Update ELM in the definition of multiple V3 compsets #6108

bishtgautam · 2023-12-07T05:24:07Z

The following compsets are updated to change ELM's mode from SP to active BGC mode with TOP solar parameterization:

F1850
F20TR
F2010
WCYCL1850_chemUCI-Linozv3
WCYCL1850_chemUCI-Linozv3-mam5
WCYCL20TR_chemUCI-Linozv3-mam5
WCYCL1850-4xCO2
WCYCL1850
WCYCL1850-1pctCO2
WCYCL20TR
WCYCL1850NS

[CC]
[non-BFB] but v3 sims already using these by runscript.
[NML]

Defines a new long compset for ELM that includes BGC and TOP solar parameterization.

The following compsets is updated to change ELM's SP mode to active BGC with TOP solar parameterization: - F1850 - F20TR - F2010 - WCYCL1850_chemUCI-Linozv3 - WCYCL1850_chemUCI-Linozv3-mam5 - WCYCL20TR_chemUCI-Linozv3-mam5 - WCYCL1850-4xCO2 - WCYCL1850 - WCYCL1850-1pctCO2 - WCYCL20TR

github-actions · 2023-12-07T05:25:22Z

PR Preview Action v1.4.6
🚀 Deployed preview to https://E3SM-Project.github.io/E3SM/pr-preview/pr-6108/
on branch `gh-pages` at 2024-01-24 19:40 UTC

bishtgautam · 2023-12-07T05:33:40Z

@rljacob @wlin7 This PR will cause DIFFs for multiple tests.

rljacob · 2023-12-07T16:56:10Z

Aren't these climate changing (if you were running master with SPBC ) ?

rljacob · 2023-12-13T20:31:41Z

Make sure user e3sm_integration runs with these changes BEFORE merging to next.

wlin7 · 2024-01-04T17:39:38Z

@bishtgautam , the config changes introduced in this PR would become incompatible with master for some compsets (e.g., F2010). This would cause many tests to fail during setup stage, if merging the PR to next, or rebasing it with master.

A test like REP_Ln5.ne4_oQU240.F2010.pm-cpu_intel would complain
ERROR : CLM build-namelist::ELMBuildNamelist::add_default() : No default value found for stream_year_first_fan

Look like certain recent land PRs in master have not accounted for F2010 with the intended ELM config opts.

Note that the problem does not exist with this PR branch itself, which was based off an older master.

rljacob · 2024-01-04T17:48:54Z

@bbye see above. If FAN is not on, then the buildnamelist should not even try to set stream_year_first_fan should it?

wlin7 · 2024-01-04T18:08:43Z

@bbye see above. If FAN is not on, then the buildnamelist should not even try to set stream_year_first_fan should it?

In ELMBuildNamelist.pm, add_default indeed is inside if( $opts->{'fan'} ). So it is interesting why it still tried to add. In another standalone F2010-NARRM test, adding use_fan=.false. in user_nl_elm didn't help.

bbye · 2024-01-05T21:26:53Z

@bbye see above. If FAN is not on, then the buildnamelist should not even try to set stream_year_first_fan should it?

In ELMBuildNamelist.pm, add_default indeed is inside if( $opts->{'fan'} ). So it is interesting why it still tried to add. In another standalone F2010-NARRM test, adding use_fan=.false. in user_nl_elm didn't help.

It shouldn't set the fan stream if fan isn't used, and this worked with the original PR, but I think the problem is in the namelist_default.xml file where the fan_stream is set in lines 730-766. I think changing the use_cn condition to a use_fan condition will resolve this issue. I have a commit to test this in my branch, but I don't know how to test it on the failing cases.

wlin7 · 2024-01-06T03:03:08Z

It shouldn't set the fan stream if fan isn't used, and this worked with the original PR, but I think the problem is in the namelist_default.xml file where the fan_stream is set in lines 730-766. I think changing the use_cn condition to a use_fan condition will resolve this issue. I have a commit to test this in my branch, but I don't know how to test it on the failing cases.

@bbye , did you mean the test mentioned above? It can be created using the command as follows if doing it on pm-cpu.

$E3SMROOT/cime/scripts/create_test REP_Ln5.ne4_oQU240.F2010.pm-cpu_intel --machine=pm-cpu

bishtgautam · 2024-01-06T04:29:04Z

Using @bbye's suggestion, I have opened #6142 which fixes the issue. After merging this branch and #6142 to next, the ELM setup proceeds without an error.

rljacob · 2024-01-11T04:51:56Z

This should be merged along with #6142

The following compsets are updated to change ELM mode from SP to active BGC mode with TOP solar parameterization: F1850 F20TR F2010 WCYCL1850_chemUCI-Linozv3 WCYCL1850_chemUCI-Linozv3-mam5 WCYCL20TR_chemUCI-Linozv3-mam5 WCYCL1850-4xCO2 WCYCL1850 WCYCL1850-1pctCO2 WCYCL20TR [CC] [non-BFB] but v3 critical path is already using these. [NML]

wlin7 · 2024-01-12T17:55:30Z

Verified master + #6108 and #6113 can BFB reproduce the current v3.LR.piControl-spinup after removing all non-IO related configuration settings in run script.

Notes:

use of nlmap for atm to ice, ocn, lnd and rof (fmap, e.g., atm2ice_fmapname_nonlinear ) still need to be specified in run script. Otherwise the field is "idmap_ignore'
fsurdat still needs to be specified, otherwise it would pick up surfdata_0.5x0.5_simyr1850_c211019.nc, which does not include '_TOP' and the run would fail. This is going to be fixed in a new PR.

rljacob · 2024-01-12T18:27:29Z

These compsets aren't changed. Should they be?
WCYCLSSP370
WCYCLSSP585
WCYCL1850NS

rljacob · 2024-01-12T18:48:39Z

There are runtime errors from ELM in the prod suite fully coupled cases. Probably because they are all using wrong finidat files?

finidat = '$DIN_LOC_ROOT/lnd/clm2/initdata_map/clmi.WCYCL1850-1pctCO2.ne30pg2_EC30to60E2r2.SMS_Ld1.c20230213.nc
 finidat = '$DIN_LOC_ROOT/lnd/clm2/initdata_map/clmi.WCYCL1850-4xCO2.ne30pg2_EC30to60E2r2.SMS_Ld1.c20230213.nc
 finidat = '$DIN_LOC_ROOT/lnd/clm2/initdata_map/clmi.WCYCL1850.ne30pg2_EC30to60E2r2.SMS_Ld1.c20230213.nc'
 finidat = '${DIN_LOC_ROOT}/lnd/clm2/initdata_map/clmi.WCYCL1850.ne30pg2_r05_EC30to60E2r2.SMS_Ld1.c20230213.nc'
 finidat = '${DIN_LOC_ROOT}/lnd/clm2/initdata_map/clmi.WCYCL1850.northamericax4v1pg2_WC14to60E2r3.SMS_PS.c20230213.nc'

ENDRUN:
ERROR:: litr1c is required on an initialization datasetERROR in ColumnDataType.
F90 at line 2570

rljacob · 2024-01-12T19:02:46Z

There are also runtime errors in the prod suite F-cases. finidat is commented out in those.
ne30pg2_r05_oECv3.F20TR.mach_comp.eam-wcprod_F20T
ne30pg2_r05_oECv3.F2010.mach_comp.eam-wcprod_F2010

 20:  surfrd_veg_all ERROR: sum of wt_nat_patch not 1.0 at nl=        4550  and t=
 20:            1
 20:  sum is:   5.088610182091506E+018
 20:  ENDRUN:
 20:  ERROR in surfrdUtilsMod.F90 at line 75

wlin7 · 2024-01-12T20:20:43Z

@rljacob , It appears so - that finidat files are incompatible with the land in BGC mode. I am going to test a fix, either using null finidat, or pulling one from v3.LR. piControl-spinup.

Do you plan to proceed with merge this and #6142 to master today?

rljacob · 2024-01-12T20:27:44Z

No I want to test the fixes. And if we can't fix the F-case fails, I'll revert this PR but will merge #6142

rljacob · 2024-01-12T22:49:45Z

Since we're headed in to a 3-day weekend, I'm going to force-push this off next and we can try again Tuesday.

rljacob · 2024-01-13T22:31:30Z

This was reverted from next.

wlin7 · 2024-01-18T18:02:25Z

The error like surfrd_veg_all ERROR: sum of wt_nat_patch not 1.0 at nl= 4550 and t= ... as seen in F-cases would also appear in coupled cases after resetting the incompatible finitdat (emptying it). There must be a systematic issue. Working on this, with an option to switch to v3 LR production grid for the wcprod tests.

wlin7 · 2024-01-19T19:39:54Z

The new commits 0549495, 8fe3157, ede5796 fix the issues for the wcprod test suite and other F-cases.

The wcprod test suite changes to use new v3.LR standard grid, ne30pg2_r05_IcoswISC30E3r5. wcprod test in atm_integration suite still uses bigrid ne30pg2_EC30to60E2r2. Ok for now to use the same testmod when not specifying an finidat. We can decide if to continue testing both bigrid and tri-grid for wcprod.

surfrd_veg_all ERROR: sum of wt_nat_patch not 1.0 at nl= 4550 and t= ... is fixed after resetting create_crop_landunit = .false. for v3 wcycl config.

@bishtgautam , the default setting for r05 simyr1850 fsurdat is set using the idea you proposed by adding additional attribute use_top_solar_rad = .true., as follows in namelist_defaults.xml, and similarly for create_crop_landunit to distinguish from the other entry that also has use_cn=.true.. It has to take the order as is otherwise it wouldn't work for v3 wcycl land config. We could alternatively specify them in the use_case files, and need to do so for each simulation period.

<fsurdat hgrid="r05"          sim_year="1850" use_crop=".false." use_cn=".false.">
lnd/clm2/surfdata_map/surfdata_0.5x0.5_simyr1850_c200609_with_TOP.nc</fsurdat>
<fsurdat hgrid="r05"          sim_year="1850" use_crop=".false." use_cn=".true." use_top_solar_rad=".true.">
lnd/clm2/surfdata_map/surfdata_0.5x0.5_simyr1850_c200609_with_TOP.nc</fsurdat>
<fsurdat hgrid="r05"          sim_year="1850" use_crop=".false." use_cn=".true.">
lnd/clm2/surfdata_map/surfdata_0.5x0.5_simyr1850_c211019.nc</fsurdat>


<create_crop_landunit use_crop=".false."                 >.false.</create_crop_landunit>
<create_crop_landunit use_crop=".false." hgrid="r05" use_cn=".true." use_top_solar_rad=".true.">.false.</create_crop_landunit>
<create_crop_landunit use_crop=".false." hgrid="r05" use_cn=".true.">.true.</create_crop_landunit>

bishtgautam · 2024-01-19T21:19:52Z

I compared the lnd_in files for all the tests in e3sm_land_developer test suite that have an active BGC and there are no differences in the baseline w.r.t. this branch. Thank you, Wuyin for those three commits to fix the test suite.

@rljacob I believe we are ready to re-merge this PR.

bishtgautam · 2024-01-19T21:20:52Z

Forgot one thing: the WCYCL1850-WW3 compset isn't updated, yet.

rljacob · 2024-01-19T21:56:47Z

Yes @sbrus89 agreed we should update that one. We can merge this early next week.

bishtgautam · 2024-01-22T19:27:07Z

@rljacob Definition of WCYCL1850-WW3 has been updated.

The following compsets are updated to change ELM's mode from SP to active BGC mode with TOP solar parameterization: F1850 F20TR F2010 WCYCL1850_chemUCI-Linozv3 WCYCL1850_chemUCI-Linozv3-mam5 WCYCL20TR_chemUCI-Linozv3-mam5 WCYCL1850-4xCO2 WCYCL1850 WCYCL1850-1pctCO2 WCYCL20TR WCYCL1850NS [CC] [non-BFB] but v3 sims already using these by runscript. [NML]

rljacob · 2024-01-23T23:47:55Z

merged to next

rljacob · 2024-01-24T18:03:26Z

There are 4 runtime fails on integration. All are BGC cases and all have the same error:
ERROR: New format surface datasets require create_crop_landunit TRUE
ERROR in surfrdMod.F90 at line 1215

The cases and compset longname.

SMS_Ln5.ne30pg2_r05_EC30to60E2r2.BGCEXP_LNDATM_CNPRDCTC_20TR
- 20TR_EAM%CMIP6_ELM%CNPRDCTCBC_MPASSI%PRES_DOCN%DOM_MOSART_SGLC_SWAV_BGC%LNDATM
SMS_Ln5.ne30pg2_r05_EC30to60E2r2.BGCEXP_LNDATM_CNPRDCTC_1850
- 1850_EAM%CMIP6_ELM%CNPRDCTCBC_MPASSI%PRES_DOCN%DOM_MOSART_SGLC_SWAV_BGC%LNDATM
SMS_Ld2.ne30pg2_r05_EC30to60E2r2.BGCEXP_CNTL_CNPECACNT_1850.xx_xx.elm-bgcexp
- 1850_EAM%CMIP6_ELM%CNPECACNTBC_MPASSI%BGC_MPASO%OIECO_MOSART_SGLC_SWAV_BGC%BCRC
SMS_Ld2.ne30pg2_r05_EC30to60E2r2.BGCEXP_CNTL_CNPRDCTC_1850.xx_xx.elm-bgcexp
- 1850_EAM%CMIP6_ELM%CNPRDCTCBC_MPASSI%BGC_MPASO%OIECO_MOSART_SGLC_SWAV_BGC%BCRC

The value of `use_top_solar_rad` is also included in the logic to create crop landunits.

Second merge of this branch. Fix 4 failing tests.

…(PR #6625) Updating CRYO compsets to v3 ELM settings Updates the ELM settings in the following compsets to the v3 settings in PR #6108: * CRYO1850 * CRYO1850-DISMF * CRYO1850-4xCO2 * CRYO1950 * CRYO1950-DISMF * CRYO20TR * CRYOSSP370 * CRYOSSP585 Adds following compsets matching the WCYCL counterparts: * CRYO1850-1pctCO2 * CRYOSSP245 [non-BFB] only for CRYO* compsets

Updating CRYO compsets to v3 ELM settings Updates the ELM settings in the following compsets to the v3 settings in PR #6108: * CRYO1850 * CRYO1850-DISMF * CRYO1850-4xCO2 * CRYO1950 * CRYO1950-DISMF * CRYO20TR * CRYOSSP370 * CRYOSSP585 Adds following compsets matching the WCYCL counterparts: * CRYO1850-1pctCO2 * CRYOSSP245 [NML] only for CRYO* compsets [non-BFB] only for CRYO* compsets

bishtgautam added 2 commits December 6, 2023 14:01

Adds a new ELM compset definition for V3

1cd5fac

Defines a new long compset for ELM that includes BGC and TOP solar parameterization.

bishtgautam requested review from acme-y9s, wlin7 and rljacob December 7, 2023 05:31

bishtgautam changed the title ~~[WIP] Updates the definition of multiple V3 compsets~~ Updates the definition of multiple V3 compsets Dec 7, 2023

rljacob changed the title ~~Updates the definition of multiple V3 compsets~~ Update ELM in the definition of multiple V3 compsets Dec 7, 2023

rljacob added Land compset BGC non-BFB PR makes roundoff changes to answers. labels Dec 7, 2023

rljacob added this to the v3.0.0-rc milestone Dec 7, 2023

wlin7 mentioned this pull request Dec 13, 2023

Enable rough topograph and add remaining atm and lnd settings for v3b02 #6113

Merged

rljacob self-assigned this Dec 14, 2023

rljacob added the CC PR is climate changing label Dec 18, 2023

rljacob approved these changes Jan 11, 2024

View reviewed changes

Updates the definition of WCYCL1850NS

7835327

wlin7 added 3 commits January 19, 2024 11:01

Unset elm's finidat for wcprod test suite

0549495

Update default settings for v3 wcycl land config

8fe3157

Change test grid for wcprod suite except for SSP & NARRM

ede5796

Updates the definition of WCYCL1850-WW3

43b642d

Updates the logic to create crop landunits

dc6c743

The value of `use_top_solar_rad` is also included in the logic to create crop landunits.

rljacob added a commit that referenced this pull request Jan 25, 2024

Merge branch 'bishtgautam/lnd/v3-compset' into next (PR #6108)

4d6f1e5

Second merge of this branch. Fix 4 failing tests.

rljacob merged commit fd65dbe into master Jan 25, 2024
3 checks passed

rljacob deleted the bishtgautam/lnd/v3-compset branch January 25, 2024 22:38

rljacob mentioned this pull request Jan 28, 2024

SMS.r05_r05.I1850ELMCN.* * .elm-qian_1948 failing #6192

Closed

darincomeau mentioned this pull request Feb 26, 2024

SMS_Lm1.ne30pg2_ECwISC30to60E2r1.WCYCL1850 broken #6230

Closed

rljacob mentioned this pull request Jun 12, 2024

Update TOP solar radiation parameterization #6467

Merged

darincomeau mentioned this pull request Sep 19, 2024

Updating CRYO compsets to v3 ELM settings #6625

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Update ELM in the definition of multiple V3 compsets #6108

Update ELM in the definition of multiple V3 compsets #6108

bishtgautam commented Dec 7, 2023 •

edited by rljacob

Loading

github-actions bot commented Dec 7, 2023 •

edited

Loading

bishtgautam commented Dec 7, 2023

rljacob commented Dec 7, 2023

rljacob commented Dec 13, 2023 •

edited

Loading

wlin7 commented Jan 4, 2024

rljacob commented Jan 4, 2024

wlin7 commented Jan 4, 2024

bbye commented Jan 5, 2024

wlin7 commented Jan 6, 2024

bishtgautam commented Jan 6, 2024 •

edited

Loading

rljacob commented Jan 11, 2024

wlin7 commented Jan 12, 2024 •

edited

Loading

rljacob commented Jan 12, 2024

rljacob commented Jan 12, 2024 •

edited

Loading

rljacob commented Jan 12, 2024 •

edited

Loading

wlin7 commented Jan 12, 2024

rljacob commented Jan 12, 2024 •

edited

Loading

rljacob commented Jan 12, 2024

rljacob commented Jan 13, 2024

wlin7 commented Jan 18, 2024

wlin7 commented Jan 19, 2024 •

edited

Loading

bishtgautam commented Jan 19, 2024

bishtgautam commented Jan 19, 2024

rljacob commented Jan 19, 2024

bishtgautam commented Jan 22, 2024

rljacob commented Jan 23, 2024

rljacob commented Jan 24, 2024

Update ELM in the definition of multiple V3 compsets #6108

Update ELM in the definition of multiple V3 compsets #6108

Conversation

bishtgautam commented Dec 7, 2023 • edited by rljacob Loading

github-actions bot commented Dec 7, 2023 • edited Loading

bishtgautam commented Dec 7, 2023

rljacob commented Dec 7, 2023

rljacob commented Dec 13, 2023 • edited Loading

wlin7 commented Jan 4, 2024

rljacob commented Jan 4, 2024

wlin7 commented Jan 4, 2024

bbye commented Jan 5, 2024

wlin7 commented Jan 6, 2024

bishtgautam commented Jan 6, 2024 • edited Loading

rljacob commented Jan 11, 2024

wlin7 commented Jan 12, 2024 • edited Loading

rljacob commented Jan 12, 2024

rljacob commented Jan 12, 2024 • edited Loading

rljacob commented Jan 12, 2024 • edited Loading

wlin7 commented Jan 12, 2024

rljacob commented Jan 12, 2024 • edited Loading

rljacob commented Jan 12, 2024

rljacob commented Jan 13, 2024

wlin7 commented Jan 18, 2024

wlin7 commented Jan 19, 2024 • edited Loading

bishtgautam commented Jan 19, 2024

bishtgautam commented Jan 19, 2024

rljacob commented Jan 19, 2024

bishtgautam commented Jan 22, 2024

rljacob commented Jan 23, 2024

rljacob commented Jan 24, 2024

bishtgautam commented Dec 7, 2023 •

edited by rljacob

Loading

github-actions bot commented Dec 7, 2023 •

edited

Loading

rljacob commented Dec 13, 2023 •

edited

Loading

bishtgautam commented Jan 6, 2024 •

edited

Loading

wlin7 commented Jan 12, 2024 •

edited

Loading

rljacob commented Jan 12, 2024 •

edited

Loading

rljacob commented Jan 12, 2024 •

edited

Loading

rljacob commented Jan 12, 2024 •

edited

Loading

wlin7 commented Jan 19, 2024 •

edited

Loading