support for catchcn ensemble run #584

weiyuan-jiang · 2022-09-22T19:12:36Z

Add CatchCN-specific export variables to EnsGridComp.
Required to support ensemble simulations with CatchCN in GEOSldas needed by @gmao-jkolassa.

Dependencies:

Depends on GCM GC PR support for catchcn ensemble run GEOSgcm_GridComp#645, which in turn depends on MAPL v2.27.1 (See merged PR add all exports from a direct child MAPL#1683.)

Prior to merging, change components.yaml accordingly.

After merging, change the following input files for the LDAS Nightly tests:

/discover/nobackup/mathomp4/LDAS_Restarts/NGHTLY_TST_TV4000/HISTORY_L4_SM.rc
/discover/nobackup/mathomp4/LDAS_Restarts/NGHTLY_TST_CS/assim/ldas179b8_C180/run/HISTORY.rc

Specifically, replace VEGDYN0000 with VEGDYN_e0000

weiyuan-jiang · 2022-09-22T19:37:45Z

@gmao-jkolassa , if you want to test ensemble run, please use this fresh checkout of the branch .

src/Applications/LDAS_App/process_hist.csh

gmao-jkolassa · 2022-09-23T21:00:50Z

This is just to confirm that an ensemble experiment using the current version of this branch works without any issues for both CatchCNCLM40 and CatchCNCLM45. Note that I only tested that it runs. Please let me know if you would like a full science-validation of this.

src/Applications/LDAS_App/process_hist.csh

- bug fix in format string in get_ensid_string() - replaced variable "id_string" with "ensid_string" for clarity - cleanup and additional clarification of 3-digit ensid string for ensemble met forcing

src/Applications/LDAS_App/ldas_setup

src/Components/GEOSldas_GridComp/GEOSlandassim_GridComp/GEOS_LandAssimGridComp.F90

src/Components/GEOSldas_GridComp/GEOSmetforce_GridComp/GEOS_MetforceGridComp.F90

src/Components/GEOSldas_GridComp/Shared/LDAS_ensdrv_Globals.F90

weiyuan-jiang · 2022-10-06T12:13:50Z

We have to change MAPL if we don't add 2 to the ens_id_with. I think the id should include "_exxxx" even we drop of the backward compatibility. @gmao-rreichle

weiyuan-jiang · 2022-10-06T13:46:05Z

@gmao-rreichle @gmao-jkolassa It should work now. But the nightly test would fail because we have hard coded HISTORY.rc for global assim and globlecs assim

gmao-rreichle · 2022-10-06T13:55:21Z

We have to change MAPL if we don't add 2 to the ens_id_with. I think the id should include "_exxxx" even we drop of the backward compatibility. @gmao-rreichle

@weiyuan-jiang, I don't understand how MAPL comes into this. I agree that we don't want to change MAPL, but maybe there's another way to distinguish between the number of digits and the total string length (ens_id_width) in LDAS.

weiyuan-jiang · 2022-10-06T14:19:24Z

These two assert should ease your concern if the users still use 4 for the width. Actually, the 6 is used in the lads_setup which should not have problem.
](

GEOSldas/src/Components/GEOSldas_GridComp/GEOS_LdasGridComp.F90

Lines 179 to 180 in a045efd

    
           _ASSERT( ens_id_width < 12, "need 1 billion ensemble members? increase ens_id_width first") 
        
           _ASSERT( ens_id_width >= 5, "need at least 5 (_exxx) for hard coded ensemble forcing ( 3, xxx)")

)

weiyuan-jiang · 2022-10-06T14:22:40Z

The way we extract the template has to change in MAPL if we change the meaning of width.
https://github.com/GEOS-ESM/MAPL/blob/160a8d720f5357d07d419818db7f53d51d471f2b/generic/MAPL_Generic.F90#L1508-L1509

gmao-jkolassa · 2022-10-06T15:10:27Z

I ran two shorts tests with the latest version of this branch. It builds successfully and runs successfully for both CatchCN40 and CatchCN45.

weiyuan-jiang · 2022-10-06T18:10:33Z

When this branch is merged, the two files for nightly test should be changed too:
/discover/nobackup/mathomp4/LDAS_Restarts/NGHTLY_TST_TV4000/HISTORY_L4_SM.rc
/discover/nobackup/mathomp4/LDAS_Restarts/NGHTLY_TST_CS/assim/ldas179b8_C180/run/HISTORY.rc

replace VEGDYN0000 with VEGDYN_e0000

@mathomp4 , We will let you know when we merge

gmao-rreichle · 2022-10-06T21:54:52Z

@weiyuan-jiang, @gmao-jkolassa, I applied a hopefully final set of minimal edits f1bd407
I don't think we need to retest the new functionality.
@weiyuan-jiang, did you run the LDAS Nightly tests with the above-mentioned edits to HISTORY.rc? If we are sure the GEOSldas PR is 0-diff for GEOSldas, we can ask @biljanaorescanin to test the matching GCM GC PR GEOS-ESM/GEOSgcm_GridComp#645 and then ask Scott to merge the matching GCM GC PR as soon as he can.
[I inadvertently closed the PR and then reopened it.]

weiyuan-jiang · 2022-10-07T13:50:09Z

it passed the test.

weiyuan-jiang added 2 commits September 22, 2022 14:57

enable catchcn ensemble run and export more variables

337b1c3

adjust component for testing

12d176d

weiyuan-jiang added 0-diff Contingent -- Do Not Approve labels Sep 22, 2022

weiyuan-jiang requested a review from a team as a code owner September 22, 2022 19:12

gmao-rreichle mentioned this pull request Sep 22, 2022

support for catchcn ensemble run GEOS-ESM/GEOSgcm_GridComp#645

Merged

gmao-rreichle changed the title ~~Feature/wjiang/catchcn ensemble new~~ support for catchcn ensemble run Sep 22, 2022

gmao-rreichle marked this pull request as draft September 22, 2022 21:25

gmao-rreichle reviewed Sep 22, 2022

View reviewed changes

src/Applications/LDAS_App/process_hist.csh Outdated Show resolved Hide resolved

gmao-rreichle reviewed Sep 23, 2022

View reviewed changes

src/Applications/LDAS_App/process_hist.csh Outdated Show resolved Hide resolved

weiyuan-jiang and others added 5 commits September 28, 2022 13:06

get CNFROOTC from ENSAVG grid comp

86e7de0

change ensid suffix from xxxx to _exxxx

9d4ac2d

change ens_id_string to _exxxx

7cea883

additional modifications of previous commit:

c24b96f

- bug fix in format string in get_ensid_string() - replaced variable "id_string" with "ensid_string" for clarity - cleanup and additional clarification of 3-digit ensid string for ensemble met forcing

revised comment line about ensid string in GEOS_MetForceGridComp.F90

5e6dc8c

gmao-rreichle reviewed Sep 30, 2022

View reviewed changes

src/Applications/LDAS_App/ldas_setup Outdated Show resolved Hide resolved

gmao-rreichle reviewed Sep 30, 2022

View reviewed changes

src/Components/GEOSldas_GridComp/GEOSlandassim_GridComp/GEOS_LandAssimGridComp.F90 Outdated Show resolved Hide resolved

gmao-rreichle reviewed Sep 30, 2022

View reviewed changes

src/Components/GEOSldas_GridComp/GEOSlandassim_GridComp/GEOS_LandAssimGridComp.F90 Outdated Show resolved Hide resolved

gmao-rreichle reviewed Sep 30, 2022

View reviewed changes

src/Components/GEOSldas_GridComp/GEOSmetforce_GridComp/GEOS_MetforceGridComp.F90 Show resolved Hide resolved

weiyuan-jiang added 2 commits October 3, 2022 09:17

change ens_id_width back to the meaning of "width of digit"

78cc7d5

fix some trivial errors

99d4140

gmao-rreichle reviewed Oct 3, 2022

View reviewed changes

src/Components/GEOSldas_GridComp/Shared/LDAS_ensdrv_Globals.F90 Outdated Show resolved Hide resolved

weiyuan-jiang added 3 commits October 5, 2022 09:47

fix id_string in landpert

864e822

minor changes.

3d0c007

add options for GetPointer when a variabe is missing

e5d4b65

add 2 to ens_id_width for "_e"

a045efd

minimal clean-up of comments and range checks re. ens_id_width

f1bd407

gmao-rreichle closed this Oct 6, 2022

gmao-rreichle reopened this Oct 6, 2022

gmao-rreichle added enhancement New feature or request infrastructure labels Oct 6, 2022

gmao-rreichle mentioned this pull request Oct 11, 2022

Bug fix: Make "lenkf.j -debug" work (again) #587

Merged

gmao-rreichle added 3 commits October 12, 2022 15:33

Merge branch 'develop' into feature/wjiang/catchcn_ensemble_new

1e50d2f

add comments about provenance of CatchCN variables

ebe8948

updating components.yaml for merge into develop

3437c07

gmao-rreichle removed the Contingent -- Do Not Approve label Oct 18, 2022

gmao-rreichle marked this pull request as ready for review October 18, 2022 16:52

gmao-rreichle approved these changes Oct 18, 2022

View reviewed changes

gmao-rreichle merged commit 9a7ec25 into develop Oct 18, 2022

gmao-rreichle deleted the feature/wjiang/catchcn_ensemble_new branch October 18, 2022 16:54

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

support for catchcn ensemble run #584

support for catchcn ensemble run #584

weiyuan-jiang commented Sep 22, 2022 •

edited by gmao-rreichle

Loading

weiyuan-jiang commented Sep 22, 2022

gmao-jkolassa commented Sep 23, 2022

weiyuan-jiang commented Oct 6, 2022

weiyuan-jiang commented Oct 6, 2022 •

edited

Loading

gmao-rreichle commented Oct 6, 2022

weiyuan-jiang commented Oct 6, 2022

weiyuan-jiang commented Oct 6, 2022 •

edited

Loading

gmao-jkolassa commented Oct 6, 2022

weiyuan-jiang commented Oct 6, 2022

gmao-rreichle commented Oct 6, 2022 •

edited

Loading

weiyuan-jiang commented Oct 7, 2022 •

edited

Loading

support for catchcn ensemble run #584

support for catchcn ensemble run #584

Conversation

weiyuan-jiang commented Sep 22, 2022 • edited by gmao-rreichle Loading

weiyuan-jiang commented Sep 22, 2022

gmao-jkolassa commented Sep 23, 2022

weiyuan-jiang commented Oct 6, 2022

weiyuan-jiang commented Oct 6, 2022 • edited Loading

gmao-rreichle commented Oct 6, 2022

weiyuan-jiang commented Oct 6, 2022

weiyuan-jiang commented Oct 6, 2022 • edited Loading

gmao-jkolassa commented Oct 6, 2022

weiyuan-jiang commented Oct 6, 2022

gmao-rreichle commented Oct 6, 2022 • edited Loading

weiyuan-jiang commented Oct 7, 2022 • edited Loading

weiyuan-jiang commented Sep 22, 2022 •

edited by gmao-rreichle

Loading

weiyuan-jiang commented Oct 6, 2022 •

edited

Loading

weiyuan-jiang commented Oct 6, 2022 •

edited

Loading

gmao-rreichle commented Oct 6, 2022 •

edited

Loading

weiyuan-jiang commented Oct 7, 2022 •

edited

Loading