[CMORization request] ssp245-covid

[tjiputra]

Mandatory information:

Full path to the case(s) of the experiment on NIRD
/projects/NS2345K/noresm/cases

experiment_id
ssp245-covid

model_id
NorESM2-LM

CASENAME(s) and years to be CMORized
NSSP245frc2covtwobli_01_f19_tn14_20200823
2015-2050

Optional information

(additional information, if the experiment is branched/hybrid restart from previous parent experiment; you may find more information relevant to the experiment_id here: https://github.com/NorwegianClimateCentre/noresm2cmor/blob/master/tables/CMIP6_CV.json)

parent_experiment_id
ssp245

parent_experiment_rip
r1i1p1

parent_time_units
'days since 1800-01-01'

branch_method
'Hybrid-restart from year 2015-01-01 of historical',

other information
(provide other information that might be useful)

Path to parent experiment: /projects/NS9560K/noresm/cases/NHIST_f19_tn14_20190710
Experiment_rip: r1i1p1f2
Do not publish first in ESGF!!

[YanchunHe]

Hi Jerry,

Please fill with correct information for:

• CASENAME(s) and years to be CMORized
• parent_experiment_rip
• parent_time_units
• branch_method
  and also:
  path to the parent experiment (original case name and path of ssp245) if available in the 'other information" part.
  thanks!
  Yanchun

[tjiputra]

Hi Yanchun,

Please check if everything is correct now. Also, I noticed for historical NorESM2-LM cmorization, in the issue requests of
ensemble #1 (#133)
ensemble #3 (#69)
both listed the identical casename. I hope this is just a typo.

many thanks,
jerry

[YanchunHe]

Hi,

I noticed the parent_experiment_id for ssp245-covid is ssp245, as described in "https://github.com/PCMDI/cmip6-cmor-tables/blob/master/Tables/CMIP6_CV.json". Not 'historical'

Also I need to know where you branched off from the parent experiment in "branch_method". Please update that.

For the NorESM2-LM historical ensemble 1, it is indeed a typo. The cmorized data should be correct. I will correct this. Thanks a lot for pointing out!

Yanchun

[tjiputra]

It's was a bit confusing for me as we diced to do the run branching from historical and nit from the ssp245 itself.

Many thanks, Yanchun.

[YanchunHe]

It's was a bit confusing for me as we diced to do the run branching from historical and nit from the ssp245 itself.

Many thanks, Yanchun.

OK, thanks!

Then I will still set the parent expid as 'ssp245' (the cmor library will anyway force use ssp245 even if I specify historical').

and the 'branch method' as: "'Hybrid-restart from year 2015-01-01 of ssp245'.

2015-01-01 of historical and ssp245 should be identical or quite close.

Anyway, it should be fine.

[YanchunHe]

Encounter an error during cmorization:

! Error: Could not find a matching variable for name: 'b_half_bnds'

likely due to missing such a property for the "PS" variable in the monthly aerosol atm/hist output.

Can you quickly locate where is the problem and probably with a fix?

@tjiputra

[DirkOlivie]

@YanchunHe @tjiputra

I have not found the reason for the problem yet. I have checked the sizes of the cam.h0 and cam.h1 files, and they all seem correct (none is much smaller - so no indication of possibly missing data).
In the cam.h0 and cam.h1 files, there are the variables hyai, hybi, hyam and hybm. I think the hybi correspond with b_half_bnds, and this array is present in the cam.h0 and cam.h1 files which I have checked. Also the variables PS and P0 are in the file.

Some of my postprocessing on nird went very slow yesterday evening and this morning. Might there have been an issue with file access and reading on nird, which also impacted the cmorization?

[YanchunHe]

Thanks @DirkOlivie ! I will try to dig more information on this error.

[YanchunHe]

Hi Ingo, @IngoBethke

In order to process this newly added ssp245-covid experiment in the DAMIP, I have to upgrade the CMIP6 Tables to the latest version, with CMOR library 3.6.0.

However, the currently install CMOR library is 3.5.0, which has inconsistency with the updated tables. That leads to the issue I mentioned above.

As you initially installed the CMOR library, could upgrade the library to 3.6.0 or install a fresh new version.

Thanks in advance!

Yanchun

[tjiputra]

Hi Yanchun, @YanchunHe

I just add a minor edit to name the realization of this experiment as "r1i1p1f2". There is a request to have this plus the other 7 members by early next week. So I will monitor this closely and submit the submit the other issues once you give a green light.

many thanks,
jerry

[YanchunHe]

Hi @tjiputra

Please go head to submit cmor requests for the other members.

I will skip those few variables defined on 'alevhalf', e.g., 'phalf', and process the others.

[YanchunHe]

This issue is fixed by omitting cell_bounds in cmor_axis() and zfactor_bounds in cmor_zfactor(). see the issue reported: PCMDI/cmor#617

updated cmor 3.6.0, cmor tables, and Makefile in commit: 9a9b025
updated recipe templates for forcing and initialization index in commit: d53e387

This issue #248 is part of #250
so close this issue now.