was taken. Here’s the photo:\n\nThe question we need to answer is whether or not this photo was taken before or after lockdown. To do this we need to know where the image was taken so we can research the relevant country’s lockdown start and end dates. Fortunately in this case the geolocation is quite easy; we can see a sign for the Alexanderplatz U-Bahn station so we know this was taken in Berlin, from approximately here:\n\nThe lockdown in Berlin was implemented in different sectors from about 19th March 2020 onwards. The lockdown is being gradually lifted there as of 6th May 2020, so we need to look for information in Julia’s photo that will help to decide if this image was taken before 19th March or after 6th May. To do this we will extract as much detail as we can from the image, see what these details can tell us about times and dates, and then assess whether these details mean it was more likely that the photo was taken before or after lockdown. Let’s look at the first clue:\n\n1. The Trees\n\nThere are two trees visible on the left side of the image. They have no leaves on them so it might be better to assume the photo was taken before lockdown, and not afterwards. Case closed, too easy – right? Not necessarily. Assuming is always dangerous and with more complex OSINT puzzles it will catch you out if you take things for granted without checking them first. It’s much more important to be able to find proof to support an argument rather than just assume it and hope for the best. So how can we do that? There are a few ways, but one of the most direct is to access a live feed of one of the many webcams that broadcast live from Alexanderplatz. This feed transmits from the Park Inn Hotel (which is just out of shot in Julia’s original photo) and also contains an archive of time lapse footage (accessed from the pop-up menu on the right of the screen). This means we can check what the trees in Alexanderplatz looked like on 6th May, when the Berlin lockdown began to be lifted:\n\nWe can show that on the day lockdown ended the trees were green and full of leaves, so now we know (rather than just assume) that the photo was taken before the lockdown started.\n\n2. Cold Weather Clothing\n\nWhat are people in the picture wearing? Most of them are wearing thick coats and a few have hats and scarves. Once again it would be easy to automatically assume that this means the photo was taken in the colder weather before lockdown started. It wise to be cautious with weather though; last weekend I got sunburn while out in my garden but three days later and it’s freezing cold, so trying to judge a date from a single snapshot in time could be risky. It’s much better to try and find a reliable source of information to back up our hypothesis. Using Wolfram Alpha’s historic weather database we can see that the weather in Berlin has been fairly mild since lockdown ended:\n\nNow compare it to the week leading up to lockdown when it was much colder:\n\nAlthough we don’t know that this image was taken in the week before lockdown, we can show objectively that it was consistently colder in the time before lockdown than during the time since it was lifted, so the coats and hats indicate the image was most likely taken before lockdown. This might seem like an excessive way to prove a simple point, but the aim is to cultivate good verification habits. Most real-life OSINT puzzles are rarely so obvious, and it is always necessary to provide evidence to support each part of your case. Arguments made without evidence can be dismissed without evidence, so it is important to be able to prove even the small things.\n\n3. The C&A Sale\n\nThe trees and seasonal clothing might be the easiest way to chronolocate this image, but what other clues are there? On the upper left of the photo we can see the Alexanderplatz branch of C&A. It appears to be having a sale, and a quick check of the C&A website shows that they’re having a sale now too – so could this be of any use? It’s tricky because retailers have sales all the time, so the fact there is a sale does not prove whether this photo was taken before or after lockdown. We need to dig a little deeper first.. The C&A website has some current information about which of its stores have reopened since the Coronavirus lockdown began to end:\n\nNone of their German stores are open yet, so it seems unlikely that this image was taken after the lockdown ended. Researching the C&A hashtag #BestDeal also isn’t too helpful. It’s been one of their marketing slogans for several years so it by itself it doesn’t help to narrow down when the image was taken. We could at least set some time parameters for the earliest possible date this image was taken. It’s beyond the remit of the quiz, but it would be useful to see if we can say that the image was take no earlier/later than a specific date. The marketing material on display in the C&A window can help a little with this. Here it is as it appears in Julia’s photo:\n\nWe can check images that people have uploaded from the area in front of the C&A store by visiting the location on Google Maps and see if the same material appears. From this image we can see that these same adverts were in place as far back as August 2019:\n\nI couldn’t find any matching images earlier than that date, so it must have been present since at least then.\n\n4. The Photographer’s Position?\n\nFrom the angle of the picture we can work out that Julia took the photo from Alexanderplatz 3, which houses the Saturn electronics store:\n\nUnlike C&A, Saturn have reopened their store in Alexanderplatz, so the fact that Julia was inside this building does not help to confirm whether this picture was taken pre or post-lockdown,\n\n5. The Galeria Kaufhof Advert\n\nIt’s a small detail, but there’s an advert on the right of the picture that says “DIE ODER KEINE” (“This or none at all”). It’s a promotional message for Galeria Kaufhof’s new loyalty card and app. The company withdrew from their previous reward scheme in January this year and this advert for a replacement reward card and app was published from early January onwards:\n\nSo this means it is likely that the image was taken in early 2020 before the lockdown began when Galeria Kaufhof were heavily promoting their new loyalty card. I’d be slightly more cautious about using this ad to answer the pre/post lockdown question though. The Galeria Kaufhof app is still available and was updated as recently as May 4th, so the fact that it was launched before lockdown does not automatically mean that the ad would disappear when lockdown ended. By itself this advert cannot be conclusive, but when we put it together with the other details it tends to support the idea that the image was taken before lockdown.\n\n6. The Fountain\n\nIn the original image the fountain is dry and the water appears to be turned off, indicating it is out of use. However a quick search for recent images of Alexanderplatz on Twitter shows this image from May 9th, just three days after the lockdown restrictions were lifted (via @PaulGaebler):\n\nThe water has been turned back on, so again this tends to suggest that the original image was more likely to be taken before the lockdown rather than afterwards.\n\n7. The Bus\n\nThere’s an orange and green bus reflected in the window as Julia took the photo. A little Google image searching for orange and green bus Berlin quickly leads to the travel company Flixbus:\n\nTheir website tells us that they suspended their operations on March 17th this year:\n\nAs of today, it seems that Flixbus is still suspended. Although they have made a few journeys to help return people to their home countries during lockdown they still haven’t resumed operations so the presence of a Flixbus in central Berlin is a strong indicator the image was taken before lockdown started.\n\n8. Crowd Behaviour\n\nEven though lockdown in Germany is now ending, there are a lot of restrictions still in place, including keeping a distance of 1.5 metres from other people and wearing face masks when out shopping:\n\nContrast this with the behaviour of the people in the photograph:\n\nNot a single face mask to be seen, and not much social distancing either. While there is a minority of people in Germany who are opposed to the lockdown regulations, such widespread and casual disregard for rules and social customs would be very un-German, wouldn’t it? The only sensible conclusion is that this photo was taken before the Coronavirus restrictions were put in place and masks and distancing were not a concern.\n\nFor my other Quiztime posts, click here.\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_001.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_011.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_005.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_013-2048x1177.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_008.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_014.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_015.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_016.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_017.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_018.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_020.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_021.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_022.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_023.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_024.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_025.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_026.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_027.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_028.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_029.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_010.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2020-05-15_lost-in-the-woods-translation-troubles-quiztime-10th-march-2020.json b/nixintel_geolocation_json/2020-05-15_lost-in-the-woods-translation-troubles-quiztime-10th-march-2020.json
new file mode 100644
index 0000000000000000000000000000000000000000..7510e6fa226c349417baa090cc62e304ac844854
--- /dev/null
+++ b/nixintel_geolocation_json/2020-05-15_lost-in-the-woods-translation-troubles-quiztime-10th-march-2020.json
@@ -0,0 +1,63 @@
+{
+ "slug": "lost-in-the-woods-translation-troubles-quiztime-10th-march-2020",
+ "title": "Lost In The Woods & Translation Troubles – Quiztime 10th March 2020",
+ "url": "https://nixintel.info/osint/lost-in-the-woods-translation-troubles-quiztime-10th-march-2020/",
+ "date": "2020-05-15T15:36:29+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
I started writing up this Quiztime post almost two months ago but then life got very busy and I never managed to finish it. This was a really fun quiz by Lars Wienand and the solution was not as obvious as I first thought it was so a little extra digging was required. Here’s the original photo that Lars posted:
\n
\n
The two questions we have to answer are:
\n
1) Where was this taken?
\n
2) What does the sign say?
\n
It was not immediately clear where the picture was taken so I focused on the sign. It’s obviously not written in a western language, but I didn’t recognise which language it was. It wasn’t Chinese, Arabic, Urdu, or any other non-Roman language that I could find, and I was slightly suspicious of its origins. I realised that there might still be other languages that I could have missed, so to avoid any doubts I decided to make use of Google Translate’s ability to recognise characters in an image and translate them to English (or any other language for that matter). I saved the image to my phone and then uploaded it to Google Translate. Sure enough the app picked out the strange characters:
\n
\n
Since I didn’t know what the language was I set it to Auto Detect. Here’s what Google Translate thought the sign said:
\n
\n
Hmm. My suspicions were right – neither I nor Google recognise this strange language, so it is probably something very niche and specific. There was not much else I could do with this text for now, so I decided to try a different approach.
\n
Incidentally if you want to see a good example of how the Google Translate technique can work for geolocation, have a read through this thread where Matthias Cantow used it to solve another OSINT challenge by translating some Japanese street signs.
\n
Crop and Search
\n
The sign wasn’t helping to locate the image, and since it wasn’t helping I decided to ignore it. Instead I focused on the other half of the image featuring the little wooden huts. If you do not have any success with a reverse image search sometimes it can be more helpful to crop parts of the image and just focus on those specific details. By far the best tool to do this is with the browser addon Search By Image. I’ve written about how to use it here and also made a short video guide with another Quiztime example here.
\n
I cropped out the huts and used Yandex to conduct a reverse image search:
\nTaking a screenshot of yourself taking a screenshot is harder than you might think.\n
Yandex is still the best search engine for matching images by visual similarity, and it came up with some promising matches straight away:
\n
\n
The first image on the left bore a strong resemblance to the huts in the quiz picture, so it was the best place to start exploring. Here’s a bigger version of the image:
\n
\n
A further reverse image search on this image led to this page, which says the huts are part of the accommodation at Waldsiedlum in Neißeaue, Germany. A quick search in Google Maps brings up the location here.
\n
The World Of Turisede
\n
So the Waldsiedlum has lots of strange little wooden huts like those in the picture, but that doesn’t mean it is the correct location because small wooden huts are hardly a unique feature. Besides there is still the issue of the strange language; it is definitely not German and we can’t just ignore it the parts of the puzzle that don’t fit. The best way forward was to explore the Waldsiedlum in more detail to see if it took me any further. If it didn’t I could always retrace my steps and go in a different direction.
\n
A quick Google search lead to the official site for the Waldsiedlum, or as it is better known, Turisede.
\n
Turisede is a woodland theme park full of huts, treehouses, and other attractions. More importantly, as I browsed through the website looking for more information I came across a few images that referred to “Ronkads Reimekaub” (Ronkad’s Rhyme Robbery). I have no idea who Ronkad is, but look at the strange text that accompanies each reference to him:
\n
\n
It’s the same strange language that appears on the sign in the original photo that Lars took. This means that Turisede is now almost certainly the correct location, but also that somewhere in this website there must be more information about these strange symbols. Google allows you to search for terms within a specific site, so I decided to search for other references to “Ronkads Reimeraub” all across the Turisede.com site. The search term goes like this:
\n
site:turisede.com “ronkads reimeraub”
\n
I am more interested in image matches rather than text matches because it is unlikely that this strange language is indexable by Google. In every instance of it I found on the website the “text” was actually a JPG file, so an image search will be better. Here are the results:
\n
\n
The Rosetta Stone
\n
Not only do we have more examples of the mystery text, but there is also what looks like a translation guide – bingo!
\n
\n
Finally it all becomes clear: the reason that neither I nor Google could understand the strange language is because it is apparently written by Trolls. We are told that Trolls have their own alphabet called “Troll Runes” but fortunately the residents of Turisede have made a handy translation guide for us. Now we can finally translate the sign to see what it says:
\n
\n
It reads: GUNTHERS HOLZWEG (“Gunther’s Wooden Path”). Mystery solved!
\n
If you enjoyed reading this some of my other Quiztime solutions can be found here.
",
+ "content_text": "I started writing up this Quiztime post almost two months ago but then life got very busy and I never managed to finish it. This was a really fun quiz by Lars Wienand and the solution was not as obvious as I first thought it was so a little extra digging was required. Here’s the original photo that Lars posted:\n\nThe two questions we have to answer are:\n\n1) Where was this taken?\n\n2) What does the sign say?\n\nIt was not immediately clear where the picture was taken so I focused on the sign. It’s obviously not written in a western language, but I didn’t recognise which language it was. It wasn’t Chinese, Arabic, Urdu, or any other non-Roman language that I could find, and I was slightly suspicious of its origins. I realised that there might still be other languages that I could have missed, so to avoid any doubts I decided to make use of Google Translate’s ability to recognise characters in an image and translate them to English (or any other language for that matter). I saved the image to my phone and then uploaded it to Google Translate. Sure enough the app picked out the strange characters:\n\nSince I didn’t know what the language was I set it to Auto Detect. Here’s what Google Translate thought the sign said:\n\nHmm. My suspicions were right – neither I nor Google recognise this strange language, so it is probably something very niche and specific. There was not much else I could do with this text for now, so I decided to try a different approach.\n\nIncidentally if you want to see a good example of how the Google Translate technique can work for geolocation, have a read through this thread where Matthias Cantow used it to solve another OSINT challenge by translating some Japanese street signs.\n\nCrop and Search\n\nThe sign wasn’t helping to locate the image, and since it wasn’t helping I decided to ignore it. Instead I focused on the other half of the image featuring the little wooden huts. If you do not have any success with a reverse image search sometimes it can be more helpful to crop parts of the image and just focus on those specific details. By far the best tool to do this is with the browser addon Search By Image. I’ve written about how to use it here and also made a short video guide with another Quiztime example here.\n\nI cropped out the huts and used Yandex to conduct a reverse image search:\n\nTaking a screenshot of yourself taking a screenshot is harder than you might think.\n\nYandex is still the best search engine for matching images by visual similarity, and it came up with some promising matches straight away:\n\nThe first image on the left bore a strong resemblance to the huts in the quiz picture, so it was the best place to start exploring. Here’s a bigger version of the image:\n\nA further reverse image search on this image led to this page, which says the huts are part of the accommodation at Waldsiedlum in Neißeaue, Germany. A quick search in Google Maps brings up the location here.\n\nThe World Of Turisede\n\nSo the Waldsiedlum has lots of strange little wooden huts like those in the picture, but that doesn’t mean it is the correct location because small wooden huts are hardly a unique feature. Besides there is still the issue of the strange language; it is definitely not German and we can’t just ignore it the parts of the puzzle that don’t fit. The best way forward was to explore the Waldsiedlum in more detail to see if it took me any further. If it didn’t I could always retrace my steps and go in a different direction.\n\nA quick Google search lead to the official site for the Waldsiedlum, or as it is better known, Turisede.\n\nTurisede is a woodland theme park full of huts, treehouses, and other attractions. More importantly, as I browsed through the website looking for more information I came across a few images that referred to “Ronkads Reimekaub” (Ronkad’s Rhyme Robbery). I have no idea who Ronkad is, but look at the strange text that accompanies each reference to him:\n\nIt’s the same strange language that appears on the sign in the original photo that Lars took. This means that Turisede is now almost certainly the correct location, but also that somewhere in this website there must be more information about these strange symbols. Google allows you to search for terms within a specific site, so I decided to search for other references to “Ronkads Reimeraub” all across the Turisede.com site. The search term goes like this:\n\nsite:turisede.com “ronkads reimeraub”\n\nI am more interested in image matches rather than text matches because it is unlikely that this strange language is indexable by Google. In every instance of it I found on the website the “text” was actually a JPG file, so an image search will be better. Here are the results:\n\nThe Rosetta Stone\n\nNot only do we have more examples of the mystery text, but there is also what looks like a translation guide – bingo!\n\nFinally it all becomes clear: the reason that neither I nor Google could understand the strange language is because it is apparently written by Trolls. We are told that Trolls have their own alphabet called “Troll Runes” but fortunately the residents of Turisede have made a handy translation guide for us. Now we can finally translate the sign to see what it says:\n\nIt reads: GUNTHERS HOLZWEG (“Gunther’s Wooden Path”). Mystery solved!\n\nIf you enjoyed reading this some of my other Quiztime solutions can be found here.\n\nShare this:\n\nTweet\n\nShare on Telegram (Opens in new window)\nTelegram\n\nShare on WhatsApp (Opens in new window)\nWhatsApp\n\nShare on Mastodon (Opens in new window)\nMastodon\n\nEmail a link to a friend (Opens in new window)\nEmail\n\nRelated",
+ "images": [
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_001-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Screenshot_20200515-092133.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Screenshot_20200515-092201.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_009.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_004.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_005-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_006.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_007-2048x601.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_008-1.png",
+ "alt": "",
+ "title": ""
+ },
+ {
+ "url": "https://nixintel.info/wp-content/uploads/2020/05/Selection_010-1.png",
+ "alt": "",
+ "title": ""
+ }
+ ]
+}
diff --git a/nixintel_geolocation_json/2022-11-03_quiztime-how-to-clean-pictures-for-better-search-results.json b/nixintel_geolocation_json/2022-11-03_quiztime-how-to-clean-pictures-for-better-search-results.json
new file mode 100644
index 0000000000000000000000000000000000000000..9fe3687a26d6e291ac83560bb6699ff340e883ac
--- /dev/null
+++ b/nixintel_geolocation_json/2022-11-03_quiztime-how-to-clean-pictures-for-better-search-results.json
@@ -0,0 +1,128 @@
+{
+ "slug": "quiztime-how-to-clean-pictures-for-better-search-results",
+ "title": "Quiztime: How To Clean Pictures For Better Search Results",
+ "url": "https://nixintel.info/osint/quiztime-how-to-clean-pictures-for-better-search-results/",
+ "date": "2022-11-03T11:49:32+00:00",
+ "author": "Nixintel",
+ "source": "NixIntel",
+ "task_type": "geolocation_reasoning",
+ "content_html": "
In last week’s Wednesday Quiz Tilman Wagner posted this photo and asked us several questions about it:
1) Who posted it and where?
\n\n\n\n
2) What was the occasion?
\n\n\n\n
3) Did Tilman participate, and if so, how?
\n\n\n\n\n\n\n\n
So where to start? We need to know the location before we can answer any of Tilman’s questions. There are a few clues in the image that might make a good starting point: the language on the screen is is in British English (organisation spelled with an ‘s’) and it suggests a networking event or meeting of some kind. There’s a WiFi sign but unfortunately the network name is too small to read, or else we could have had some fun with Wigle. No quick wins so far.
\n\n\n\n
An important rule in geolocation (and investigation generally) is that the more unique or distinctive a feature is, the more useful it is for finding the origin of a photo. By far the most distinctive feature is the unusual pyramid shaped building that it visible from the window. If we can locate this, we can work out where the photo was taken from.
\n\n\n\nThis is the most distinguishing feature in the photo and will be the key to geolocating it.\n\n\n\n
Cropping the image and conducting a reverse image search seems to be the most obvious way to find the building, but there’s a big problem with that approach in this instance. When we look through a window with human eyes, we subconsciously ignore details like window frames and curtains to focus on what is outside. Unfortunately reverse image search engines aren’t smart enough to do this yet and they don’t distinguish between the foreground noise and the background detail that we are interested in. If you tried to solve this puzzle with this method you probably got a lot of matches that looked like this:
\n\n\n\nYandex search results find a lot of matches…but only for window frames.\n\n\n\n
Reverse image search engines show their limitations here. They focus on the window frame and curtains but don’t focus on the building outside that we’re interested in. So what can we do?
\n\n\n\n
Removing Noise From The Image
\n\n\n\n
What we need is an easy way to remove the foreground clutter from the image so we can conduct a better reverse image search. Cleanup.pictures is very easy to use and makes it easy to remove unwanted features (like window frames) from images. Instead of leaving gaps, the software guesses what should appear in the image based on the rest of the image content and fills in the details. It will be perfect for what we need.
\n\n\n\n
First we upload our cropped image to the website.
\n\n\n\nCropped image ready to upload to Cleanup.pictures.\n\n\n\n
Next we use the eraser tool to highlight the details we want to remove.
\n\n\n\nHighlight the areas for removal.\n\n\n\n
Here’s the final result – not too bad!
\n\n\n\nThe result with all the foreground noise removed.\n\n\n\n
By using the smart eraser Cleanup.pictures removes the unwanted details and fills in the gaps based on the unaltered parts of the image. The result you can see above gives a clutter-free image that will give us a much better chance of success with a reverse image search. Here are the reverse image search results with the foreground clutter removed:
\n\n\n\nSearch results without the foreground noise.\n\n\n\n
Much better! With the window frame removed, we get much better matches for the building in the background. We learn that it is the National Library of Latvia in Riga. By examining the area in Google Maps, we can see that there is only one building that Tilman’s photo could have been taken from: the Wellton Riverside Spa Hotel. The other buildings opposite the National Library do not have glass fronted balconies like the location in the original picture.
\n\n\n\nThe Wellton Riverside Spa Hotel in Riga.\n\n\n\n
By way of confirmation, a quick look at the conference facilities on the hotel website brings up a room that matches the one in Tilman’s picture. The National Library is even visible out of the windows.
\n\n\n\nThe conference room at the Wellton Riverside Spa.\n\n\n\n
So we know the location – how do we find out about the conference?
\n\n\n\n
Finding The Event
\n\n\n\n
Tilman posted this challenge on Wednesday 26th October and the photo was posted on the same day, so it’s reasonable to assume that the event happened in the few days leading up to the 26th. This means a good starting point is to found out what events were held at the Wellton Riverside Spa that week. It’s a simple piece of information but it’s surprisingly tricky to find out. The hotel website doesn’t list upcoming private events and search engines aren’t really too helpful. Eventbrite has a great search feature that allows you to search for events by location, but unfortunately it only works for current and future events, not past ones. It’s much easier to start with an event and find the location, but it seems more difficult the other way around.
\n\n\n\n
So what techniques might help us make the link between an event and a location? Geotagged social media posts are often helpful to tackle this problem. Both Facebook and Instagram have geographic search filters, so they are a good place to start.
\n\n\n\n
Instagram is my preferred resource for this because even though its search features are quite limited, it is still really good for searching photos by location. Simply entering “Wellton Riverside Spa” into the search bar brings up a few locations and hashtags of interest.
\n\n\n\nAll the matching hashtags and location tags for the Wellton Riverside Spa. \n\n\n\n
One of the geotagged images leads to this image that was posted on 25th October. It refers to the hashtag #MILSummit2022 – so we will need to find out exactly what this summit was.
\n\n\n\nThe image is geotagged at the hotel location.\n\n\n\nA second image reveals some of the themes of the event.\n\n\n\n
The same user posted another photo seemingly taken at the same event. Whatever the MIL summit is, the themes of media literacy, disinformation and verification are very common. We know Tilman works in this field, so this is probably a good indicator we are on the right track. We also learn from the second image that the full name of the conference is the “Baltic MIL summit”.
\n\n\n\nThis post links the MIL event at the hotel to the Ziniu Economic Forum,\n\n\n\n
By researching the “Baltic MIL Summit” we quickly find that on Monday 24th October the Žinių Ekonomikos Forumas was hosting the MIL Summit at the hotel. The Facebook page for the event contains hundreds of photos of the participants taking part in the conference. Not only do these photos help to confirm the location matches the one in the original photo, but eventually we even find the exact same photo that Tilman used for the quiz. Bingo!
\n\n\n\nA familiar photo on the ZEF Facebook page.\n\n\n\n
So we now know that the image was posted by the Ziniu Economic Forum on their own Facebook page, and that the event at the hotel was the Baltic MIL media literacy summit held at the Wellton Riverside in Riga. To finish the quiz we just need to find out whether Tilman took part (if indeed he did).
\n\n\n\n
Linking Tilman To The Conference
\n\n\n\n
Clicking through all 230+ of the event photos was useful because not only did it help to identify the source of the original image, but despite all the attendees being very heavily photographed there was not a single photo with Tilman in it. (If you didn’t know what Tilman looked like already, a little Googling will bring up this bio with a picture.)
\n\n\n\nTilman’s picture and bio details.\n\n\n\n
So there are two possibilities: either Tilman was there and he is very, very camera shy, or he was not there in person at all. On the balance of probabilties, it seems more likely that he did not physically attend the event, but we need to try and verify this first. Are there any other sources of information that might help?
\n\n\n\n
Looking through the images, it’s clear that there are strong links between the Deutsche Welle Akademie (where Tilman works) and the MIL event in Riga. We need to try and explore this connection a little bit more.
\n\n\n\n
A good place to start might be with the conference organisers, Ziniu Economic Forum. Most conferences have at least some publicity and a schedule of events. If we can find these, we might be able to learn a little bit more about the links between Deutsche Welle and the MIL Summit.
The first article on the ZEF website refers to the summit and also includes a schedule for the event (PDF download). There are a lot of speakers – including several from DW Akademie – but Tilman is not one of them.
\n\n\n\nDay Two of the summit, but no Tilman…\n\n\n\n
So we know some of Tilman’s colleagues were speaking at the event, and even though we have not found any evidence that he attended in person, the schedule indicates that some of the sessions were streamed over the internet – so perhaps he took part by watching online? Perhaps he did, but it is hard to verify this directly. Can we do any better than just saying Tilman is linked to the conference because his colleagues were speaking there?
\n\n\n\n
We have already made a one-way link from the ZEF to the DW Akademie via the conference photos and speaker schedule, but what if we approach from the other direction and see what DW Akademie have to say in their own event publicity?
By starting at the DW Akademie website we can see that they have two social media news feeds, one for Twitter and one for Facebook. By focusing on the dates of 24-25th October we can see that a live feed of the conference was available (so perhaps this was how Tilman took part?), but the other main publicity feature from DW is a talk by Lena Ritsche on their new game “Go Verify”. Here it is on Facebook:
\n\n\n\nFacebook mention of Go Verify.\n\n\n\n
And here on Twitter…
\n\n\n\nTwitter reference to Go Verify.\n\n\n\n
The Twitter post contains a direct link to the Go Verify game that is the main focus of Deutsche Welle’s publicity around the Baltic MIL Summit. By following the link and reading about Go Verify we find a familiar name at the bottom of the article:
\n\n\n\n\n\n\n\n
And there we have the closest link between Tilman and the conference so far. It seems unlikely that he was present, and we can’t say for sure whether he watched online or not – but we can say that the Go Verify game that he helped to develop was the subject of one of the talks.
\n\n\n\n
If you want to learn more about how to use Cleanup.pictures you can wach Technisette and I demonstrate it on David Bombal’s webcast here.
LS \nHet hotel waar hierover gesproken wordt, wat tegenover de bibliotheek zou moeten staan, staat er niet. Gelukkig is het maar een spelletje, groet..
\n
\n
Loading...
\n
\n
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t \n\t\t\t\t\t\t\t
\t\t\t\t\t\n\n\t\t\t\t\n\t\t\t\t
\n\n\t\t\t\t\t\n\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t
Nixintel
\t\t\t\t\t\t\t\t
\t\t\t\t\t\t\t\t \n\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t
The hotel is directly opposite the library on the other side of the Daugava River?
\n
\n
Loading...
\n
\n
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t \n\t\t\t\t\t\t\t
\t\t\t\t\t\n\n\t\t\t\t
\n\n
\n\t\t\t\t
\n\n\t\t\t\t\t\n\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\t
Sarmad Nadeem
\t\t\t\t\t\t\t\t
\t\t\t\t\t\t\t\t \n\t\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t
I use Imgcleaner.com to remove unwanted objects from the image. Its good.
Great OSINT quiz on cleaning pictures! Using PicPicAi image editing tools can enhance photo clarity, making search results more accurate. A fun, practical guide for better visual investigations!
![\"brown1\"](\"http://bellingcat.com/app/uploads/2014/07/brown1.png\")
![\"brown2\"](\"http://bellingcat.com/app/uploads/2014/07/brown2.png\")
![\"brown3\"](\"http://bellingcat.com/app/uploads/2014/07/brown3.png\")
![\"brown5\"](\"http://bellingcat.com/app/uploads/2014/07/brown5.png\")
![\"brown6\"](\"http://bellingcat.com/app/uploads/2014/07/brown6.png\")
![\"group7\"](\"http://bellingcat.com/app/uploads/2014/07/group7.png\")
![\"group8\"](\"http://bellingcat.com/app/uploads/2014/07/group8.png\")
![\"group9\"](\"http://bellingcat.com/app/uploads/2014/07/group9.png\")
![\"group11\"](\"http://bellingcat.com/app/uploads/2014/07/group11.png\")
![\"group12\"](\"http://bellingcat.com/app/uploads/2014/07/group12.png\")
![\"hello1\"](\"http://bellingcat.com/app/uploads/2014/07/hello1.png\")
![\"hello2\"](\"http://bellingcat.com/app/uploads/2014/07/hello2.png\")
![\"hello3\"](\"http://bellingcat.com/app/uploads/2014/07/hello3.png\")
![\"hello4\"](\"http://bellingcat.com/app/uploads/2014/07/hello4.png\")
![\"hello5\"](\"http://bellingcat.com/app/uploads/2014/07/hello5.png\")
![\"hello6\"](\"http://bellingcat.com/app/uploads/2014/07/hello6.png\")
![\"hello7\"](\"http://bellingcat.com/app/uploads/2014/07/hello7.png\")
![\"hello8\"](\"http://bellingcat.com/app/uploads/2014/07/hello8.png\")
![\"hello9\"](\"http://bellingcat.com/app/uploads/2014/07/hello9.png\")
![\"IMAGE](\"http://bellingcat.com/app/uploads/2014/07/IMAGE-1.jpg\")
![\"IMAGE](\"http://bellingcat.com/app/uploads/2014/07/IMAGE-2.jpg\")
![\"IMAGE](\"http://bellingcat.com/app/uploads/2014/07/IMAGE-3.jpg\")
![\"IMAGE](\"http://bellingcat.com/app/uploads/2014/07/IMAGE-4.jpg\")
![\"IMAGE](\"http://bellingcat.com/app/uploads/2014/07/IMAGE-5.jpg\")
![\"IMAGE](\"http://bellingcat.com/app/uploads/2014/07/IMAGE-6.jpg\")
![\"IMAGE](\"http://bellingcat.com/app/uploads/2014/07/IMAGE-7.jpg\")
![\"IMAGE](\"https://bellingcat.com/app/uploads/2014/07/IMAGE-14.jpg\")
![\"IMAGE](\"https://bellingcat.com/app/uploads/2014/07/IMAGE-22.jpg\")
![\"IMAGE](\"https://bellingcat.com/app/uploads/2014/07/IMAGE-32.jpg\")
![\"IMAGE](\"https://bellingcat.com/app/uploads/2014/07/IMAGE-42.jpg\")
![\"IAMGE](\"https://bellingcat.com/app/uploads/2014/07/IAMGE-5.jpg\")
![\"bukmap1\"](\"https://bellingcat.com/app/uploads/2014/07/bukmap1.jpg\")
![\"bukmap2\"](\"https://bellingcat.com/app/uploads/2014/07/bukmap2.jpg\")
![\"bukmap3\"](\"https://bellingcat.com/app/uploads/2014/07/bukmap3.jpg\")
![\"bukmap4\"](\"https://bellingcat.com/app/uploads/2014/07/bukmap4.jpg\")
![\"bukmap5\"](\"https://bellingcat.com/app/uploads/2014/07/bukmap5.jpg\")
![\"bukmap6\"](\"https://bellingcat.com/app/uploads/2014/07/bukmap6.jpg\")
![\"bukmap7\"](\"https://bellingcat.com/app/uploads/2014/07/bukmap7.jpg\")
![\"buk](\"https://bellingcat.com/app/uploads/2014/07/buk-possible-site.jpg.jpg\")
![\"map1\"](\"https://bellingcat.com/app/uploads/2014/07/map1.jpg\")
![\"7Wseuzow\"](\"https://bellingcat.com/app/uploads/2014/07/7Wseuzow.jpg\")
![\"map3\"](\"https://bellingcat.com/app/uploads/2014/07/map3.jpg\")
![\"map2\"](\"https://bellingcat.com/app/uploads/2014/07/map2.jpg\")
![\"Map4\"](\"https://bellingcat.com/app/uploads/2014/07/Map4.jpg\")
![\"المؤتمر_الصحفي_للمجلس_المحلي_في_زملكا_بريف_دمشق_-_YouTube-11\"](\"https://bellingcat.com/app/uploads/2014/08/المؤتمر_الصحفي_للمجلس_المحلي_في_زملكا_بريف_دمشق_-_YouTube-11.jpg\")
![\"Rocket](\"https://bellingcat.com/app/uploads/2014/08/Rocket-map.jpg\")
![\"Comp](\"https://bellingcat.com/app/uploads/2014/08/Comp-2.jpg\")
![\"Comp](\"https://bellingcat.com/app/uploads/2014/08/Comp-1-1200x426.jpg\")
![\"Comp](\"https://bellingcat.com/app/uploads/2014/08/Comp-3-1200x387.jpg\")
![\"2\"](\"https://bellingcat.com/app/uploads/2014/08/2.jpg\")
![\"1\"](\"https://bellingcat.com/app/uploads/2014/08/1.jpg\")
![\"3\"](\"https://bellingcat.com/app/uploads/2014/08/3-1200x710.jpg\")
![\"5\"](\"https://bellingcat.com/app/uploads/2014/08/5-1200x667.jpg\")
![\"4\"](\"https://bellingcat.com/app/uploads/2014/08/4.jpg\")
![\"6\"](\"https://bellingcat.com/app/uploads/2014/08/6-1200x670.jpg\")
![\"7\"](\"https://bellingcat.com/app/uploads/2014/08/7.jpg\")
![\"12\"](\"https://bellingcat.com/app/uploads/2014/08/12.jpg\")
![\"8\"](\"https://bellingcat.com/app/uploads/2014/08/8-1200x674.jpg\")
![\"OPCW3\"](\"https://bellingcat.com/app/uploads/2014/08/OPCW3-1200x716.jpg\")
![\"OPCW2\"](\"https://bellingcat.com/app/uploads/2014/08/OPCW2.jpg\")
![\"OPCW1\"](\"https://bellingcat.com/app/uploads/2014/08/OPCW1.jpg\")
![\"positions\"](\"https://bellingcat.com/app/uploads/2014/08/positions1.jpg\")
![\"OPCW4\"](\"https://bellingcat.com/app/uploads/2014/08/OPCW4-1200x657.jpg\")
![\"OCPW5\"](\"https://bellingcat.com/app/uploads/2014/08/OCPW5.jpg\")
![\"OPCW6\"](\"https://bellingcat.com/app/uploads/2014/08/OPCW6-1200x621.jpg\")
![\"IMAGE-3\"](\"https://bellingcat.com/app/uploads/2014/08/IMAGE-3.jpg\")
![\"2\"](\"https://bellingcat.com/app/uploads/2014/08/21.jpg\")
![\"1\"](\"https://bellingcat.com/app/uploads/2014/08/11.jpg\")
![\"3\"](\"https://bellingcat.com/app/uploads/2014/08/31.jpg\")
![\"4\"](\"https://bellingcat.com/app/uploads/2014/08/41.jpg\")
![\"5\"](\"https://bellingcat.com/app/uploads/2014/08/51.jpg\")
![\"6\"](\"https://bellingcat.com/app/uploads/2014/08/61.jpg\")
![\"7\"](\"https://bellingcat.com/app/uploads/2014/08/71.jpg\")
![\"8\"](\"https://bellingcat.com/app/uploads/2014/08/81.jpg\")
![\"9\"](\"https://bellingcat.com/app/uploads/2014/08/9.jpg\")
![\"10\"](\"https://bellingcat.com/app/uploads/2014/08/10-1200x1000.jpg\")
![\"11\"](\"https://bellingcat.com/app/uploads/2014/08/111.jpg\")
![\"12\"](\"https://bellingcat.com/app/uploads/2014/08/121.jpg\")
![\"13\"](\"https://bellingcat.com/app/uploads/2014/08/13.jpg\")
![\"14\"](\"https://bellingcat.com/app/uploads/2014/08/14.jpg\")
![\"15\"](\"https://bellingcat.com/app/uploads/2014/08/15.jpg\")
![\"16\"](\"https://bellingcat.com/app/uploads/2014/08/16.jpg\")
![\"17\"](\"https://bellingcat.com/app/uploads/2014/08/17.jpg\")
![\"19\"](\"https://bellingcat.com/app/uploads/2014/08/19.jpg\")
![\"18\"](\"https://bellingcat.com/app/uploads/2014/08/18.jpg\")
![\"21\"](\"https://bellingcat.com/app/uploads/2014/08/211.jpg\")
![\"20\"](\"https://bellingcat.com/app/uploads/2014/08/20.jpg\")
![\"23\"](\"https://bellingcat.com/app/uploads/2014/08/23.jpg\")
![\"22\"](\"https://bellingcat.com/app/uploads/2014/08/22.jpg\")
![\"25\"](\"https://bellingcat.com/app/uploads/2014/08/25.jpg\")
![\"24\"](\"https://bellingcat.com/app/uploads/2014/08/24.jpg\")
![\"29\"](\"https://bellingcat.com/app/uploads/2014/08/29.jpg\")
![\"28\"](\"https://bellingcat.com/app/uploads/2014/08/28.jpg\")
![\"27\"](\"https://bellingcat.com/app/uploads/2014/08/27.jpg\")
![\"26\"](\"https://bellingcat.com/app/uploads/2014/08/26.jpg\")
![\"Angle](\"https://bellingcat.com/app/uploads/2014/08/Angle-5.jpg\")
![\"Sat](\"https://bellingcat.com/app/uploads/2014/08/Sat-map-Raqqa.jpg\")
![\"Raised](\"https://bellingcat.com/app/uploads/2014/08/Raised-1.jpg\")
![\"Raised](\"https://bellingcat.com/app/uploads/2014/08/Raised-2.jpg\")
![\"Raised](\"https://bellingcat.com/app/uploads/2014/08/Raised-3.jpg\")
![\"Break](\"https://bellingcat.com/app/uploads/2014/08/Break-1.jpg\")
![\"Break](\"https://bellingcat.com/app/uploads/2014/08/Break-2.jpg\")
![\"Location](\"https://bellingcat.com/app/uploads/2014/08/Location-1.jpg\")
![\"Jut](\"https://bellingcat.com/app/uploads/2014/08/Jut-1.jpg\")
![\"Jut](\"https://bellingcat.com/app/uploads/2014/08/Jut-2.jpg\")
![\"Outcrop\"](\"https://bellingcat.com/app/uploads/2014/08/Outcrop.jpg\")
![\"Location](\"https://bellingcat.com/app/uploads/2014/08/Location-2.jpg\")
![\"Background](\"https://bellingcat.com/app/uploads/2014/08/Background-1.jpg\")
![\"Background](\"https://bellingcat.com/app/uploads/2014/08/Background-2.jpg\")
![\"Background](\"https://bellingcat.com/app/uploads/2014/08/Background-3.jpg\")
![\"Background](\"https://bellingcat.com/app/uploads/2014/08/Background-4.jpg\")
![\"Background](\"https://bellingcat.com/app/uploads/2014/08/Background-5.jpg\")
![\"1\"](\"https://bellingcat.com/app/uploads/2014/09/16.jpg\")
![\"2\"](\"https://bellingcat.com/app/uploads/2014/09/2.jpg\")
![\"3\"](\"https://bellingcat.com/app/uploads/2014/09/31.jpg\")
![\"4\"](\"https://bellingcat.com/app/uploads/2014/09/41.jpg\")
![\"5\"](\"https://bellingcat.com/app/uploads/2014/09/52.jpg\")
![\"6\"](\"https://bellingcat.com/app/uploads/2014/09/6.jpg\")
![\"7\"](\"https://bellingcat.com/app/uploads/2014/09/71.jpg\")
![\"8\"](\"https://bellingcat.com/app/uploads/2014/09/81.jpg\")
![\"9\"](\"https://bellingcat.com/app/uploads/2014/09/91.jpg\")
![\"10\"](\"https://bellingcat.com/app/uploads/2014/09/101.jpg\")
![\"11\"](\"https://bellingcat.com/app/uploads/2014/09/111.jpg\")
![\"12\"](\"https://bellingcat.com/app/uploads/2014/09/121.jpg\")
![\"13\"](\"https://bellingcat.com/app/uploads/2014/09/131.jpg\")
![\"14\"](\"https://bellingcat.com/app/uploads/2014/09/141.jpg\")
![\"15\"](\"https://bellingcat.com/app/uploads/2014/09/151.jpg\")
![\"16\"](\"https://bellingcat.com/app/uploads/2014/09/161.jpg\")
![\"17\"](\"https://bellingcat.com/app/uploads/2014/09/17.jpg\")
![\"july-01-fedoseevka\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-01-fedoseevka.jpg\")
![\"july-02-fedoseevka\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-02-fedoseevka.jpg\")
![\"july-03-staryoskol\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-03-staryoskol.jpg\")
![\"july-04-staryoskol\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-04-staryoskol.jpg\")
![\"truck-comparison-left-right\"](\"https://www.bellingcat.com/app/uploads/2014/11/truck-comparison-left-right.png\")
![\"july-05-staryoskol\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-05-staryoskol.jpg\")
![\"july-06-staryoskol\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-06-staryoskol.jpg\")
![\"july-07-staryoskol\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-07-staryoskol.jpg\")
![\"july-08-neznamovo\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-08-neznamovo.jpg\")
![\"july-09-neznamovo\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-09-neznamovo.jpg\")
![\"july-11-alexeyevka\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-11-alexeyevka.jpg\")
![\"july-12-olkhovatka\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-12-olkhovatka.jpg\")
![\"map-july1\"](\"https://www.bellingcat.com/app/uploads/2014/11/map-july1.jpg\")
![\"july-10-kamensk-shakhtinsky\"](\"https://www.bellingcat.com/app/uploads/2014/11/july-10-kamensk-shakhtinsky.jpg\")
![\"map-july2\"](\"https://www.bellingcat.com/app/uploads/2014/11/map-july2.jpg\")
![\"01-troitsky\"](\"https://www.bellingcat.com/app/uploads/2014/11/01-troitsky.jpg\")
![\"02-staryoskol\"](\"https://www.bellingcat.com/app/uploads/2014/11/02-staryoskol.jpg\")
![\"03-tel-comparison\"](\"https://www.bellingcat.com/app/uploads/2014/11/03-tel-comparison.jpg\")
![\"04-popovy-dvory-sign\"](\"https://www.bellingcat.com/app/uploads/2014/11/04-popovy-dvory-sign.jpg\")
![\"05-stary-oskol-bldg\"](\"https://www.bellingcat.com/app/uploads/2014/11/05-stary-oskol-bldg.jpg\")
![\"06-stary-oskol-deleted\"](\"https://www.bellingcat.com/app/uploads/2014/11/06-stary-oskol-deleted.jpg\")
![\"07-stary-bridge\"](\"https://www.bellingcat.com/app/uploads/2014/11/07-stary-bridge.jpg\")
![\"08-stary-rail\"](\"https://www.bellingcat.com/app/uploads/2014/11/08-stary-rail.jpg\")
![\"09-alexeyevka\"](\"https://www.bellingcat.com/app/uploads/2014/11/09-alexeyevka.jpg\")
![\"10-alexeyevka-rural\"](\"https://www.bellingcat.com/app/uploads/2014/11/10-alexeyevka-rural.jpg\")
![\"map-june-mb\"](\"https://www.bellingcat.com/app/uploads/2014/11/map-june-mb.jpg\")
![\"1\"](\"https://www.bellingcat.com/app/uploads/2014/11/1.jpg\")
![\"2\"](\"https://www.bellingcat.com/app/uploads/2014/11/2.jpg\")
![\"3\"](\"https://www.bellingcat.com/app/uploads/2014/11/3.jpg\")
![\"4\"](\"https://www.bellingcat.com/app/uploads/2014/11/4.jpg\")
![\"5\"](\"https://www.bellingcat.com/app/uploads/2014/11/5.jpg\")
![\"6\"](\"https://www.bellingcat.com/app/uploads/2014/11/6.jpg\")
![\"7\"](\"https://www.bellingcat.com/app/uploads/2014/11/7.jpg\")
![\"8\"](\"https://www.bellingcat.com/app/uploads/2014/11/8.jpg\")
![\"9\"](\"https://www.bellingcat.com/app/uploads/2014/11/9.jpg\")
![\"10\"](\"https://www.bellingcat.com/app/uploads/2014/11/10.jpg\")
![\"11\"](\"https://www.bellingcat.com/app/uploads/2014/11/11.jpg\")
![\"12\"](\"https://www.bellingcat.com/app/uploads/2014/11/12.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-1.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-2.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-3.png\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-4.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-6.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-5.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-7.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-8.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-9.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-10.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-11.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-12.png\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-13.png\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-14.jpg\")
![\"Tunis](\"https://www.bellingcat.com/app/uploads/2014/12/Tunis-15.jpg\")
![\"screen_shot_2014-12-19_at_1.48.54_pm_1024\"](\"https://www.bellingcat.com/app/uploads/2014/12/screen_shot_2014-12-19_at_1.48.54_pm_1024.png\")
![\"23\"](\"https://www.bellingcat.com/app/uploads/2015/03/23.jpg\")
![\"G8NQGeq7u4Y\"](\"https://www.bellingcat.com/app/uploads/2015/03/G8NQGeq7u4Y.jpg\")
![\"dash\"](\"https://www.bellingcat.com/app/uploads/2015/03/dash.jpg\")
![\"SV](\"https://www.bellingcat.com/app/uploads/2015/03/SV-dash.jpg\")
![\"SV](\"https://www.bellingcat.com/app/uploads/2015/03/SV-comp-verge.jpg\")
![\"SV](\"https://www.bellingcat.com/app/uploads/2015/03/SV-poll-comp.jpg\")
![\"Crack\"](\"https://www.bellingcat.com/app/uploads/2015/03/Crack.jpg\")
![\"1\"](\"https://www.bellingcat.com/app/uploads/2015/04/1.jpg\")
![\"2\"](\"https://www.bellingcat.com/app/uploads/2015/04/2.jpg\")
![\"3\"](\"https://www.bellingcat.com/app/uploads/2015/04/3.jpg\")
![\"4\"](\"https://www.bellingcat.com/app/uploads/2015/04/4.jpg\")
![\"5\"](\"https://www.bellingcat.com/app/uploads/2015/04/5.jpg\")
![\"6\"](\"https://www.bellingcat.com/app/uploads/2015/04/6.jpg\")
![\"7\"](\"https://www.bellingcat.com/app/uploads/2015/04/7.jpg\")
![\"8\"](\"https://www.bellingcat.com/app/uploads/2015/04/8.jpg\")
![\"9\"](\"https://www.bellingcat.com/app/uploads/2015/04/9.jpg\")
![\"10\"](\"https://www.bellingcat.com/app/uploads/2015/04/10.jpg\")
![\"11\"](\"https://www.bellingcat.com/app/uploads/2015/04/11.jpg\")
![\"12\"](\"https://www.bellingcat.com/app/uploads/2015/04/12.jpg\")
![\"13\"](\"https://www.bellingcat.com/app/uploads/2015/04/13.jpg\")
![\"14\"](\"https://www.bellingcat.com/app/uploads/2015/04/14.jpg\")
![\"15\"](\"https://www.bellingcat.com/app/uploads/2015/04/15.jpg\")
![\"16\"](\"https://www.bellingcat.com/app/uploads/2015/04/16.jpg\")
![\"header\"](\"https://www.bellingcat.com/app/uploads/2015/05/header.jpg\")
![\"CENTCOM\"](\"https://www.bellingcat.com/app/uploads/2020/05/CENTCOM.jpg\")
![\"mod\"](\"https://www.bellingcat.com/app/uploads/2020/05/mod.jpg\")
![\"hud1\"](\"https://www.bellingcat.com/app/uploads/2020/05/hud1.jpg\")
![\"date](\"https://www.bellingcat.com/app/uploads/2020/05/date-and-time.jpg\")
![\"signal\"](\"https://www.bellingcat.com/app/uploads/2020/05/signal.jpg\")
![\"zoomout\"](\"https://www.bellingcat.com/app/uploads/2020/05/zoomout.jpg\")
![\"GE\"](\"https://www.bellingcat.com/app/uploads/2020/05/GE.jpg\")
![\"Bing\"](\"https://www.bellingcat.com/app/uploads/2020/05/Bing.jpg\")
![\"bingdate\"](\"https://www.bellingcat.com/app/uploads/2020/05/bingdate.jpg\")
![\"relation](\"https://www.bellingcat.com/app/uploads/2020/05/relation-to-ramadi.jpg\")
![\"happy\"](\"https://www.bellingcat.com/app/uploads/2020/05/happy.jpg\")
![\"taqaddum\"](\"https://www.bellingcat.com/app/uploads/2020/05/taqaddum.jpg\")
![\"mi-17\"](\"https://www.bellingcat.com/app/uploads/2020/05/mi-17.jpg\")
![\"mourning\"](\"https://www.bellingcat.com/app/uploads/2015/05/mourning.jpg\")
![\"luh1\"](\"https://www.bellingcat.com/app/uploads/2015/05/luh1.jpg\")
![\"GHMRwTU71H8\"](\"https://www.bellingcat.com/app/uploads/2015/05/GHMRwTU71H8.jpg\")
![\"Screenshot](\"https://www.bellingcat.com/app/uploads/2015/05/base.jpg\")
![\"RussianArmy1\"](\"https://www.bellingcat.com/app/uploads/2015/05/RussianArmy11.jpg\")
![\"chimney\"](\"https://www.bellingcat.com/app/uploads/2015/05/chimney.jpg\")
![\"geoloc_map\"](\"https://www.bellingcat.com/app/uploads/2015/05/geoloc_map.jpg\")
![\"cfsvpbuw0aaayyv__1_\"](\"https://www.bellingcat.com/app/uploads/2015/05/cfsvpbuw0aaayyv__1_.jpg\")
![\"pav1\"](\"https://www.bellingcat.com/app/uploads/2015/05/pav1.jpg\")
![\"pav2\"](\"https://www.bellingcat.com/app/uploads/2015/05/pav2.jpg\")
![\"pav3\"](\"https://www.bellingcat.com/app/uploads/2015/05/pav31.jpg\")
![\"uFA3scj\"](\"https://www.bellingcat.com/app/uploads/2015/05/uFA3scj.jpg\")
![\"3fgnLfz\"](\"https://www.bellingcat.com/app/uploads/2015/05/3fgnLfz.jpg\")
![\"1\"](\"https://www.bellingcat.com/app/uploads/2015/05/11.jpg\")
![\"2\"](\"https://www.bellingcat.com/app/uploads/2015/05/2.jpg\")
![\"3\"](\"https://www.bellingcat.com/app/uploads/2015/05/3.jpg\")
![\"4\"](\"https://www.bellingcat.com/app/uploads/2015/05/4.jpg\")
![\"5\"](\"https://www.bellingcat.com/app/uploads/2015/05/5.png\")
![\"6\"](\"https://www.bellingcat.com/app/uploads/2015/05/6.jpg\")
![\"m1\"](\"https://www.bellingcat.com/app/uploads/2015/05/m1.jpg\")
![\"Plates1\"](\"https://www.bellingcat.com/app/uploads/2015/05/Plates1.jpg\")
![\"m2\"](\"https://www.bellingcat.com/app/uploads/2015/05/m2.jpg\")
![\"m3\"](\"https://www.bellingcat.com/app/uploads/2015/05/m3.jpg\")
![\"m4\"](\"https://www.bellingcat.com/app/uploads/2015/05/m4.jpg\")
![\"m5\"](\"https://www.bellingcat.com/app/uploads/2015/05/m5.jpg\")
![\"m6\"](\"https://www.bellingcat.com/app/uploads/2015/05/m6.jpg\")
![\"strips\"](\"https://www.bellingcat.com/app/uploads/2015/06/strips.jpg\")
![\"globe\"](\"https://www.bellingcat.com/app/uploads/2015/06/globe.jpg\")
![\"coords\"](\"https://www.bellingcat.com/app/uploads/2015/06/coords.jpg\")
![\"globe_search\"](\"https://www.bellingcat.com/app/uploads/2015/06/globe_search.jpg\")
![\"globe_acquisition\"](\"https://www.bellingcat.com/app/uploads/2015/06/globe_acquisition.jpg\")
![\"globe_results\"](\"https://www.bellingcat.com/app/uploads/2015/06/globe_results.jpg\")
![\"globe_max\"](\"https://www.bellingcat.com/app/uploads/2015/06/globe_max.jpg\")
![\"globe_save\"](\"https://www.bellingcat.com/app/uploads/2015/06/globe_save.jpg\")
![\"ge_7.24.14\"](\"https://www.bellingcat.com/app/uploads/2015/06/ge_7.24.14.jpg\")
![\"globe_7242014\"](\"https://www.bellingcat.com/app/uploads/2015/06/globe_7242014.jpg\")
![\"ge_july2\"](\"https://www.bellingcat.com/app/uploads/2015/06/ge_july2.jpg\")
![\"globe_july2\"](\"https://www.bellingcat.com/app/uploads/2015/06/globe_july2.jpg\")
![\"ge_may30\"](\"https://www.bellingcat.com/app/uploads/2015/06/ge_may30.jpg\")
![\"globe_may30\"](\"https://www.bellingcat.com/app/uploads/2015/06/globe_may30.jpg\")
![\"Russia](\"https://www.bellingcat.com/app/uploads/2015/05/Russia-MoD-2.jpg\")
![\"An](\"https://www.bellingcat.com/app/uploads/2015/06/MoD-1.jpg\")
![\"1\"](\"https://www.bellingcat.com/app/uploads/2015/07/11.jpg\")
![\"2\"](\"https://www.bellingcat.com/app/uploads/2015/07/21.jpg\")
![\"3\"](\"https://www.bellingcat.com/app/uploads/2015/07/31.jpg\")
![\"4\"](\"https://www.bellingcat.com/app/uploads/2015/07/4.jpg\")
![\"5\"](\"https://www.bellingcat.com/app/uploads/2015/07/51.jpg\")
![\"BswldgYIUAAV7A_\"](\"https://www.bellingcat.com/app/uploads/2015/07/BswldgYIUAAV7A_.jpg\")
![\"house\"](\"https://www.bellingcat.com/app/uploads/2015/08/house.jpg\")
![\"One](\"https://www.bellingcat.com/app/uploads/2015/08/DSCN4325.jpg\")
![\"8-5-2015](\"https://www.bellingcat.com/app/uploads/2015/08/8-5-2015-6-54-14-PM.jpg\")
![\"8-6-2015](\"https://www.bellingcat.com/app/uploads/2015/08/8-6-2015-5-07-14-PM.jpg\")
![\"8-5-2015](\"https://www.bellingcat.com/app/uploads/2015/08/8-5-2015-6-52-46-PM.jpg\")
![\"1\"](\"https://www.bellingcat.com/app/uploads/2015/09/1.jpg\")
![\"2\"](\"https://www.bellingcat.com/app/uploads/2015/09/2.jpg\")
![\"3\"](\"https://www.bellingcat.com/app/uploads/2015/09/3.jpg\")
![\"4\"](\"https://www.bellingcat.com/app/uploads/2015/09/4.jpg\")
![\"7\"](\"https://www.bellingcat.com/app/uploads/2015/09/7.jpg\")
![\"8\"](\"https://www.bellingcat.com/app/uploads/2015/09/8.jpg\")
![\"2015-11-10](\"https://www.bellingcat.com/app/uploads/2015/11/2015-11-10-16_27_35.gif\")
![\"video](\"https://www.bellingcat.com/app/uploads/2015/11/video-1.jpg\")
![\"Video](\"https://www.bellingcat.com/app/uploads/2015/11/Video-21.jpg\")
![\"ge2\"](\"https://www.bellingcat.com/app/uploads/2015/11/ge2.jpg\")
![\"Video](\"https://www.bellingcat.com/app/uploads/2015/11/Video-31.jpg\")
![\"ge3\"](\"https://www.bellingcat.com/app/uploads/2015/11/ge3.jpg\")
![\"1\"](\"https://www.bellingcat.com/app/uploads/2016/02/1-2.jpg\")
![\"2\"](\"https://www.bellingcat.com/app/uploads/2016/02/2-2.jpg\")
![\"3\"](\"https://www.bellingcat.com/app/uploads/2016/02/3-2.jpg\")
![\"4\"](\"https://www.bellingcat.com/app/uploads/2016/02/4-2.jpg\")
![\"5\"](\"https://www.bellingcat.com/app/uploads/2016/02/5-2.jpg\")
![\"6\"](\"https://www.bellingcat.com/app/uploads/2016/02/6-1.jpg\")
![\"7\"](\"https://www.bellingcat.com/app/uploads/2016/02/7-1.jpg\")
![\"8\"](\"https://www.bellingcat.com/app/uploads/2016/02/8-1.jpg\")
![\"9.1\"](\"https://www.bellingcat.com/app/uploads/2016/02/9.1.jpg\")
![\"strike-1\"](\"https://www.bellingcat.com/app/uploads/2016/10/Strike-1-1200x472.jpg\")
![\"strike-2\"](\"https://www.bellingcat.com/app/uploads/2016/10/strike-2-1.jpg\")
![\"strike-3\"](\"https://www.bellingcat.com/app/uploads/2016/10/Strike-3-1200x387.jpg\")
![\"strike-5\"](\"https://www.bellingcat.com/app/uploads/2016/10/strike-5-1200x487.jpg\")
![\"strike-6\"](\"https://www.bellingcat.com/app/uploads/2016/10/strike-6-1200x498.png\")
![\"overview\"](\"https://www.bellingcat.com/app/uploads/2016/10/overview.png\")
![\"vandeput\"](\"https://www.bellingcat.com/app/uploads/2016/10/vandeput-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/03/0mvEb4QXPjw.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/03/lRNxJUF.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/comp_mykhailo.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/mykhailo_comp2.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/2014buksmykhiailo.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/comb-1-1200x446.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/features_mykhailo.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/YT-interface-1-1200x884.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/number-of-airstrikes-per-1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/estimate-1-1200x584.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/05/locations-1200x554.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/geolocated-airstrkes-per-country.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/closest-location-governorate.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/checkcoalition-figure-3-1-1200x426.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/checkcoalition-figure-4-1-1200x141.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/checkcoalition-figure-5-1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/checkcoalition-figure-6-1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/example-1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/manbij-exmaple.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/fighting-position-example-2.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/vehicle-example.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/04/Untitled-1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/territorial-contorl-1200x826.jpg\" "\\"LiveUAmap\\"")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Werfalli-resignation-1200x675.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Incident-One-Still-1-1200x675.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Incident-One-3-1200x576.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Incident-One-2.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Incident-One-4-1200x293.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/DI826MwWsAErDJ7.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Incident1_15-03-2017_matching.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Incident-Four-1200x640.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/DI448YPW4AAVh7o.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Panorama-Video6.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/electricity_poles_visible.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Incident-Six-1200x675.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Incident-Seven-1200x341.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/syrian-archive-1200x447.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-2.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-3.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-4.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-5.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-6.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-7.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-8.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-9.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-10.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-11.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-12.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-13.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-14.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-15.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/09/Image-16.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/10/9OF67tBB15o.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/10/donetsketoukraina-1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/10/gate_geolocation.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/10/notches.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/10/buk_comparison.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2017/10/map-kurazh.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS1.jpeg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS2-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS3-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS4.png\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/vs5.png\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS6.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS7.png\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS8.png\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS9.png\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS10.png\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS11.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS12.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS13.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS14.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS15.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS16.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS17.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/09/VS18.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/Europol_tweet_and_image-1200x670.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/Tweet_and_image_Olli_Enne.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_overview_lines.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_photo_location_2014.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_photo_view_Google-1200x718.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/buildings_right_far_away_and_huge_building_missing.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/Google_view_2002_2008-1200x523.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_first_location-1200x834.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_restaurant_shops-1200x449.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_map_comparison-1200x339.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/Shenzhen_roll_area.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/satellite-receiver_and-dish-00002-1200x675.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/3.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/4.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/5.gif\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/6.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/7.gif\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/8.gif\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/9.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/10.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/11.gif\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/12.gif\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/13.gif\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/14.gif\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/01/pax-logo-web-fc-900.jpg\"){kind=logo}
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/1-7-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/2-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/3-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/4-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/5-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/6-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/8-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/9-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/10-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/11-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/12-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/13-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/14-CAMEROON.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/15-CAMEROON.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/16-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/17-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/18-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/19-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/20-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/21-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/22-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/23-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2018/11/24-CAMEROON.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/babis-4.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/image-1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/image-2.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/image-3.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/image-5.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/andrej-babis-3762053_1280.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/image-6.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/image-7.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtIHQf9XcAAAJlJ.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtJn9jXU8AA7vX2.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtJpTBFVAAE-3_u.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtIHQfSWsAAYBYU.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtKjuGTUwAAc-Ky.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtM79_3X4AA9njb.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtM78qeXgAIG9_s.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtNFW-pW0AAeniP.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtNN0FTWoAIT6Pj.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtNT8k0WwAAFVbe-150x150.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtNT7yYWoAIFui3-150x150.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtNTa7QXQAA4ou7-150x150.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtNT9itXcAA2TK0-150x150.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtNb2tfXQAA5fBc-1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtNb4AhWwAAIswR.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Screenshot-2019-02-06-at-21.48.58.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Screenshot-2019-02-06-at-21.50.43.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtIHQfEWwAIfwH3.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/3_0015.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtLM2iAXQAAQIsv.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Screenshot-2019-02-06-at-21.57.01.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/716819-top_foto1-6yqv9.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/DtKfrTvWkAUflhW.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Dtw0WD_XQAEUPUr.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Dtw4NkaWoAAhy7v.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Dtw2wqpW4AIa-n-.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Dtw2xvbXQAMc2zc.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/image-10.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Dtw2ycyWoAAvbJt.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/YANDEX.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/BLACK-PANTHER-PHOTOS.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/air-hockey.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/reference-photo.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Screenshot-2019-02-06-at-22.22.53.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Dtw7isMXcAEWbF6.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Dtw-LrYXcAEN8__.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Dtw-huNXgAAu9-k.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/02/Dtw-kXkW0AAZ_7J.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/D0Zi8LnXgAAtE2o.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/D0Zi8LjX4AAX47X.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/Chrash-Site-2-Panorama.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/D0ZzusUWkAA9uvE.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/D0bojyjX0AE6c19.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/Place-Horran-Kotla.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/Kotla-Mapcarta-1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/rocks.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/chrash-site-place-2.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/ge-terrain.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/plantangles.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/3.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/4.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/12.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/Screen-Shot-2019-02-22-at-3.32.22-pm.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/5.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/13.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/6.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/7.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/8.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/9.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/10.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/11.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/Screen-Shot-2019-02-22-at-3.59.40-pm.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/Screen-Shot-2019-02-22-at-4.03.39-pm.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/15.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/16.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/17.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/18.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/19.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/20.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/2019-03-19_13-01-50.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/Picture1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/Picture3.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/22-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/2019-03-19_19-38-31.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/Picture4.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/5-3.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/locati.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/as.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/at.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/al.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/ee.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/or.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/ob.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/ol.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/panorama2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/qq.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/qw.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/qa.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/qeeq.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/ee33.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/1313.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/03/christmas.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/summer-house-collection-1200x525.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/summer-house-album-1200x449.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/Lake-1200x602.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/video-location-1200x334.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/enhanced-1200x263.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/comparison.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/skylines-1200x424.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/thunder_stone_2004_skyline_and_Europol_overlay-1200x768.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/Building_and_text.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/Zhongyu_compare-1200x400.jpeg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/3D-model-compare-1200x436.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/Jiangmen_building_compare.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/Times_hotel-1200x385.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/meeting_rooms-1200x424.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/mallchina_coke-1200x437.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/07/2018-6-c17-compare-1200x451.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F2.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F3.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F4.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F5.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F6.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F7.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F8.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F9.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F10.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F11.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F12.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F13.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F14.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F15.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F16.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F17.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F18-847x1200.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/09/F19.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/1-scaled.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/3-scaled.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/4.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/5.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/6.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/7.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/8-scaled.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/9.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/10-scaled.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/11-scaled.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/12-scaled.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/13.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/14.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/15.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/16.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/17.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/18.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/19.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/Table-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/20.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/21.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/22.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/23.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/24.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/25.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/26.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/27.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/28.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/29-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/30-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/31-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/32-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/33-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/34-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/35-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/36-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/37-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/38-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/39-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/40-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/41-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/42-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/43.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/44.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/45-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/46-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/Table-2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2019/12/47-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/2020-01-09_13-20-59.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/image-38.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/image-11.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/ENxkCymUUAAPtXB.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/EN2FJHnW4AAYrAZ.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/Men.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/Diputados-Lago.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/1-Bank-Name-Image-Search.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/2-Bank-Sofia-Results.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/Bank01.jpeg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/4.-ThreeOptions-scaled.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/5-Google-Street-View-Bank.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/6-Geolocation-1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/7.-Mapillary-Bank-scaled.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/8-Roads.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/9-Lake-Comparison-scaled.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/10-Hombres-1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/MountainView1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/01/13-Men-Geolocation-scaled.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/12/suncalc1-1200x729.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/11/image7-1-1200x865.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/11/image10.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/11/image6-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/11/image8.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2020/11/image1.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/1.-John-Doe-29-Warning.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/Screenshot-2021-04-30-at-10.19.34.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/3.-FBI-BB001-and-BB002.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/4.-google-maps_2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/5.-Verification-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/6.-Banana-Leaf.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/7.-Banana-map.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/8.-The-turkish-flag_2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/9.-Metal-trash-can.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/10.-The-sun-lounger-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/11.-3D-model_2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/12.-3D-model.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/13.-Side-Turkey.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/14.-Pool-reflection.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/15.-Verification-2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/16.-Verification-3_2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/17.-Verification-4.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/Screenshot-2021-04-30-at-10.53.29.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/04/Screenshot-2021-04-30-at-10.55.18.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image14-300x160.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image19-300x200.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image13-139x300.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image4-138x300.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/static.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image7-1-300x188.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image8-300x156.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image20-224x300.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image3-300x206.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image11-225x300.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image2-300x263.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image12-225x300.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image15-300x262.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image16-225x300.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image6-300x128.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image10-300x170.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image1-300x223.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image18-300x131.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image17-224x300.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/image9-300x199.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/Temple-mount-1st.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/Temple-Mount-2nd.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/Temple-Mount-Suncalc.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/Zabid-First-picture.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/Zabid-chronolocation.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/Zabid-Suncalc.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/05/pasted-image-0.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image6-4.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image7-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image9-2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image11-2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image8-2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image3-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image1-3.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image13-2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image15-2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image10-2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image2-2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image4-2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image14-3.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2022/10/image5-2.png\")
![\"Illustration](\"https://www.bellingcat.com/app/uploads/2023/02/LUCY_BELLINGCAT-final-1200x977.jpg\")
![\"Image](\"https://www.bellingcat.com/app/uploads/2023/02/output_0006.png\")
![\"Image](\"https://www.bellingcat.com/app/uploads/2023/02/dos-Santos_SaltBae-150x150.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/02/SaltBae-1-e1675419192317-150x150.png\")
![\"Isabel](\"https://www.bellingcat.com/app/uploads/2023/02/Trove-Dos-Santos.png\")
![\"An](\"https://www.bellingcat.com/app/uploads/2023/02/Sardinia-dos-Santos-edited.jpeg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/02/IMG_5257-738x1200.jpg\")
![\"Composite](\"https://www.bellingcat.com/app/uploads/2023/02/TicketQatar.png\")
![\"Image](\"https://www.bellingcat.com/app/uploads/2023/02/Gym1.png\")
![\"Image](\"https://www.bellingcat.com/app/uploads/2023/02/vis-feaxtures-257x300.jpeg\")
![\"Composite](\"https://www.bellingcat.com/app/uploads/2023/02/Gym3.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/02/DW-Interview.png\")
![\"Two](\"https://www.bellingcat.com/app/uploads/2023/02/Pink-Shirt-Selfie.jpeg\")
![\"Composite](\"https://www.bellingcat.com/app/uploads/2023/02/Composite-selfie.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/02/Oldcurtainpic-1.jpeg\")
![\"Old](\"https://www.bellingcat.com/app/uploads/2023/02/furtinure-dw-1200x741.jpeg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-03-at-11.44.38-1200x673.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2021/09/2021-06-23T195324Z_582482922_UP1EH6N1J8YUH_RTRMADP_3_SOCCER-EURO-GER-HUN-REPORT-1200x795.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-03-at-11.49.34-1200x791.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-10.48.45-1200x893.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-10.51.49.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-10.53.53.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-10.56.59.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-11.12.47.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-11.25.55-1200x539.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-11.27.47-1200x442.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-11.27.56-1200x579.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-11.50.37.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.09.10.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.27.21-1200x545.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.28.49.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.30.29-1200x168.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.31.35-1200x551.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.33.00-1200x696.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.35.10-1200x305.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Image-from-insta-709x1200.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.41.18-1200x775.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.41.49-1200x278.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.43.02-1200x552.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.45.14-1200x311.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.56.07-1200x590.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.57.48-1200x669.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-12.58.39-1200x534.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.06.55-1200x601.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.07.50-1200x596.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.11.20-1200x449.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.17.23-1200x669.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.18.18-1200x598.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.19.53-1200x458.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/05/Screenshot-2023-05-08-at-13.21.25-1200x202.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image19-1200x606.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image7-1-1200x677.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image5-3-1200x829.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image17-1200x743.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image20-1200x825.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image22-1200x1054.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image8-1-1200x841.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image4-3.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image6.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image10.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image24-1200x663.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image12-1-300x109.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image9-1-300x65.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image23-300x227.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image21-300x105.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image14.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image11-1200x800.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image18.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image2-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image3-2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image1-4.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image25.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image27-1200x642.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image26-1200x744.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image28-1200x865.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image15-3-1200x861.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image16-1-1200x411.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/04/image13-1200x412.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/06/image3.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/06/image12-1-1200x726.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/06/image-11.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/06/image-9-1200x960.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/06/image2-1-1200x737.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/06/image8-1-1200x1062.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/06/image-8-1200x700.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/06/image-7.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/06/image-5-1200x667.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/06/image-10-1200x674.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/06/image9-1-1200x669.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.40.18.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.43.33-1200x650.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.44.59-1200x261.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.45.23-1200x670.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.45.38.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.45.50-1200x386.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.46.09-1200x525.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.46.26-1200x586.png\")
![](\"https://lh4.googleusercontent.com/fj_STl0caRbRrPIVNi0Wd8PfZdW5z43y7PTD-ZX_G_ImexlkKz2QCo4Ru7TFXSACxqLY23UhVX736yN1vPXSeVGZjQ7kjNY1jGJRpS2_J6mV4eKyD003jSxUsMArRw-hMUhB_j_2O5d9juzquI55tbI\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.47.29.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-13-at-09.06.35-1200x569.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-13-at-09.06.44-1200x415.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.47.52-1200x264.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.48.11-1200x590.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.48.23.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.48.49-1200x593.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.49.00-1200x593.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.49.14-1200x546.png\")
![](\"https://lh5.googleusercontent.com/Qk1UzoaUXXdnpDdvCxsv75cb4VvrTupyV2XqunUkVBmpkDpSk3VVdgQDPBgWq8ExGcQ3joeaesJwtP4JOxtNqwlJl2hgSRAnwqy2Kxk_UMwY6fOEWWTcPEFUoCM4Hx4Eql-NxfAoAPcpaHDLDuqcNDQ\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.49.26-1200x643.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-13-at-10.35.06-1200x696.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-13-at-11.52.27.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/image2-1200x541.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/image3-1200x540.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/07/image4-1200x541.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/09/Screenshot-2023-09-06-at-19.06.46-965x1200.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/09/image3-1200x623.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/09/image5-1200x568.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/09/image4.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/09/image1-1200x547.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/09/image2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-8.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-9-1200x749.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-10-1200x673.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-11-1200x748.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-12-1200x673.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/Screenshot-2023-10-19-at-17.00.10-1188x1200.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-13-1200x785.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-5-1200x675.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-6-1200x675.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed.gif\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-7-1200x675.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-8-1200x675.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-9-1200x675.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-10-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-11.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-14-1200x675.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-12-1200x669.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-13-1200x687.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-14-1200x687.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-15.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-16-1200x675.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/10/unnamed-17.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/unnamed-9-1200x285.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-4-1200x800.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-3.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-2-1200x801.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-7-1200x950.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-6-1200x675.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-5.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-9-1200x504.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-10.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-11.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-12.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-13.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/diagram-1200x658.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2023/11/image-15-1200x908.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/image1-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/image6-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/image5-1.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/image2-1200x900.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/Screenshot-2024-03-15-at-12.51.06-1200x931.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/Screenshot-2024-03-15-at-12.49.26-1-1200x446.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/Bundy-Map.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/2023-08-29T211200Z_703442659_MT1ABCPR000X63BH6_RTRMADP_3_ABACA-PRESS-2-1200x785.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/image14.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/image11-1200x1067.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/image13-1040x1200.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/image9-1-1200x1035.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/image2-1-675x1200.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/03/image10-1200x483.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/07/image7.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/07/image11-1200x777.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/07/image2.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/07/image5.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/07/image8-1200x741.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/07/image4.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/07/image9-1200x691.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/08/image8-1-1200x753.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/08/image12-1200x675.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/08/image10-1200x675.jpg\")
![](\"https://lh7-qw.googleusercontent.com/docsz/AD_4nXe44jAhSjvhVnB5OCQwf2uSJ9KapxF_irbcY68PujxARgZJuoGF_Q719iCU4ifRhxAkUsl-yYniPNE6RKlmiztkeU7QlKKGxk2nqOettm7LqhkIB1_DFzyLb_f97OTleYqMhO5LWXzz7rDSvcv63Mu1j4Q?key=wO7lkzKX9VzMSrx0SMMocQ\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/08/ShadowTool.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/08/image12.jpg\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/08/image1.png\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/08/image3.png\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/08/image4.png\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/08/image11-1200x683.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/08/image5-1-1200x712.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/10/quick-annotation-1200x775.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/10/gazebo-and-red-building-annotation-1200x671.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/10/white-residential-tower-annotation-1200x673.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/10/tiles-and-background-building-anntoated-1200x675.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/10/gazebo-planet-imagery-1200x707.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/10/white-residential-tower-comparison-1200x673.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/10/drive-by-video-comparison-VERSION-2-1-1200x1013.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/10/window-annotation-1200x513.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/10/red-building-annotation-1200x479.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/10/august-22-sinwar-location-annotated.png\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2024/10/september-9-sinwar-location-annotated.png\"/)
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/06/Untitled-design-1-1200x960.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/05/question-mark.png\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/06/snowy-highway-Copy-1200x900.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/06/field-hills-Copy-1200x904.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/06/alley-903x1200.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/06/beach-forest-1200x900.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/06/apc.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/08/Untitled-design-1-1200x675.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/08/test25-1200x904.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/08/Screenshot-2025-08-14-at-11.05.24.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/11/RSF-shoulder-patch.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/11/Map.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/11/Map-with-images.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/11/Panorama-1200x338.jpg\")
![\"\"](\"https://www.bellingcat.com/app/uploads/2025/11/Abu-Lulu.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/D97j7ZiW4AA7hfE.jpg-orig-1024x576.jpeg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Screenshot-2019-06-26-at-22.17.24-1024x421.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/55832920_boris1.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Screenshot-2019-06-26-at-22.43.28-1024x357.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/abandoned-spanish-consulate-building-sidi-450w-421852126.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/07_16grenze_portbou-1-1024x738.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Screenshot-2019-06-26-at-22.33.32-1024x463.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/french-spanish-border-crossing-above-portbou-in-catalonia-spain-sept-E8K29N-1024x752.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Screenshot-2019-06-26-at-22.32.40-1024x591.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Selection_001-2-1024x743.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Selection_003-3.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Selection_004-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Selection_005.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Selection_006-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Selection_007-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Selection_008-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/06/Selection_009-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/brochure.jpglarge-768x1024.jpeg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_003-1024x604.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_010.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_010_box.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_006_box.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_004-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Screenshot-2019-07-05-at-19.49.47.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Screenshot-2019-07-05-at-19.54.49-1024x572.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/D-uItj5XkAEw1C5.jpglarge.jpeg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_005-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_002.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/D-t32d6XkAADqd1.jpglarge-JPEG-Image-1536-×-2048-pixels-Mozilla-Firefox_007-1024x562.png\"){kind=tracking}
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_008-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Yandex.Images-search-for-similar-images-Mozilla-Firefox_009-1024x562.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/holmes.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/D-koskVWsAcCcDt.jpglarge-1024x768.jpeg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Screenshot-2019-07-17-at-19.44.39.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_031-1024x290.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_027a.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_030.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_027-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_026.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_029.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/07/Selection_028.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/parismetro-859x1024.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Metro-red-682x1024.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Screenshot_2019-08-11-Google-Search-1024x743.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_030.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_028-1-1024x488.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/slovenia-1024x311.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Screenshot-2019-08-11-edited-at-11.11.32-1024x602.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/EC1QuFiXoAArc2z-768x1024.jpeg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/original.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/leopard2a4.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_027-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/1280px-CV90_photo-003-1024x768.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/1280px-Combined_Resolve_III_141106-A-SG416-002-1024x723.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_029-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_028-2.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_030-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/1280px-Royal_Danish_Army_2019-1024x718.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_001-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/DIB_armpatch.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_002-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_023-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_018-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/EC1QuFiXoAArc2z.-highlights-768x1024.jpeg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_004-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Denmark-map-motorways.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_005-2.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_024-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2019/08/Selection_025-2-1024x575.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/01/EPjMUYUWAAEy7vJ-1536x2048.jpeg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/01/EPjMUZKXUAU2vRr-2048x1536.jpeg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/01/Selection_003-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/01/Selection_004-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/01/Selection_005-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/01/Selection_007-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/01/Selection_008-2-2048x1041.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/01/Selection_009-2.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/01/Selection_010-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_001.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_002.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_003.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_004-2048x410.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_005.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/800px-Athlone_-_St_Marys_Church_-_20180921174650.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Screenshot-2020-02-08-at-12.17.42.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Screenshot-2020-02-08-at-12.20.05.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Screenshot-2020-02-08-at-12.22.02.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_001-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_002-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_003-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/apache1.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/apache2.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/apache3.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/apache4.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/apache6.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/apache3lg-notes-2048x1365.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_004-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_005_notes-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_012.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/EQQ3g-1WsAMIZpb.jpeg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/apache7-2-2048x1365.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_019.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_014.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_015-2.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_016.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/02/Selection_018-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_001.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_002.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_003.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/72551738_2538535489593992_7739630611849019392_o.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_064.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_065.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_005-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_009-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_010-4.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_006-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_007.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_008-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/72551738_2538535489593992_7739630611849019392_o-annotated.jpg\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_011-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_012-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_013-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_014.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_015-1-2048x807.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_017.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_019-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_018.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/03/Selection_020-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_001.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_011.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_005.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_013-2048x1177.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_008.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_014.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_015.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_016.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_017.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_018.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_020.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_021.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_022.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_023.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_024.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_025.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_026.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_027.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_028.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_029.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_010.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_001-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Screenshot_20200515-092133.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Screenshot_20200515-092201.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_009.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_004.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_005-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_006.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_007-2048x601.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_008-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2020/05/Selection_010-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-18.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-1.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/Screenshot-2022-11-01-205952.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-2.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-3.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-4.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-5.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-6.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-10.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-11-2048x1196.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-12.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-7.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-8.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-9.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-13-2048x1249.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-14.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-15.png\")
![\"\"](\"https://nixintel.info/wp-content/uploads/2022/11/image-17.png\")
![original photo](\"https://pp.vk.me/c624521/v624521010/2b6a1/uNFTPKD7r8Y.jpg\"){kind=link}
![backup](\"https://content.screencast.com/users/Voelspriet/folders/Snagit/media/ffcbc673-a40c-4e82-b717-38860617b2e2/2019-03-19_12-27-23.png\"){kind=link}
![photograph](\"https://web.archive.org/web/20190401012843/https:/www.europol.europa.eu/sites/default/files/styles/europol_large/public/images/2018-6-c16.png?itok=Sm9Mcy7E\"){kind=link}
![October 2004](\"http://geocaching.su/photos/albums/4789.jpg\"){kind=link}
![3 defence rings](\"http://armchairgeneral.com/rkkaww2/maps/1941S/Odessa_1_41.jpg\"){kind=link}
![image of the Hotel from 2013](\"https://media-cdn.tripadvisor.com/media/photo-o/04/a0/b8/bd/angkor-sayana-hotel-spa.jpg\"){kind=link}
![1.5m/year](\"https://www.google.com/imgres?imgurl=http%3A%2F%2Fonline.fliphtml5.com%2Fflqg%2Foozf%2Ffiles%2Flarge%2F1.jpg&imgrefurl=http%3A%2F%2Ffliphtml5.com%2Fflqg%2Foozf%2Fbasic&docid=Ne-mN9feepg_nM&tbnid=WE1eN3sjbxwTRM%3A&vet=10ahUKEwi6m9zKxO_lAhWIUMAKHddKCQoQMwhMKAQwBA..i&w=1391&h=1800&client=safari&bih=1273&biw=1646&q=mango%20tree%20growth%20asia%20year%201-5&ved=0ahUKEwi6m9zKxO_lAhWIUMAKHddKCQoQMwhMKAQwBA&iact=mrc&uact=8\"){kind=link}
![midway on the missile,](\"https://www.bellingcat.com/app/uploads/2020/01/EN1rHcnX0AAwuOA.png\"){kind=link}
![this](\"https://metropolitan.realestate/wp-content/uploads/property/MP-S-36984/Apartment-Bulgari-Resort-Residences-18.jpg\"){kind=link}
![this](\"https://images.bayut.com/thumbnails/283363168-800x600.webp\"){kind=link}
![banners](\"https://web.archive.org/web/20230406144522oe_/https://cdn1.ntv.com.tr/gorsel/uMAc6RRK-0-wutoLVrrASA.jpg?width=1000&mode=both&scale=both&v=1475066536758\"){kind=link}
![published](\"https://www.nepalpress.com/wp-content/uploads/2023/06/nepalese-in-russian-army-3.jpg\"){kind=link}
![this image](\"https://avangard.center/assets/images/photogallery/events/12-2021/1/iuhjdcjz-iy.jpg\"){kind=link}
![scanned letter](\"https://www.nepalpress.com/wp-content/uploads/2023/06/nepalese-in-russian-army-2.jpg\"){kind=link}
![green areas near Rushon identified in our aforementioned strategy](\"https://www.bellingcat.com/app/uploads/2023/07/Screenshot-2023-07-12-at-14.45.50.png\"){kind=link}
![photo](\"https://sobesednik.com/storage/images/20221111/5rmlZT0rGDKtqsR4467O48LNnaUINl3kwAsuNfTs.jpg\"){kind=link}
![PropSearch.ae](\"https://static.propsearch.ae/dubai-locations/burj-vista_YZ158_xl.jpg\"){kind=link}
![structural overview of the Danish Army](\"https://en.wikipedia.org/wiki/Royal_Danish_Army#/media/File:Royal_Danish_Army_2019.png\"){kind=link}
![official badge of the 2nd Brigade](\"https://en.wikipedia.org/wiki/Royal_Danish_Army#/media/File:Emblem_for_the_Danish_2_Brigade.svg\"){kind=link}
Pingback: Quiztime: How To Clean Pictures For Better Search Results – We are OSINTCurio.us
\n\t\t\t\tLS
\n\nHet hotel waar hierover gesproken wordt, wat tegenover de bibliotheek zou moeten staan, staat er niet. Gelukkig is het maar een spelletje, groet..
\n\t\t\t\t- \n\n\t\t\t\t\t
\n\t\t\t\t\t\n\t\t\t\t\t\t![\"\"](\"https://secure.gravatar.com/avatar/a8dbec59c215a7304ee0e9f72018ce377e1a7869da6ef4dc675dc1de3278de1e?s=50&d=mm&r=g\")
Nixintel \n\t\t\t\t\t\t\t\t\nLoading... \n \n\t\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t \n\t\t\t\t\t\t\t \t\t\t\t\t \n\n\t\t\t\t \n
\nThe hotel is directly opposite the library on the other side of the Daugava River?
\nI use Imgcleaner.com to remove unwanted objects from the image. Its good.
\nGreat OSINT quiz on cleaning pictures! Using PicPicAi image editing tools can enhance photo clarity, making search results more accurate. A fun, practical guide for better visual investigations!
\n