I noticed that when reading LLC4320 data using intake on the google cloud, the SSS variable looks slightly different from the other fields.
While all variables have only the main variable as the data variable:
sst = cat.LLC4320_SST.to_dask()
gives only SST as a data variable,
‘sss = cat.LLC4320_SSS.to_dask()’ gives a xarray dataset with SSS, XC, YC, and iter as Data variables. It is not really a problem, but can raise an annoying bug when trying to do a xr.merge later . which can also be easily resolved, but requires handling salinity different than the other variables.