You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have done a comparision between different sources of ADM 2 data, including your upload of the GAUL2015 data.
My results are:
GADM has the most polygons (356.508), however it seems like it will not be updated anymore (as a new update was announched for fall 2023 but was never released)
GeoBoundries has the second most polygons (49.346) and also polygons in the area's where I am interested in. The source seems to be maintained
The ESRI Global Polio Adminstrative Boundries has 48.745 polygons but focuses mostly on Africa.
The GAUL2015 has 38.258 polygons but seems to be not updated anymore.
The OSM has no worldwide coverage of ADM 2 regions
Therefore my advice is:
Use Geoboundries as main source of ADM 2 data
Supplement countries that are not well covered in the ADM 2 data (e.g. Australia, Brazil, and many more) of GeoBoundries with other ADM 2, ADM 3 and ADM 4 sources from Geoboundries, Who's On First and OSM-Boundries.
Hope this is helpful. Let me know if you have any questions.
Thanks Levi for the contribution.
Indeed GeoBoundaries seems to be the most solid and updated at the moment; however, for inclusion in WB tools, boundaries must first be approved by the WB GeoSpatial team. While GAUL 2015 were previously approved, other sources are not.
This is particularly important when releasing public maps under WB logo, specifically for those area that are disputed / conflict area (India/Pakistan/China boundaries, and others).
So, unfortuately we are stuck with GAUL 2015 until a new standard is approved.
The global boundaries currently in use are based on GAUL 2015, which were the source selected for the official world bank boundaries. They have then been optimised in several ways for the purpose of use in TH.
As of 2023, they are outdated and need to be replaced for all administrative levels (ADM0 - ADM1 - ADM2)
The text was updated successfully, but these errors were encountered: