search

E3SM-Project / E3SM

Energy Exascale Earth System Model source code. NOTE: use "maint" branches for your work. Head of master is not validated.
https://docs.e3sm.org/E3SM
Other
332 stars 334 forks source link

Update v3 SSP compsets/configurations/tests and use_case files #6335

Closed wlin7 closed 1 week ago

wlin7 commented 1 month ago

Add/update SSP compsets with standard v3 configurations. The changes include:

[non-BFB] only for SSP tests [NML] changes only for SSP tests

Checklist and Notes

github-actions[bot] commented 1 month ago

PR Preview Action v1.4.7 :---: :rocket: Deployed preview to https://E3SM-Project.github.io/E3SM/pr-preview/pr-6335/ on branch gh-pages at 2024-05-03 14:11 UTC

tangq commented 1 month ago

Hi @wlin7 , regarding your E90 checklist and notes below,

I assume the file created in 2017 refers to this file /lcrc/group/e3sm/data/inputdata/atm/cam/chem/emis/CMIP6_emissions_1750_2015/emissions_E90global_surface_1750-2100_0.9x1.25_c20170322.nc, which was before E90 was implemented for E3SMv3. We certainly do NOT want to use this file in E3SM.

For E3SMv3, we should use this E90 file for ALL SSP scenarios because E90 is an artificial tracer which doesn't change with SSPs. /lcrc/group/acme/public_html/inputdata/atm/cam/chem/trop_mozart/ub/emissions_E90_surface_1750-2101_1.9x2.5_c20231222.nc, which was created by Jinbo a couple months ago. (@jinboxie can provide details if needed.)

Checklist and Notes

E90 emission for SSP scenarios need to be inspected. The one specified was created in 2017 and uses variable emis_e90, while emis_anthro is used for piControl/historical and F compsets

wlin7 commented 1 month ago

Thank you @tangq and @jinboxie . It is great you have already created one for all v3 SSP scenarios.

wlin7 commented 1 month ago

Hi @hwangacme @mingxuanwupnnl @tangq @keziming @jinboxie , @keziming @bishtgautam This SSP PRs involve many files you helped create. Please help review if the settings are all good.

You may start with the checklist and notes in the PR message. Available historical and SSP tests may be helpful for comparing the settings.

wlin7 commented 1 month ago

Hi @tangq , the settings for the prescribed methane section has been all removed from the SSP use_case files via 9a0ab28 . Note that similar settings still exist in other use_case _files (e.g., 1850, 2010, 20TR). Related settings still show up in atm_in. Complete removal requires removal from namelist_defaults_eam.xml and changes of logics in build-namelist (and probably configure). This is not done.

rljacob commented 2 weeks ago

@wlin7 is this ready for merging?

wlin7 commented 2 weeks ago

@rljacob , the land settings need further review/input. @acme-y9s and @thorntonpe , can you please take a look? It is about the use_case settings to work with the BGC land.

acme-y9s commented 2 weeks ago

@wlin7 @bishtgautam For the nitrogen deposition (nden) and population density (popdens) files, we do have the transient files that we used for BGCv1 simulation (like SSP585). For the phosphorus deposition file (pden) we only one year. If you do not specify the those files in your use_case file, the model will have default setting files (maybe based on you simulation year ), but the default files may not what you need. There should be ndep and popdens files for each SSPs, you can specify thoses files in each SSP's use_case file. I did specify those files for our BGCv1 SSP585 case, you can have a look if you want. Or, if you need help, feel free to let me know.

wlin7 commented 2 weeks ago

@wlin7 @bishtgautam For the nitrogen deposition (nden) and population density (popdens) files, we do have the transient files that we used for BGCv1 simulation (like SSP585). For the phosphorus deposition file (pden) we only one year. If you do not specify the those files in your use_case file, the model will have default setting files (maybe based on you simulation year ), but the default files may not what you need. There should be ndep and popdens files for each SSPs, you can specify thoses files in each SSP's use_case file. I did specify those files for our BGCv1 SSP585 case, you can have a look if you want. Or, if you need help, feel free to let me know.

Thanks, @acme-y9s definitely need more help from you on this. Can you please provide the files (the path if ready on our machine) and the proper year-align settings? We can update the specification in a next PR.

wlin7 commented 2 weeks ago

Merged to next.

wlin7 commented 1 week ago

Remerged to next after restoring the config settings for 4xCO2 and 1pctCO2 compset. Pattern _EAM%CMIP6_ can not match the CO2 compset longname, which uses - in place of _ following the string CMIP6.

wlin7 commented 1 week ago

Only expected SSP test diffs with this PR on next (and compy's SSP370 FAIL due to permission to download additional forcing files).

The PR is merged to master (to create a master repo with SSP compsets). Requests for bless SSP test DIFFs are submitted but before they are blessed, both next and master will see diff on Monday.

peterdschwartz commented 3 days ago

@wlin7 Maybe due to machine issues at the time, but this PR may still have DIFFs on pm-cpu that need to be blessed(for example https://my.cdash.org/viewTest.php?onlyfailed&buildid=2561805).