Failure in trying to reproduce a SSP585-ESM simulation with CESM2 done at Cheyenne

[m_kliphuis@uu_nl]

Hi CESM experts,

I work at the Institute for Marine and Atmospheric research Utrecht (IMAU) in the Netherlands and I would like to rerun the following CESM2 case (b.e21.BSSP585_BPRPcmip6.f09_g17.CMIP6-esm-ssp585.001) which was run at machine Cheyenne:

https://data.ucar.edu/dataset/cmip6...nteractive-land-clm5-coupled-ocean-pop2-with-

I tried to reproduce the run on the dutch national supercomputer 'Snellius' and I think I set it up properly with the command:

./create_newcase --run-unsupported --case b.e21.BSSP585cmip6.f09_g17.02 --compset SSP585_CAM60_CLM50%BGC-CROP-CMIP6DECK_CICE%CMIP6_POP2%ECO%ABIO-DIC_MOSART_CISM2%NOEVOLVE_WW3_BGC%BPR --res f09_g17 --output-root /projects/0/prace_imau/prace_2013081679/cesm2_1_3/

Note the option BPRP for the biochemistry model in the ocean, meaning that CO2 in the ocean is calculated instead of prescribed. Also note that I’m using CESM2.1.3.

On this forum at How does one set up a SSP585-ESM simulation with CESM2
I found (at the end) Keith Lindsay mentioning that the initial condition files should not be the default initial files coming from a BDRD experiment but that I should use BPRP initial files.

I contacted Gary Strand and thankfully he (very quickly) offered me the files, more specific the year 2015 restart files from case: BHIST_BPRP.f09_g17.CMIP6-esm-hist.002.

However after running the case for 25 years the run does not give similar output as the NCAR run mentioned above. The NCAR run is much warmer, especially on the north and south pole ice sheets as you can see in the attached figure, showing the difference in 'Surface Temperature' averaged over the first year resp. the first 25 years.

I thought that perhaps I should have used the default year 2015 restart files (from case: b.e21.BHIST.f09_g17.CMIP6-historical.010_v2) but also then I get similar results.

Apparantly I am missing something. I have 2 questions:

1. Was the NCAR run indeed initialized with the b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.002 initial files?
2. Are there other settings that I forgot?

This is how I setup my case:

./create_newcase --run-unsupported --case b.e21.BSSP585_BPRPcmip6.f09_g17.001 --compset SSP585_CAM60_CLM50%BGC-CROP-CMIP6DECK_CICE%CMIP6_POP2%ECO%ABIO-DIC_MOSART_CISM2%NOEVOLVE_WW3_BGC%BPR --res f09_g17 --output-root /projects/0/prace_imau/prace_2013081679/cesm2_1_3/

with in env_run.xml

<entry id="RUN_REFDIR" value="cesm2_init">
<entry id="RUN_REFCASE" value="b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.002">
<entry id="RUN_REFDATE" value="2015-01-01">
<entry id="RUN_REFTOD" value="00000">
<entry id="GET_REFCASE" value="TRUE">
<entry id="RUN_STARTDATE" value="2015-01-01">

After that I simply let it run for 25 years.

Thanks in advance …

best,

Michael

 

[mlevy]

Was the NCAR run indeed initialized with the b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.002 initial files?

I think I found the original case root for this run on the NCAR system in /glade/work/cmip6/cases/C4MIP/b.e21.BSSP585_BPRPcmip6.f09_g17.CMIP6-esm-ssp585.001, and it does look like that is the correct case reference case. All of the values you listed from env_run.xml look right to me:

$ grep entry.*REF env_run.xml
    <entry id="RUN_REFDIR" value="cesm2_init">
    <entry id="RUN_REFCASE" value="b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.002">
    <entry id="RUN_REFDATE" value="2015-01-01">
    <entry id="RUN_REFTOD" value="00000">
    <entry id="GET_REFCASE" value="TRUE">
$ grep entry.*RUN_STARTDATE env_run.xml
    <entry id="RUN_STARTDATE" value="2015-01-01">

Are there other settings that I forgot?

I'm not sure of the answer to this question, so I'm going to move this CESM Community Project forum

 

[strandwg]

I can verify that the ESM SSP585 run

b.e21.BSSP585_BPRPcmip6.f09_g17.CMIP6-esm-ssp585.001

was indeed started from

b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.002

at year 2015.

I've also attached the CASEROOT for b.e21.BSSP585_BPRPcmip6.f09_g17.CMIP6-esm-ssp585.001

 

[m_kliphuis@uu_nl]

Thanks a lot Michael (Levy) for moving my ticket and bringing it back under attention.

And Gary, to you many thanks for sharing the CASEROOT. This was very helpful!

I compared it with my own CASEROOT and in the end the problem was caused by a simple typo :(

I set up my case using this compset:

SSP585_CAM60_CLM50%BGC-CROP-CMIP6DECK_CICE%CMIP6_POP2%ECO%ABIO-DIC_MOSART_CISM2%NOEVOLVE_WW3_BGC%BPR

which I copied from the 4th message (of Erik Kluzek) on the forum page:

How does one set up a SSP585-ESM simulation with CESM2

When I did this I already though; why are the last characters not BPRP instead of BPR but I figured perhaps this was done because the complete string would become too long or something like that.

In the message of Erik a missing 'P' should be added after 'BPR', I will sent Erik a message about this.

Adding the 'P' and running for 5 years indeed gave similar results :)

Thanks again!

All the best,
Michael

 

[strandwg]

Great!

 

[Besuyi]

I can verify that the ESM SSP585 run

b.e21.BSSP585_BPRPcmip6.f09_g17.CMIP6-esm-ssp585.001

was indeed started from

b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.002

at year 2015.

I've also attached the CASEROOT for b.e21.BSSP585_BPRPcmip6.f09_g17.CMIP6-esm-ssp585.001

Hi, strandwg
I want to study the detalied settings of BHIST_BPRP-esm-hist, could you please share me the CASEROOT files for "b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.001"? It showed Not Fould while I opened the Link.
Thank you in advance!

 

[strandwg]

Hi, strandwg
I want to study the detalied settings of BHIST_BPRP-esm-hist, could you please share me the CASEROOT files for "b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.001"? It showed Not Fould while I opened the Link.
Thank you in advance!

Please retrieve the file

ftp://ftp.cgd.ucar.edu/pub/strandwg/b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.001.tar.gz

 

[Besuyi]

Please retrieve the file

ftp://ftp.cgd.ucar.edu/pub/strandwg/b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.001.tar.gz

Thank you,Sir
I appreciate with your help and I have downloaded the file.

 

[Lil]

Please retrieve the file

ftp://ftp.cgd.ucar.edu/pub/strandwg/b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.001.tar.gz

Dear strandwg,

Hope this message finds you well. I have been following this and others' threads on the b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.001 restart files. I apologize for any inconvenience, but it seems that I am unable to locate the files on the link you previously shared also in CESM2.1 BHIST restart files.

If it’s not too much trouble, I kindly request your assistance in obtaining the 2015-01-01 restart files. Your help with this matter would be greatly appreciated.

Thank you very much for your time and attention :)
Lili

 

[strandwg]

Dear strandwg,

Hope this message finds you well. I have been following this and others' threads on the b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.001 restart files. I apologize for any inconvenience, but it seems that I am unable to locate the files on the link you previously shared also in CESM2.1 BHIST restart files.

If it’s not too much trouble, I kindly request your assistance in obtaining the 2015-01-01 restart files. Your help with this matter would be greatly appreciated.

Thank you very much for your time and attention :)
Lili

Please retrieive

ftp://ftp.cgd.ucar.edu/b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.001.rest.2015-01-01-00000.tar

 

[Lil]

Please retrieive

ftp://ftp.cgd.ucar.edu/b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.001.rest.2015-01-01-00000.tar

Thank you so much for your prompt assistence. I have successfully downloaded the data : )

Best regards,
Lili

 

[kyha0830]

Please retrieive

ftp://ftp.cgd.ucar.edu/b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.001.rest.2015-01-01-00000.tar

Dear strandwg,

I have been struggling to find the restart files for a BPRP future projection (from 2015) and arrived here. By setting REFCASE to b.e21.BHIST_BPRP.f09_g17.CMIP6-esm-hist.002
following the above comments, the `case.build` failed due to the absence of the case at https://svn-ccsm-inputdata.cgd.ucar.edu/trunk/inputdata/cesm2_init/
Could you please provide the restart files for me as well ?
# Just in case it matters, I am using CESM 2.1.5 (the latest scientifically valid version) not 2.1.3. And I do not specifically demand that above-mentioned case (...-hist.002) if there exists another restart data made from a BHIST_BPRP.f09_g17 run that is usable with v2.1.5.

Best regards,
Young-Ha