CDO

hi!

Some sample of your data would be helpful

cheers
ralf

attached. thank you ralf.

CDO needs ntecdf data generally following the CF convention: http://cfconventions.org/

Your data does not:

        int nCorners(nCorners) ;
                nCorners:long_name = "Ground Pixel Corner" ;
                nCorners:units = "1" ;
                nCorners:description = "The dimension representing the ground pixel corners.  There are four corners per ground pixel." ;
        double nTimes(nTimes) ;
                nTimes:long_name = "Time at Start of Scan (TAI93)" ;
                nTimes:units = "seconds since 1993-01-01" ;
                nTimes:description = "The dimension representing the observation times in the granule.  The values are in TAI93 (time in continuous seconds since 12 a.m. UTC on January 1, 1993)." ;
        float nWavel(nWavel) ;
                nWavel:long_name = "TOMS Wavelength" ;
                nWavel:units = "nm" ;
                nWavel:description = "The dimension representing the TOMS wavelengths.  The values range from 331 to 380 nm." ;
        int nXtrack(nXtrack) ;
                nXtrack:long_name = "Cross-Track Position (1-Based)" ;
                nXtrack:units = "1" ;
                nXtrack:description = "The dimension representing the cross-track position of the observation in the scan line.  The values range from 1 to 35." ;

group: ANCILLARY_DATA {
  variables:
        float SurfaceAlbedoOceanCorrected(nTimes, nXtrack, nWavel) ;
                SurfaceAlbedoOceanCorrected:_FillValue = -1.267651e+30f ;
                SurfaceAlbedoOceanCorrected:long_name = "Surface Albedo Ocean Corrected at 331, 340, 360 and 380 nm" ;
                SurfaceAlbedoOceanCorrected:units = "1" ;
                SurfaceAlbedoOceanCorrected:description = "The surface albedo ocean corrected at 331, 340, 360 and 380 nm." ;
                SurfaceAlbedoOceanCorrected:valid_min = 0.f ;
                SurfaceAlbedoOceanCorrected:valid_max = 1.f ;
        float TerrainPressure(nTimes, nXtrack) ;
                TerrainPressure:_FillValue = -1.267651e+30f ;
                TerrainPressure:long_name = "TerrainPressure" ;
                TerrainPressure:units = "hPa" ;
                TerrainPressure:valid_min = -20.f ;
                TerrainPressure:description = "The terrain pressure (in hPa) at the center of the ground pixel (from the TOMS Level 1B file)." ;
                TerrainPressure:valid_max = 1000.f ;

CDO is not a general netcdf editor, but you could use the NCO package - it supports the netcdf-standard in a broader sense.

hth
ralf

Could you tell me a little more where this file fails CF-convention?

I tried
http://puma.nerc.ac.uk/cgi-bin/cf-checker.pl
and it gave a pass.

Thanks,

Ok, CDO is designed for working with climate/weather, so it needs some kind of grid. There is no such coodinate in your file, nothing that could be interpreted as longitudes or latitudes. You can have a look into the documentation about it.

sorry for being unclear

Esp. I dont see your file following the coordinates recommendations: http://cfconventions.org/cf-conventions/cf-conventions.html#coordinate-system

Is it possible to run cdo setgrid to fix this issue?

In the group GEOLOCATION_DATA, two variables: Latitude/Longitude could be used ?

Here is a related example. (attached)

Panoply shows:

dimensions:
lat = 1800;
lon = 3600;

but ncdump shows:

dimensions:
lat = 1800 ;
phony_dim_1 = 3600 ;

cdo sinfo shows:

Grid coordinates :
1 : generic : points=3600

CDO needs a data variable for be able to work with that file. you could put in an empty one, that makes use of the given dimensions.

Is it true CDO does not support groups? It's an issue reported 6 years ago. Looks like v1.6.9 does not.

What do you mean by a data variable? lat/lon are both dimensions and variables in the file.