Small issues with the Millan 22 dataset on Theia

[fmaussion]

The dataset is available on https://www.sedoo.fr/theia-publication-products/?uuid=55acbdd5-3982-4eac-89b2-46703557938c

I'm just writing all this here until we find a better place to report this:

• [ ] there are two files for the same tile 17.7: THICKNESS_RGI-17.7_2021July09.tif and THICKNESS_RGI-17.7_2021July23.tif. The current workaround is to ignore the latter (7_2021July23)
• [ ] there is one orphan file in the velocity dataset (V_RGI-1.6_2021July01.tif): there is no other variable for the tile 1.6 than V. The current workaround is to ignore this file.
• [ ] some ERRTHICKNESS files do not match the spatial extent or even coordinates of their equivalent THICKNESS files. This is a problem to compute errors on the same grid. The workaround is to reproject the data yourself. Possibly this occurs with velocity as well, I didn't check
• [ ] Region 05 (Greenland) does not have velocity data. This is acknowledged on the website (one has to refer to MEASURES) but still there is a file to download with some tiffs in it. This is a tiny problem though.

[fmaussion]

Here are the files with coordinate issues for THICKNESS:

RGI-1.1_2021July01 has a different grid size for thickness and error
RGI-5.1_2022February24 has different coordinates for thickness and error
RGI-5.5_2022February24 has a different grid size for thickness and error
RGI-10.1_2021July09 has different coordinates for thickness and error

[fmaussion]

For velocity its only one file:

RGI-1.1_2021July01 has a different grid size for v and vx/vy