Updated Data Plans
Hi I was wondering if there are any plans to update this dataset to include the remaining data from Sept 2025 to Present? Thanks
Hi, there isn't for this, sorry. I do have an ongoing archive of data available here though (although the newer data is only available in grib format, I haven't been able to convert it): https://source.coop/bkr/icon
Hi @jacobbieker ,
I saw your note that the newer ICON-Global data in your Source Cooperative mirror is only in GRIB and hasn't been converted yet. I'd be glad to help with that, unpaid. I'm a retired software architect (40 years) and now spend my time on data-engineering help for open science projects.
What I have in mind:
- Start from OCF's download_combine_upload_icon.py, so the output matches the existing archive's layout (variables, chunking, native grid).
- Convert a few sample runs first and check them against the older OCF Zarr data, so you can see whether it's consistent before anything larger runs.
- Hand over the code and the samples for you to review. You'd stay in control of what goes into the archive and of running it at scale.
One question: the icon-global-raw folder appears to start on March 12, 2026, while the OCF archive ends in August 2025. Is the September 2025 to March 2026 period stored somewhere else, or is that gap permanent?
Would this be useful? If you have preferences, such as the format, which variables or the period to start with, or another reason it's been hard to convert, I'd like to hear them first.
Thanks for keeping this archive going. It's clearly widely used.
Mousa Shaya
github.com/mshaya · shayaml.com
@Jack-Kelly @sol-ocf , gently bumping this in case it got buried. My offer
still stands: I can write and test the GRIB-to-Zarr conversion for the
ICON-Global data from Sept 2025 onward, based on the existing OCF
pipeline script, validate it against the old Zarr output, and hand over
the code for you to run at scale.
If someone else is already on this, or you've decided not to continue the
archive, just let me know and I'll stop asking. Thanks!