CDO

Hi Tarek,
the land sea mask file already has the missing value set to zero (type unsigned short). The following example demonstrates how to use your lsm file.

Example: Generate the global topography data, remap to same grid as the land sea mask, and multiply it with the land sea mask

cdo -f nc -mul land10km_wgs84_con.nc -remapbil,land10km_wgs84_con.nc -topo outfile.nc

Hi Karin,

Thanks for the swift response. Yes, the missing values are set to zero type ubyte. I want to change the variable data type to float and assign miss value to e.g., -9999 to avoid clipping my map that already has zeros as data (0 precipitation at a given location at a given timestep). This is how it looks like (please see attached).

Best regards,
Tarek

Try

cdo -b f32 -setmissval,-9999. -setmisstoc,-9999. land10km_wgs84_con.nc lsm.nc

Hi again,

Thanks Karin, I arrived at the same result at the same time, which I ended up using. However, this was a bit more problematic for another set of data that already has a sea land mask. So, I resorted to first applying the lsm.nc mask to the data which gave me a lot of 0 values beyond the land (i.e., the ocean), and then divided the data itself by another mask with miss value = 0, which removed the zeros from the ocean only using

cdo div data_file.nc lsm_zeros.nc outfile_masked.nc

I mention this in case some has encountered a similar issue dealing with already masked data.

Again, thanks for the help and quick responses.

Best regards,
Tarek