Issues with ERA5 ml2pl conversion
Added by Benjamin Kirbus almost 4 years ago
Dear all,
I struggle with a simple task, which has been asked here prior. But none of the proposed fixes seem to work. I use ERA5 data on model levels (e.g. ml 80 to 137). Now I want to convert from ml to pl, using the cdo operator "ml2pl". My workflow is as follows:
- Download 2D & 3D fields from MARS catalogue, e.g. PS & Q
- Convert grib -> netCDF, merge
- cdo -f nc4c copy PS.grib PS.nc
- cdo -f nc4c copy Q.grib Q.nc
- cdo -merge PS.nc Q.nc testdata.nc
- Now, I try to use ml2pl:
- cdo ml2pl,90000,100000 testdata.nc test_ml2pl.nc
No matter what I do, it always fails with "Warning: No 3D variable with hybrid sigma pressure coordinate found!". I will upload the sample dataset; maybe I make a very obvious mistake..?
Thanks a lot, all the best
Benjamin
| TEST_data.nc (3.37 MB) TEST_data.nc |
Replies (3)
RE: Issues with ERA5 ml2pl conversion - Added by Uwe Schulzweida almost 4 years ago
Dear Benjamin,
The operator ml2pl expects the input 3D variables to be present on all hybrid layers. In your netcdf file only 58 of the 137 layers are present. Unfortunately ml2pl is not implemented for a subset of layers, sorry.
Cheers,
Uwe
RE: Issues with ERA5 ml2pl conversion - Added by Benjamin Kirbus almost 4 years ago
Hi Uwe,
thank you for your quick response - that makes an awful lot of sense! I will think about downloading all layers from now on.
Cheers,
Benjamin
RE: Issues with ERA5 ml2pl conversion - Added by Benjamin Kirbus almost 4 years ago
FYI, (not surprisingly) it all works as intended when downloading all model levels. Thanks again, Uwe!