when switching to data sources other than ERA5, you can generate the data using, but the model fails when trying to run see notebooks in https://github.com/eWaterCycle/ewatercycle-pcrglobwb/tree/fix-cmip-forcing for an error. I think multiple issues are at play here: - naming of variables? - input needs to be regridded to match parameter set resolution, which the recipe does not do - input needs to be padded broader than the domain for some reason