Bias-adjustment ISIMIP3BASD

[saschahofmann]

Yesterday I mentioned in the talk with @coxipi and @Zeitsperre, we are interested in the bias adjustment method I think used for the IPCC AR6: ISIMIP3BASD. I can see that its already part of ibicus, so I am looking forward to the xsdba split 👍

[aulemahal]

Yes! This is exactly one of the reason for the split : avoid redoing all the work. Preliminary tests for wrapping ibicus in a xsdba/xarray setup did give good results, but did not go very far.

FYI : the test is in the wrap-ibicus branch. The conclusion are hidden in a private convo I think, but depending on the settings chosen, the wrap gave a similar performance to the original, while abstracting the multi-processing overhead. However, I did only test one of the simplest method. ISIMIP does not look simple at all...