Apparently, the spatial trend plot for GRACE is wrong. I don't know if I can revise this notebook myself since I don't have access to AWS or any zarr stuff in the beginning of notebook.
I have rerun the notebook with the repaired himatpy, so the trend map is now correct.
I ran this in a new instance of Pangeo where I could deploy dask workers as needed.
Apparently, the spatial trend plot for GRACE is wrong. I don't know if I can revise this notebook myself since I don't have access to AWS or any zarr stuff in the beginning of notebook.