I just realized that without data for non-flood years then the model will probably overpredict flood.
I will like to join your project. I just finished learning time series analysis
yes.. you are right: am trying to get data for 2014,15 -19 flood/rainfall record in malawi
It's easy to find climate data, but flood extent? I was able to find a flood extent for 2019 but that's because it was a flood year.
So, it seems that we have to derive flood extent from remote sensing imagery for every year. This looks like it would be a good baseline: https://developers.google.com/earth-engine/datasets/catalog/JRC_GSW1_1_YearlyHistory
I would be happy to join forces!