CDO

Hi Julien!

Have you checked the results when using the --eccodes option? It enables grib-handling for GRIB1/2 using eccodes,only. I think with that backend the tables might be more accurate. You need a cdo-1.9.6 or later.

hth
ralf

Hi Ralf,

Thanks for the quick response! Indeed, using the --eccodes option instead of -t ecmwf worked for previously unknown variables (e.g., tciw or tclw).

The only small issue now is that the -R option (to convert from reduced to regular Gaussian grid) doesn't seem to work with --eccodes, so for example I can't convert directly from Grib/reduced GG to netcdf/regular GG in a single operation if I use the two options together (-R --eccodes). The only way I found to make it work is to do it in two steps with an intermediate output file in-between.

As a note, I am using cdo-1.9.6.

Thanks,
Julien

if you upload some sample of the input I could give it a try

Attached is a file containing ERA5 monthly mean data for total cloud ice water on a reduced Gaussian grid.
Thanks!

Use setgridtype,regular instead of the CDO option -R:

cdo --eccodes  sinfon -setgridtype,regular era5_moda_sfc_199802.grib

   File format : GRIB
    -1 : Institut Source   T Steptype Levels Num    Points Num Dtype : Parameter name
     1 : ECMWF    unknown  v instant       1   1    819200   1  P16  : tciw          
   Grid coordinates :
     1 : gaussian                 : points=819200 (1280x640)  F320
                              lon : 0 to 359.7188 by 0.28125 degrees_east  circular
                              lat : 89.78488 to -89.78488 degrees_north
   Vertical coordinates :
     1 : surface                  : levels=1
   Time coordinate :  1 step
     RefTime =  1998-01-31 18:00:00  Units = hours  Calendar = proleptic_gregorian
  YYYY-MM-DD hh:mm:ss  YYYY-MM-DD hh:mm:ss  YYYY-MM-DD hh:mm:ss  YYYY-MM-DD hh:mm:ss
  1998-02-01 06:00:00

Perfect! Thank you so much for your help!