CDO

hi Anne!

are your realy sure that sss_clim_core.nc is the reference file with the right dimensions??

netcdf sss_clim_core {
dimensions:
        x = 180 ;
        y = 193 ;
        time = 12 ;
variables:
        double sss(time, y, x) ;
                sss:missing_value = -99. ;
                sss:units = "psu" ;
                sss:long_name = "Sea surface salinity" ;
data:

 sss =
  -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, 
    -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, 
    -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, 
    -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, 
    -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, 
    -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, 
    -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, 
    -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, 
    -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, 
    -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, 
    -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, 
    -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, -99, .....

is has nothing like data variables for the dimensions or units or so. To me the other file seems to be in a good shape, though..

cheers
ralf

anyway: to set the time axis of a file (no matter if a time dimension already exist or not), you can use the operator settaxis like this

cdo -f nc -settaxis,2003-12-12,12:00:00,1days sss_clim_core_LGM_smoothed.nc sss_clim_core_withTime.nc

sss_clim_core.nc is not CF-conform, so you get a warning with the above call, but still a proper time axis is created.

hth
ralf

Hey Ralf,

Thank you. Yes, the sss_clim_core.nc is the reference file, and I agree it is a bit of a strange file... The other one is the one I made, with a new sss variable field which should replace the one in the reference file. The thing is, changing the time axis of the sss_clim_core_LGM_smoothed.nc file does not remove the time dimension (while keeping 12 data frames) like present in the reference file. But I managed in the meantime to successfully extract the field (they are both 180x193x12, and that is the most important) without changing the sss_clim_core_LGM_smoothed file. Making a non-CF-conform file using CDO is probably also not what is meant to be done with it

Thanks again and have a nice day,
Anne