Question on different list of countries/groupings in different datasets and IOSID fractions

Created by: rosapietroiusti

Hi! Thanks so much for making this available to the ISIMIP community, I have already found it super useful.

I see there are slightly different lists of countries for example between countryData.json (195 official UN countries), countrymasks_fractional.nc (which understandably has more, 224 groupings excluding 'world') but also that there are differences between countrymasks_fractional.nc and countrymasks.geojson (with the latter having 208 countries/groupings, excluding for example IOSID, PSID etc. that are in the countrymasks). It would be great to have a metadata file that has all the groupings that are in the final countrymasks_fractional file and was wondering if there were ways or plans to harmonize this.

(I originally opened this issue thinking some official/observer 195 UN countries were missing in countrymasks_fractional, but see now that they are missing in countrymasks.geojson but mostly present in countrymasks_fractional.nc, except for very small countries like Liechtenstein, San Marino and Monaco, which is unlikely to be an issue in analyses).

Also, maybe this is a separate issue, but I am wondering if some grid cells are assigned a fraction of their area unnecessarily to groupings like IOSID, PSID? For example, part of Mauritius seems to be assigned to IOSID, and part of Singapore also to IOSID, and I am wondering if this reflects something administrative or is unnecessary and the fraction could be assigned respectively to Mauritius and Singapore. Here for example the situation in a pixel in Mauritius: