Coupled BGC enabling and enhancements to ocean BGC #1801

maltrud · 2017-09-21T17:04:32Z

This PR defines several coupled BGC (CBGC) compsets. in order to do this, enhancements in the ocean/sea ice BGC codes and configuration scripts were needed. in addition, there are some ocean BGC bug fixes and diagnostic enhancements.

Conflicts: components/mpas-o/cime_config/buildnml components/mpas-o/model

jonbob

I tested this with PET_Ln9.T62_oQU240.GMPAS-NYF.edison_intel and it all looks good. Only things that snuck in and maybe shouldn't are:

env_mach_specify.py? I don't think we want to mess with that
in mpas-o/cime_config/buildnml, you change the restoring file for oEC60to30v3?
in mpas-o/cime_config/buildnml, you take out he "activeTracers" var_array?

maltrud · 2017-09-21T17:28:43Z

@jonbob thanks for pointing out those differences.

env_mach_specific.py: the "module list" command failed for some reason on grizzly so i got rid of it. definitely should be restored to original and i'll just deal with it when on grizzly.
mpas-o/cime_config/buildnml: i wanted to make sure i wasn't using the old file with December SSS set to zero. this should also be restored to the original since it is the corrected file.
mpas-o/cime_config/buildnml: we want all tracers to be output by default. if no BGC tracers are enabled then this includes just the activeTracers, so no functionality is lost.

jgfouca · 2017-09-21T17:53:34Z

cime/scripts/lib/CIME/XML/env_mach_specific.py

@@ -112,7 +112,8 @@ def list_modules(self):
            source_cmd = ""

        if (module_system in ["module", "module_lmod"]):
-            return run_cmd_no_fail("%smodule list" % source_cmd, combine_output=True)
+          # return run_cmd_no_fail("%smodule list" % source_cmd, combine_output=True)


What's going on here?

jgfouca

Restore list_modules plz.

jonbob

Issues found during initial review are fixed

jonbob · 2017-09-21T18:02:11Z

@jgfouca - I saw that as well. It should be fixed now, if you want to update your review

jgfouca · 2017-09-21T18:07:22Z

@maltrud could you describe the error you were getting from list_modules?

susburrows · 2017-09-21T18:16:51Z

@maltrud and @jonbob , thanks for getting this in! Just want to make sure @rljacob et al are aware that this PR is critical for the CBGC team and it would be good to get it merged as soon as possible -- ideally before the next batch of CIME updates. We would like to have further testing, development, and performance optimization of the CBGC configurations running from master instead of a feature branch, which should make things easier and more efficient moving forward.

maltrud · 2017-09-21T18:20:37Z

@jgfouca here's what happens if i don't remove the "module list" bit:
mm@gr-fe1.lanl.gov {1006}% ./case.setup
Machine/Decomp/Pes configuration has already been done ...skipping
Calling drv buildnml

….etc….

Finished creating component namelists
ERROR: Command: 'module list' failed with error '/bin/sh: module: command not found'

since it isn't at all necessary, i just remove it and am able to proceed with no problems.

jgfouca · 2017-09-21T18:38:37Z

@maltrud it sounds like the machine is configured incorrectly. It should be using modulecmd, not the module bash function.

susburrows · 2017-09-21T23:03:39Z

@rljacob it wouldn't hurt to have a tag so that we have a clear idea of which version we're referring to. Do you have a specific naming convention for the tags?

rljacob · 2017-09-21T23:36:00Z

Yes we do: https://acme-climate.atlassian.net/wiki/spaces/Docs/pages/2523172/Branch+Tag+and+Version+name+conventions

susburrows · 2017-09-21T23:48:09Z

Thanks! Since we are on beta tags now, does that mean we would add a beta tag for this version? Since there are new bug fixes and enhancements in this version, maybe an alpha tag would be more appropriate?

rljacob · 2017-09-21T23:51:33Z

It would be a beta tag. We've left alpha for v1.0.

susburrows · 2017-09-21T23:57:41Z

OK, so then it would just be the next increment of the beta tag?

Do you have a strategy for sequencing the currrent PRs? I know @maltrud is really eager to get this in before a lot of additional changes (particularly in CIME) go into the model that might make merging more difficult.

rljacob · 2017-09-22T01:55:03Z

Land model has a bunch of PRs they need first but some are non-BFB so will take a while. Is this really BFB? Could go in next week.
Don't fear the CIME changes. run_acme and the test suites are all tested with it and doing well.

rljacob · 2017-09-22T02:03:50Z

You should probably look at the pending land model PRs and see if you need them. Many are bug fixes which I assume would impact BGC in some way.

susburrows · 2017-09-25T19:41:56Z

@rljacob , we talked about this in the CBGC meeting and the consensus seems to be that we should wait until all PRs (this one and the pending land PRs) are in, and then create a tag. Also, since this PR modifies ocean and sea ice, can it be merged in parallel to merging the land PRs?

rljacob · 2017-09-25T19:43:43Z

Sure. This PR has the "Bug fix" label. What bug does it fix? (Should be an issue in github.)

jonbob · 2017-09-25T19:47:32Z

No bug -- that was probably just included by mistake from the issue list

maltrud · 2017-09-25T19:55:50Z

there were a couple of small things (like a unit conversion issue) that i noticed as i was putting together the branch, so i added Bug Fix. but no issues to link to. sorry about that.

Coupled BGC enabling and enhancements to ocean BGC This PR defines several coupled BGC (CBGC) compsets. in order to do this, enhancements in the ocean/sea ice BGC codes and configuration scripts were needed. in addition, there are some ocean BGC bug fixes and diagnostic enhancements. Tested with: * ERS_Ld5.T62_oEC60to30v3.GMPAS-OECO-IAF.cori-knl_gnu [NML] [BFB]

jonbob · 2017-10-02T18:14:57Z

Merged to next

Test suite: scripts_regression_tests.py, PRE.f09_f09_ADESP_TEST.hobart_nag, PRE.f19_f19.ADESP_TEST.hobart_nag Test baseline: Test namelist changes: Test status: bit for bit Fixes #1801 User interface changes?: Update gh-pages html (Y/N)?: Code review:

Remove mask specification for test using DOCN%SOMAQP. The mg16 and mg17 grid has a hole over Greenland. This was causing Nans when running SOM aquaplanet at f09_f09_mg17. The fix is not to specify a mask when using DOCN%SOMAQP. Test suite: scripts_regression_tests.py, PRE.f09_f09_ADESP_TEST.hobart_nag, PRE.f19_f19.ADESP_TEST.hobart_nag Test baseline: Test namelist changes: Test status: bit for bit Fixes #1801 User interface changes?: Update gh-pages html (Y/N)?: Code review: jedwards

Coupled BGC enabling and enhancements to ocean BGC This PR defines several coupled BGC (CBGC) compsets. in order to do this, enhancements in the ocean/sea ice BGC codes and configuration scripts were needed. in addition, there are some ocean BGC bug fixes and diagnostic enhancements. Tested with: * ERS_Ld5.T62_oEC60to30v3.GMPAS-OECO-IAF.cori-knl_gnu [NML] [BFB]

jonbob · 2017-10-04T17:37:18Z

merged to master

susburrows · 2017-10-04T18:07:56Z

Great news, thank you Jon!

jonbob · 2017-10-04T18:12:48Z

Well @susburrows and @maltrud, you all have worked a long time to get this done -- glad I could be part of it

maltrud · 2017-10-04T22:08:57Z

yes, indeed. thanks @jonbob for all of your help with this. now for the next round of updates.....

mark-petersen · 2017-10-05T00:17:59Z

Yes, thanks to everyone. After a long summer, the log jam is finally clearing.

Coupled BGC enabling and enhancements to ocean BGC This PR defines several coupled BGC (CBGC) compsets. in order to do this, enhancements in the ocean/sea ice BGC codes and configuration scripts were needed. in addition, there are some ocean BGC bug fixes and diagnostic enhancements. Tested with: * ERS_Ld5.T62_oEC60to30v3.GMPAS-OECO-IAF.cori-knl_gnu [NML] [BFB]

Test suite: scripts_regression_tests.py, PRE.f09_f09_ADESP_TEST.hobart_nag, PRE.f19_f19.ADESP_TEST.hobart_nag Test baseline: Test namelist changes: Test status: bit for bit Fixes #1801 User interface changes?: Update gh-pages html (Y/N)?: Code review:

Remove mask specification for test using DOCN%SOMAQP. The mg16 and mg17 grid has a hole over Greenland. This was causing Nans when running SOM aquaplanet at f09_f09_mg17. The fix is not to specify a mask when using DOCN%SOMAQP. Test suite: scripts_regression_tests.py, PRE.f09_f09_ADESP_TEST.hobart_nag, PRE.f19_f19.ADESP_TEST.hobart_nag Test baseline: Test namelist changes: Test status: bit for bit Fixes #1801 User interface changes?: Update gh-pages html (Y/N)?: Code review: jedwards

maltrud and others added 6 commits September 14, 2017 14:09

intial changes to ACME config/bld files for MPAS-O/CICE BGC

281e4c1

Conflicts: components/mpas-o/cime_config/buildnml components/mpas-o/model

more changes for CBGC missed in previous commit

ae8091b

add BGC updates and new CBGC compsets

4d1a4ff

Update ocean submodule for BGC updates.

5118fec

Update mpas-o and mpas-cice and supporting bld changes

ddd13c4

Update mpas-cice hash and supporting build scripts

02c1e26

maltrud added BFB PR leaves answers BFB bug fix PR compset Coupled Model NML mpas-ocean Scripts mpas-seaice labels Sep 21, 2017

maltrud assigned jonbob Sep 21, 2017

maltrud requested review from mark-petersen and jonbob September 21, 2017 17:04

jonbob requested changes Sep 21, 2017

View reviewed changes

rljacob requested a review from jgfouca September 21, 2017 17:16

jgfouca reviewed Sep 21, 2017

View reviewed changes

jgfouca requested changes Sep 21, 2017

View reviewed changes

Fix issues found during review.

09f664c

jonbob approved these changes Sep 21, 2017

View reviewed changes

jgfouca approved these changes Sep 21, 2017

View reviewed changes

rljacob removed the bug fix PR label Sep 25, 2017

rljacob added this to the v1.0beta3 milestone Sep 26, 2017

jonbob merged commit 09f664c into master Oct 4, 2017

jonbob deleted the maltrud/ocean/CBGC_initial_version branch October 4, 2017 17:37

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Coupled BGC enabling and enhancements to ocean BGC #1801

Coupled BGC enabling and enhancements to ocean BGC #1801

maltrud commented Sep 21, 2017

jonbob left a comment

maltrud commented Sep 21, 2017

jgfouca Sep 21, 2017

jgfouca left a comment

jonbob left a comment

jonbob commented Sep 21, 2017

jgfouca commented Sep 21, 2017

susburrows commented Sep 21, 2017 •

edited

Loading

maltrud commented Sep 21, 2017

jgfouca commented Sep 21, 2017

susburrows commented Sep 21, 2017

rljacob commented Sep 21, 2017

susburrows commented Sep 21, 2017

rljacob commented Sep 21, 2017

susburrows commented Sep 21, 2017

rljacob commented Sep 22, 2017

rljacob commented Sep 22, 2017

susburrows commented Sep 25, 2017

rljacob commented Sep 25, 2017

jonbob commented Sep 25, 2017

maltrud commented Sep 25, 2017

jonbob commented Oct 2, 2017

jonbob commented Oct 4, 2017

susburrows commented Oct 4, 2017

jonbob commented Oct 4, 2017

maltrud commented Oct 4, 2017

mark-petersen commented Oct 5, 2017

Coupled BGC enabling and enhancements to ocean BGC #1801

Coupled BGC enabling and enhancements to ocean BGC #1801

Conversation

maltrud commented Sep 21, 2017

jonbob left a comment

Choose a reason for hiding this comment

maltrud commented Sep 21, 2017

jgfouca Sep 21, 2017

Choose a reason for hiding this comment

jgfouca left a comment

Choose a reason for hiding this comment

jonbob left a comment

Choose a reason for hiding this comment

jonbob commented Sep 21, 2017

jgfouca commented Sep 21, 2017

susburrows commented Sep 21, 2017 • edited Loading

maltrud commented Sep 21, 2017

jgfouca commented Sep 21, 2017

susburrows commented Sep 21, 2017

rljacob commented Sep 21, 2017

susburrows commented Sep 21, 2017

rljacob commented Sep 21, 2017

susburrows commented Sep 21, 2017

rljacob commented Sep 22, 2017

rljacob commented Sep 22, 2017

susburrows commented Sep 25, 2017

rljacob commented Sep 25, 2017

jonbob commented Sep 25, 2017

maltrud commented Sep 25, 2017

jonbob commented Oct 2, 2017

jonbob commented Oct 4, 2017

susburrows commented Oct 4, 2017

jonbob commented Oct 4, 2017

maltrud commented Oct 4, 2017

mark-petersen commented Oct 5, 2017

susburrows commented Sep 21, 2017 •

edited

Loading