How to mask fill values in existing netcdf file?
In case ncdump -h command does not recognises your fill or missing values you can do the following.
Mask the NaNs in fillvalue of the file (with a variable gpp) using xarray.
import xarray as xr
ds = xr.open_dataset("abc.nc")
ds_new = ds.fillna(1.e+36)
\# ds_new = ds.fillna(1.e+36).drop_vars("areacella") # incase you also want to remove a variable.
ds_new.gpp.attrs["_FillValue"] = 1.e+36
ds_new.gpp.attrs["missing_value"] = 1.e+36
ds_new.to_netcdf(f"abc_new.nc", unlimited_dims={'time':True})
Change the lat/lon coordinates using xarrays
lon_values = np.asarray(ds.lon.values-180, dtype=float)
ds = ds.assign_coords({"lon" : (lon_values )})