A single NetCDF file from this model contains 500 years of daily data.
Right. This is probably the issue. Specify chunks={"time": 365} in open_mfdataset. You should see increased parallelism after that.
A single NetCDF file from this model contains 500 years of daily data.
Right. This is probably the issue. Specify chunks={"time": 365} in open_mfdataset. You should see increased parallelism after that.