Closed oloapinivad closed 5 years ago
Paolo, I really do think that tos in the 3hr table should be extracted from IFS, I believe ece2cmor3 chooses IFS by 'realm matching' for this variable...
...and I am not sure whether 3hr-hourly ocean output is properly supported...
BTW in the newest CMIP6 tables the tos modelling realm in the 3hr table is ocean, so I guess we should have NEMO produce this variable? See #224
I see your point. I guess that - theoretically - it should be a little bit more correct to start from NEMO for coupled runs and from IFS when you have AMIP ones. I am pretty sure that in an older commit I was able to cmorize correctly from the NEMO 3h file. My present concern is that in the actual configuration I cannot cmorize it neither from IFS:
Indeed, if I set a varlist as:
{
"3hr": [
"tos"
]
}
and I launch the --atm cmorization with ifspar.json I get nothing
INFO:ece2cmor3.cmor_target:CMOR tables data_specs_version : 01.00.23
INFO:ece2cmor3.cmor_target:CMOR tables cmor_version : 3.3
INFO:ece2cmor3.cmor_target:CMOR tables Conventions : CF-1.7 CMIP-6.2
INFO:ece2cmor3.cmor_target:CMOR tables table_date : 21 March 2018
INFO:ece2cmor3.taskloader:Found 1 cmor target variables in input variable list.
INFO:ece2cmor3.taskloader:Created 0 ece2cmor tasks from input variable list.
INFO:ece2cmor3.ece2cmorlib:Selected 0 IFS tasks from 0 input tasks
INFO:ece2cmor3.grib_filter:Searching for previous month file of /marconi_scratch/userexternal/pdavini0/tmp_cmor/qctr_26134/atm_1950_1/linkdata/ICMGGqctr+195001
INFO:ece2cmor3.grib_filter:Found previous month gridpoint file /marconi_scratch/userexternal/pdavini0/tmp_cmor/qctr_26134/atm_1950_1/linkdata/ICMGGqctr+194912 and spectral file /marconi_scratch/userexternal/pdavini0/tmp_cmor/qctr_26134/atm_1950_1/linkdata/ICMSHqctr+194912
INFO:ece2cmor3.ifs2cmor:Inspecting 0 tasks.
INFO:ece2cmor3.ifs2cmor:Validated 0 tasks for processing.
INFO:ece2cmor3.ifs2cmor:Executing 0 IFS tasks...
Which tables are this? Jon Seddon's primavera ones or the new ones?
Still Jon Seddon's ones.
Il giorno mar 4 set 2018 alle ore 17:18 Gijs van den Oord < notifications@github.com> ha scritto:
Which tables are this? Jon Seddon's primavera ones or the new ones?
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/EC-Earth/ece2cmor3/issues/231#issuecomment-418407584, or mute the thread https://github.com/notifications/unsubscribe-auth/AbCe60pRpbsUABEM7cdvbsJwNCZEqMceks5uXpmqgaJpZM4WZKwS .
Hmm I suspect it is the realm key. Can you change line 24 in ece2cmor3/ece2cmor3/cmor_target.py to:
realm_key = "realm"
instead of "modeling_realm"? Then rerun python setup.py install
and test whether it works?
Hi,
setting the realm_key as you said works. I have now the tos from the IFS files. I only have to double check if this compromises any other variable. More in general, I still feel that it would be more correct using the NEMO files, if available. Thanks, P
Il giorno mar 4 set 2018 alle ore 18:04 Gijs van den Oord < notifications@github.com> ha scritto:
Hmm I suspect it is the realm key. Can you change line 24 in ece2cmor3/ece2cmor3/cmor_target.py to:
realm_key = "realm"
instead of "modeling_realm"? Then rerun python setup.py install and test whether it works?
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/EC-Earth/ece2cmor3/issues/231#issuecomment-418425100, or mute the thread https://github.com/notifications/unsubscribe-auth/AbCe6x07tlFX5FMHWBrHhFeEyDBFD6Piks5uXqSpgaJpZM4WZKwS .
@goord, is this fix permanent? I can close the issue in this case.
Hi Paolo, I propose to make a new primavera-stream 2 branch to put this fix in...
done, created branch primavera-stream2
Hi all,
I am trying to cmorizing 3hrs tos from the correspondent grid_T file, but I am getting this error.
However, the variable is there:
The nemopar.json and the cmip6 tables seem to point correctly to tos.
Any experience on this? I saw a few issues involving tos but this seems new.