--- license: apache-2.0 pretty_name: canVODpy reference stores tags: - GNSS - GNSS Transmissometry - vegetation optical depth - VOD - Icechunk - Zarr - remote sensing size_categories: - n<1K --- # canVODpy reference stores The reference stores of [canVODpy](https://github.com/nfb2021/canvodpy), which computes vegetation optical depth (VOD) from GNSS signals (GNSS Transmissometry). They are GNSS stores and VOD stores (Icechunk) built from the canVODpy test data with a fixed set of settings, called setting profiles. New canVODpy code must reproduce them: a change that alters the stored observations, the satellite geometry or the VOD shows up as a difference. The stores were built by canVODpy at commit [`bcb6b808`](https://github.com/TUW-GEO/canvodpy/commit/bcb6b808ba87a99a8b956258c14b2ed0fd097516) (with [`7ec890b8`](https://github.com/TUW-GEO/canvodpy/commit/7ec890b897a82afa6aade2126a644237b3d6d3e2) merged in) from the test data at commit [`d4ab73e`](https://github.com/TUW-GEO/canvodpy-test-data/commit/d4ab73eae6ce5568e5df152acfc52fb5410ead47). The releases of the test data are archived on Zenodo ([10.5281/zenodo.19708759](https://doi.org/10.5281/zenodo.19708759), all versions). ## Setting profiles | Profile | Sites | Settings | |---|---|---| | `default` | Rosalia (RINEX v3.04, SBF, NMEA), MOFLUX (RINEX v2.11), ExampleSite (stripped RINEX v3.05), Hainich (NMEA); one day each | final SP3 orbits (CODE), up to 15 signals (SIDs) per system, no preprocessing, SNR only, VOD of each receiver pair | | `sbf-broadcast` | Rosalia (SBF) | as `default`, but the satellite geometry from the broadcast ephemerides the SBF files carry, instead of the final SP3 orbits | | `gridded` | as `default` | as `default`, plus observations aggregated to 1-minute means and assigned to the cells of a 2° equal-area grid | The profiles themselves are in the canVODpy repository, under `reference/profiles/`, and the [reference stores guide](https://canvodpy.eu/guides/reference-stores/) describes them in full. ## Files One zip per store: `/-gnss.zip` and `/-vod.zip`, 26 zips and 165 MB in all. `SHA256SUMS` lists the sha256 of every zip. Each zip unpacks into `stores///`, so the zips of one profile can be unpacked into the same folder. The zips are deterministic: packing the same stores again gives the same sha256. They hold only the stores: no settings files, no logs and no local paths. The stores name no real person or computer: their metadata is the made-up identity of canVODpy's tests, and their commits don't record the computer that built them. Each store records the receiver position its test data carries, links to this dataset and names the other store of its site (GNSS or VOD). ## Use canVODpy downloads, checks and unpacks the zips itself. From a checkout of the canVODpy repository with the test data: ```bash just reference-check ``` This runs every profile again and compares each new store with its reference store. It reads the zips of one commit of this repository, named in `canvodpy/src/canvodpy/reference-stores/published.yaml`, and checks each zip's sha256, so a later change here never changes what a given canVODpy version compares with. ## Citation Cite canVODpy ([10.5281/zenodo.18496233](https://doi.org/10.5281/zenodo.18496233)) and the test data the stores come from ([10.5281/zenodo.19708759](https://doi.org/10.5281/zenodo.19708759)). ## License Apache License 2.0, as the test data.