Romain Hugonnet

Hypsometric + Proximity should certainly work. However only if you account for changes in proximity over time. So one always need to interpolate between the old and new outlines!

My little finger is telling me that the most robust co-registration method will indeed be a weighted co-registration, possibly with Bayesian snow-cover drift estimation :) I have a MSc student...

Looks like there is a Python API fairly practical to use: https://github.com/ethz-asl/libpointmatcher/blob/master/examples/python/icp.py But the install doesn't look very user-friendly indeed, see: https://libpointmatcher.readthedocs.io/en/latest/CompilationPython/

Errors due to SciPy 1.13 fixed by using the slower method in #185 (see #186). Method `dok_matrix.update` still does not work, but there might be a new solution to keep...

Hi @StrangeFrog, TBC means "To Be Communicated". I have updated it by "only available on demand", because these data are too large to distributed by our institutions.

@StrangeFrog The added note in the README should make it clear that it is me you need to contact, and is also a good reminder that this data is very...