Account for stake measurements that are located outside the glacier outlines provided by RGI

[JulianBiesheuvel]

Some stake measurements might be so outdated that they no longer intersect with the recorded RGI glacier outlines due to glacier retreat. As a result, these stakes cannot be matched with a specific glacier or assigned an RGI code. However, to make these historical data usable, it's important to find a way to still associate these measurements with a glacier and obtain an RGI code.

One possible solution is to group or cluster nearby stakes and identify those with a similar RGI. Another approach could involve using a graphical user interface (GUI) that allows users to select points and manually assign them to a glacier.

[khsjursen]

We could possibly use geopandas.sjoin_nearest to do this: https://geopandas.org/en/stable/docs/reference/api/geopandas.sjoin_nearest.html

I think this would work if the problem is generally that these points are associated with retreating glacier tongues such that we can expect that the nearest geometry is the correct one.

If we use this we might set a default max_distance such that in case of incorrect POINT_LAT and POINT_LON we do not connect the point with an arbitrary RGIId.