CDO

All,

I'm wondering if CDO can do something or not (my guess, not, but you never know). I have some station data (one grid point) that looks like this:

$ cdo sinfon LEF_200901_co2.nc4
   File format: netCDF4 classic
    -1 : Institut Source   Ttype    Levels Num  Gridsize Num Dtype : Parameter name
     1 : unknown  unknown  instant       6   1         1   1  F64  : CO2        
     2 : unknown  unknown  instant       6   1         1   1  F64  : Uncertainty
   Grid coordinates :
     1 : lonlat       > size      : dim = 1  nx = 1  ny = 1
                        lon       : first = -90.2732  degrees_east
                        lat       : first = 45.9451  degrees_north
   Vertical coordinates :
     1 : generic                m : 11 30 76 122 244 396 
   Time coordinate :  744 steps
     RefTime =  2009-01-01 00:00:00  Units = minutes  Calendar = standard
  YYYY-MM-DD hh:mm:ss  YYYY-MM-DD hh:mm:ss  YYYY-MM-DD hh:mm:ss  YYYY-MM-DD hh:mm:ss
  2009-01-01 00:00:00  2009-01-01 01:00:00  2009-01-01 02:00:00  2009-01-01 03:00:00
  2009-01-01 04:00:00  2009-01-01 05:00:00  2009-01-01 06:00:00  2009-01-01 07:00:00
  2009-01-01 08:00:00  2009-01-01 09:00:00  2009-01-01 10:00:00  2009-01-01 11:00:00
  2009-01-01 12:00:00  2009-01-01 13:00:00  2009-01-01 14:00:00  2009-01-01 15:00:00
...

So for the month of January, there is station data on 6 levels, and most months look like this. However, for some months, only certain levels are reported, e.g., July:

$ cdo sinfon LEF_200907_co2.nc4
   File format: netCDF4 classic
    -1 : Institut Source   Ttype    Levels Num  Gridsize Num Dtype : Parameter name
     1 : unknown  unknown  instant       3   1         1   1  F64  : CO2        
     2 : unknown  unknown  instant       3   1         1   1  F64  : Uncertainty
   Grid coordinates :
     1 : lonlat       > size      : dim = 1  nx = 1  ny = 1
                        lon       : first = -90.2732  degrees_east
                        lat       : first = 45.9451  degrees_north
   Vertical coordinates :
     1 : generic                m : 30 122 396 
   Time coordinate :  744 steps
     RefTime =  2009-07-01 00:00:00  Units = minutes  Calendar = standard
  YYYY-MM-DD hh:mm:ss  YYYY-MM-DD hh:mm:ss  YYYY-MM-DD hh:mm:ss  YYYY-MM-DD hh:mm:ss
  2009-07-01 00:00:00  2009-07-01 01:00:00  2009-07-01 02:00:00  2009-07-01 03:00:00
  2009-07-01 04:00:00  2009-07-01 05:00:00  2009-07-01 06:00:00  2009-07-01 07:00:00
  2009-07-01 08:00:00  2009-07-01 09:00:00  2009-07-01 10:00:00  2009-07-01 11:00:00
...

Note how for July, only 30, 122, and 396 levels are reported.

A straight mergetime fails, of course:

$ cdo mergetime LEF_200901_co2.nc4 LEF_200907_co2.nc4 test.nc4

cdo mergetime (Abort): Input streams have different number of records per timestep!

but what I'd like in the end is to mergetime all the monthly files into a single yearly file, and, where months are missing levels, have missing value for those levels for that month.

Obviously, the challenge is the "adding missing levels". Once that is done, mergetime can do the rest. Is there a good way to do this with CDO, or would it be easier to write a netCDF Python script, say, to just make new files based on the old ones where I create and fill in the missing levels.

Thanks,
Matt