OGGM / oggm

Open Global Glacier Model
http://oggm.org
BSD 3-Clause "New" or "Revised" License
218 stars 105 forks source link

Replacing calculated centerlines with user created centerline shapefile #1325

Open MarySorensen6 opened 2 years ago

MarySorensen6 commented 2 years ago

Hello

I am a grad student trying to reconstruct the LIA glacial ice extent for an alpine glacier in Chile (Mt. Sierra Nevada). The problem is that the calculated centerlines do not go down the right valley; it does not follow the geomorphology of the landscape or the glacier itself (it does not flow perpendicular to the crevasses). Is it possible to have OGGM run off of user created shapefiles?

fmaussion commented 2 years ago

Interesting. Out of curiosity, which glacier is that? Any screenshot?

There is no fundamental obstacle to this use case, but it will require you to either dig into OGGM's code or us offering options to do that in a general sense (quite some work). First it might be a good idea to really look at it and see if we can do something else first.

MarySorensen6 commented 2 years ago

Attempting to Reconstruct LIA conditions and Ice Extent at Mt. Sierra Nevada, Chile.pdf

Here is a brief description of what I have done so far and my results

fmaussion commented 2 years ago

OK yes I see. Nice work!

Well, I think that here you have a quite complex configuration for a simple flowline glacier such as OGGM. You may manage to trick the model as you have attempted indeed, but many questions will remain about the validity of the flowline assumption in your case, etc. Still, I see a few ways forward:

To prescribe the geometries into OGGM, you may trick the model by rewriting the output of the tasks after they have been run. For example rewriting centerlines.pkl after running compute_centerlines, and same for the downstream line. I think the in your case it might be meaningful to have only one flowline instead of multiple ones.