Currently, the model output of potential temperature is compared to climatology of in-situ temperature from WOA13.

Need to use potential temperature (referred to ocean surface) calculated from t_an and s_an and depth from WOA13.

To enable run the tool on the ipcc.nird.sigma2.no node.

grid file:
Betzy: /cluster/shared/noresm/inputdata/ocn/blom/grid/grid_tnx0.125v4_20200722.nc

This issue is to include another diagnostic component for the land ice.

The currently implementation is by default plot all the time series of model run (TRENDS_ALL=1).
If two runs are compared, the time-axis of the second case (-i2) is abandoned but use the time-axis information from the first case (-i1) instead.
(This was designed as it was, but reported by some users as bugs).

To do:

• numerous fixes needed due to NIRD upgrade/maintenance
• port to Betzy.
• miscellaneous improvements

Would it be possible to add to the HAMOCC diagnostics a comparison of the model output ('bigdelta14clvl') with observed Δ14C?

The file with the gridded product of pre-industrial Δ14C data (from Key et al. (2004)) is available here in NetCDF format: http://vedlegg.uib.no/?id=0f9edf8d3e3ab53645b5d3561535f37d

The model output will only be there if the model is run with the isotopes enabled (like for d13C).

I put an example file with such output on NIRD: /projects/NS2345K/temp/NOINYOC_tn21_18b.micom.hbgcy.3970-4000.nc

(A. Moree)

Comparing multiple NorESM experiments/versions without cmorized outputs.
(in connection to KeyClim WP6 and iNES wp4).

Currently, it is in the job script running in the login node after the submitted job of diag_run in diag_srun batch job is finished

(A. Moree)

I noticed some changes in the output of the diag package. Before, I would successfully generate section transports and global averages, but this does not work at the moment:

For example compare

http://ns2345k.web.sigma2.no/noresm_diagnostics/NOINYOC_tn21_20A/MICOM_DIAG/yrs3950to4000-obs/indexnew.html

to

http://ns2345k.web.sigma2.no/noresm_diagnostics/NOINYOC_tn21_20A/MICOM_DIAG/yrs2950to3000-obs/indexnew.html

This NOINYOC_tn21_20A run has a different grid, causing some of the information to already be missing in the 2950-3000 diag (which is OK), but in the 3950-4000 diag I do not get any section transport anymore.

The logs are here: /projects/NS2345K/noresm_diagnostics/out/anne89/MICOM_DIAG/config/NOINYOC_tn21_20A/logs

It says several variables are not present, but that would be surprising because I did not change the output variables during the run. The output files are here: /projects/NS2980K/NORSTORE_OSL_DISK/NS2980K/moree/NOINYOC_tn21_20A/ocn/hist

The CICE and HAMOCC output is the same as before as far as I see now.

Do you see what happened here?

To private home folders on NIRD or other machine.

For example:

/projects/NS2345K/noresm_diagnostics/bin/diag_run -m blom,cice -c1 NOIIAJRAOC20TR_TL319_tn14_20190710_repro -s1 1653 -e1 1653 -i1 /projects/NS2345K/cgu025/cases -c2 NOIIAJRAOC20TR_TL319_tn14_20190710 -s2 1653 -e2 1653 -o /projects/NS2345K/cgu025/noresm_diagnostics/

Then diag_run gives warnings like -

WARNING: end_year in NOIIAJRAOC20TR_TL319_tn14_20190710_repro is smaller than start_year: please correct this in blom_diag_template.sh.
-CHANGING CNTL CASE to NOIIAJRAOC20TR_TL319_tn14_20190710 in blom_diag_template.sh
-CHANGING CNTL to USER in blom_diag_template.sh
-CHANGING FIRST CLIMO YR IN CNTL CASE to 1653 in blom_diag_template.sh
WARNING: end_year of NOIIAJRAOC20TR_TL319_tn14_20190710 is smaller than start_year: please correct this in blom_diag_template.sh.
-SETTING UP TIME-SERIES DIAGNOSTICS FOR ENTIRE EXPERIMENT

At config.log it says -

At the very least, the following variables in blom_diag_template.sh must be configured prior to running the script:

NYRS_CLIMO1
NYRS_CLIMO2
(only if CNTL==USER)
TRENDS_ALL

Can compare the experiments by BLOM and previous experiment by MICOM

I think there are some more recent versions of the NCAR diagnostics including new observations.

Thomas have you used / seen the path to any of those?

If these versions are readily available it would be nice if you can find the time to update our diag_run version.
(ø.Seland)