search

EC-Earth / ece2cmor3

Post-processing and cmorization of ec-earth output
Apache License 2.0
13 stars 6 forks source link

No nemo variables cmorized in Emon table #209

Closed aearamos closed 6 years ago

aearamos commented 6 years ago

I'm having some issues with the latest tables and ece2cmor3. I have a few nemo variables in my varlist.json such as thetaot, thetaot2000/300/700 and t20d that can be found in the Emon table. However, when I run cmorization for the ocean, I only get cmorized nemo variables for tables Omon and SImon, but not Emon.

I noticed then that some of these variables are part of the list-of-identified-missing-cmpi6-requested-variables.xlsx file, so I tried deleting them from the table, but no success so far. I still don't have the nemo variables in the Emon table, and I don't get any error messages regarding these variables.

I'm sure I have the sources/targets in my nemopar.json file, I have the outputs in my nemo netcdf files and I have the right tables.

Any clues? Or did anyone have a similar problem? Is there a way to "force" cmorization of identified missing variables?

Cheers.

aearamos commented 6 years ago

@treerink after some debugging I found what my problem was. It wasn't related to the identified missing tables, but to a difference in the name for the "realm" variable in cmor_target.py. In ece2cmor3, the realm_key is called "realm", while in the CMIP6 tables, they changed the name to "modelling_realm". This worked fine for most of the variables, but for me, the program thought the variables such as thetaot, thetaot2000/300/700, and t20d had a "realm" of land, instead of ocean. All I did was change the realm_key in cmor_target to "modeling_realm".

I hope this works.

goord commented 6 years ago

Hi @aearamos have you committed your fix into a branch?

aearamos commented 6 years ago

Hi, @goord. I did not but will do it ASAP.