Update mpas-source for Langmuir ocean vertical mixing parameterization #2952

mark-petersen · 2019-05-29T16:57:24Z

This merge adds Langmuir mixing parameterization via CVMix

Change the CVMix repository to use the 'theory-wave' approximation of Li et al., 2017 to estimate the Langmuir enhancement factor from 10-meter wind, surface friction velocity and boundary layer depth
New entries in mpaso namelist allowing different options of Langmuir mixing parameterization

This PR is climate changing if new flags are used to turn on the new mixing.

[BFB]
[NML]
[FCC]

mark-petersen · 2019-05-29T17:42:05Z

Please see discussion on testing at MPAS-Dev/MPAS-Model#134. This
PR also updates cvmix version for python 3, described at MPAS-Dev/MPAS-Model#251

Right now the 'Files changed' tab will include the changes in #2852. After that is merged, it will only show the changes for this PR.

The mpas-source submodule currently points to the branch e3sm/testing. After #2852 is successfully merged, then MPAS-Dev/MPAS-Model#134 and MPAS-Dev/MPAS-Model#251 can be merged into ocean/develop and e3sm/develop, and the changes here will only show the Langmuir updates in this PR.

mark-petersen · 2019-05-29T17:48:56Z

@jonbob and @vanroekel this branch is ready for the namelist updates and testing.

mark-petersen · 2019-05-30T18:25:50Z

All the threading code is now merged. This PR is ready to go. @vanroekel and @qingli411, please look at the 'Files changed' above to make sure you see what you expect. Then run a few tests on this branch to make sure it is the same as your long testing previously.

qingli411 · 2019-05-30T19:11:53Z

@mark-petersen Great. The 'Files changed' looks good to me. I will do some tests on this branch.

mark-petersen · 2019-05-30T19:23:29Z

Passed my standard tests:

PASS SMS_D_Ln9.T62_oEC60to30v3wLI.GMPAS-DIB-IAF-ISMF.cori-knl_gnu RUN time=272
PASS SMS_D_Ln9.T62_oEC60to30v3wLI.GMPAS-DIB-IAF-ISMF.cori-knl_intel RUN time=1116
PASS PEM_Ln9.ne30_oECv3wLI.A_WCYCL1850-DIB-ISMF_CMIP6.cori-knl_gnu COMPARE_base_modpes
PASS PET_Ln3.T62_oEC60to30v3wLI.GMPAS-DIB-IAF-ISMF.cori-knl_gnu COMPARE_base_single_thread

Everything for the mpas-source is done. I won't be able to work further on this PR, but I think you all can finish it.

@jonbob, there are three new namelist flags:

config_cvmix_kpp_use_theory_wave
config_cvmix_kpp_langmuir_mixing_opt
config_cvmix_kpp_langmuir_entrainment_opt

vanroekel · 2019-06-19T21:45:23Z

@jonbob should we append the buildnml changes to activate langmuir mixing to this PR?

qingli411 · 2019-06-19T22:00:37Z

@jonbob @vanroekel These are the changes to buildnml for mpas-ocean I need to activate Langmuir mixing in E3SM:
qingli411@fa73b3a#diff-f27161481241c47c1d8caad816cf6a24

jonbob · 2019-06-25T18:32:42Z

@qingli411 , @vanroekel - we have scripts that pick up changes in the Registry file and auto-generate something close to the right build-namelist files. So thanks for pointing me at the required changes, but that's not necessary. Are the defaults in Registry the ones you want in E3SM??

<config_cvmix_kpp_use_theory_wave>.false.</config_cvmix_kpp_use_theory_wave>
<config_cvmix_kpp_langmuir_mixing_opt>'NONE'</config_cvmix_kpp_langmuir_mixing_opt>
<config_cvmix_kpp_langmuir_entrainment_opt>'NONE'</config_cvmix_kpp_langmuir_entrainment_opt>

vanroekel · 2019-06-25T19:00:04Z

@jonbob I think we want in order

config_cvmix_kpp_use_theory_wave = .true.
config_cvmix_kpp_langmuir_mixing_opt = 'LF16'
config_cvmix_kpp_langmuir_entrainment_opt = 'LF17'

Is that right @qingli411?

qingli411 · 2019-06-25T19:05:53Z

@vanroekel @jonbob If we want to have the Langmuir turbulence parameterization turned on by default, we need

config_cvmix_kpp_use_theory_wave = .true.
config_cvmix_kpp_langmuir_mixing_opt = 'LF16'
config_cvmix_kpp_langmuir_entrainment_opt = 'LF17'

There is another modification that we need. To use the 'theory wave' option we need 10-meter wind speed being passed from coupler to mpas-ocean in components/mpas-ocean/driver/ocn_comp_mct.F.

mark-petersen · 2019-06-25T19:51:32Z

@qingli411 for second, you mean

config_cvmix_kpp_langmuir_mixing_opt = 'LWF16'

right?

I'm changing MPAS-Ocean to run stand-alone tests with these defaults as well. Mainly that just exercises the code in our tests.

qingli411 · 2019-06-25T20:13:33Z

@mark-petersen Yes, it's LWF16. Sorry.

jonbob · 2019-06-25T20:38:14Z

@qingli411 - do you need just the magnitude of the 10-m winds? And put into what mpas-ocean field? windSpeed10m?

mark-petersen · 2019-06-25T20:59:00Z

@vanroekel and @qingli411: When I run the MPAS-Ocean stand alone nightly test suite with these flags:

config_cvmix_kpp_use_theory_wave = .true.
config_cvmix_kpp_langmuir_mixing_opt = 'LWF16'
config_cvmix_kpp_langmuir_entrainment_opt = 'LF17'

I get a divide by zero. Adding 1e-15 to the denominator fixes it:

+++ b/src/core_ocean/shared/mpas_ocn_vmix_cvmix.F
@@ -469,10 +469,10 @@ contains
               ! compute Langmuir number and Langmuir enhancement factor
               ! TODO: theory-wave is the only option to get Langmuir number and enhancement factor for now <06-09-18, Qing Li> !
               if (config_cvmix_kpp_use_theory_wave .and. iceFraction(iCell) .lt. 0.05_RKIND) then
-                 langmuirNumber =  sqrt(surfaceFrictionVelocity(iCell) / &
+                 langmuirNumber =  sqrt(surfaceFrictionVelocity(iCell) / ( &
                          cvmix_kpp_ustokes_SL_model(windSpeed10m(iCell), &
                                                     boundaryLayerDepth(iCell), &
-                                                    cvmix_global_params))
+                                                    cvmix_global_params)+1e-15_RKIND) )
                  langmuirEnhancementFactor =  &
                          cvmix_kpp_EFactor_model(windSpeed10m(iCell), &
                                                  surfaceFrictionVelocity(iCell), &

I’m sure it’s because the 10-meter wind field comes in as zero, so you get a divide by zero. Still, we don’t want a potential divide by zero in the code. Should I add that 1e-15 into MPAS? Or would it mess something else up? It's possible that you could have an exact zero in E3SM somewhere.

qingli411 · 2019-06-25T23:38:16Z

@jonbob Yes, I only need the magnitude of the 10-meter wind speed and I was using 'windSpeed10m'

qingli411 · 2019-06-25T23:42:55Z

@mark-petersen I don't see a problem adding 1e-15 to the denominator as you did. Thanks for catching that.

qingli411 · 2019-06-25T23:56:05Z

Here is the new feature description page for Langmuir turbulence parameterization on Confluence:
W4_MPAS-Ocean_Langmuir_turbulence

I have run a 4-year B-case simulation of this PR and compared it with the first 4 years of a previous simulation I did for the MPAS-Ocean PR #134. The results are qualitatively similar. Here are the link to MPAS-Analysis:

B-case with this PR (year 1-4):
20190619.Langmuir-LF17.A_WCYCL1850S.ne30_oECv3_ICG.anvil1-4

First 4 year of the B-case run 20181129.Langmuir-LF17.A_WCYCL1850S.ne30_oECv3_ICG.anvil:
20181129.Langmuir-LF17.A_WCYCL1850S.ne30_oECv3_ICG.anvil_years1-4

vanroekel · 2019-06-26T01:59:46Z

agreed, adding 1E-15 seems like a very good idea. Thanks for catching this @mark-petersen

jonbob · 2019-07-01T15:41:07Z

Passes SMS.ne30_oECv3_ICG.A_WCYCL1850S_CMIP6.anvil_intel

mark-petersen · 2019-10-02T13:11:11Z

I just rebased on today's master. @jonbob could you run a few standard tests on this branch now?

@vanroekel and @qingli411 we need to fill out the rest of this page:
https://acme-climate.atlassian.net/wiki/spaces/ED/pages/988840065/W4+MPAS-Ocean+Langmuir+turbulence
Please add what you can based on the documents you have, and the tests you've already run. We will try to merge this in this week.

mark-petersen · 2019-10-02T13:19:49Z

@vanroekel I think you've checked this over before. Please approve this PR when you are ready.

vanroekel

changes and flags look good. Just need @jonbob to regenerate the buildnml bits with the auto generation scripts.

mark-petersen · 2019-10-08T15:20:57Z

components/mpas-ocean/bld/namelist_files/namelist_defaults_mpaso.xml

@@ -238,6 +238,9 @@
 <config_cvmix_kpp_stop_OBL_search>100.0</config_cvmix_kpp_stop_OBL_search>
 <config_cvmix_kpp_use_enhanced_diff>.true.</config_cvmix_kpp_use_enhanced_diff>
 <config_cvmix_kpp_nonlocal_with_implicit_mix>.false.</config_cvmix_kpp_nonlocal_with_implicit_mix>
+<config_cvmix_kpp_use_theory_wave>.true.</config_cvmix_kpp_use_theory_wave>


Change to false for this PR.

mark-petersen · 2019-10-08T15:21:17Z

components/mpas-ocean/bld/namelist_files/namelist_defaults_mpaso.xml

@@ -238,6 +238,9 @@
 <config_cvmix_kpp_stop_OBL_search>100.0</config_cvmix_kpp_stop_OBL_search>
 <config_cvmix_kpp_use_enhanced_diff>.true.</config_cvmix_kpp_use_enhanced_diff>
 <config_cvmix_kpp_nonlocal_with_implicit_mix>.false.</config_cvmix_kpp_nonlocal_with_implicit_mix>
+<config_cvmix_kpp_use_theory_wave>.true.</config_cvmix_kpp_use_theory_wave>
+<config_cvmix_kpp_langmuir_mixing_opt>'LWF16'</config_cvmix_kpp_langmuir_mixing_opt>
+<config_cvmix_kpp_langmuir_entrainment_opt>'LF17'</config_cvmix_kpp_langmuir_entrainment_opt>


Change to 'none' for this PR, for both of previous two.

mark-petersen · 2019-10-08T15:32:59Z

After speaking with @vanroekel, we realized that this PR will add the relevant flags and be BFB. A later PR will be climate changing and change these three flags back:

config_cvmix_kpp_use_theory_wave = .true.
config_cvmix_kpp_langmuir_mixing_opt = 'LWF16'
config_cvmix_kpp_langmuir_entrainment_opt = 'LF17'

mark-petersen · 2019-10-08T15:41:32Z

@jonbob This is rebased and ready to merge. Please give it a quick test.

Update mpas-source for Langmuir ocean vertical mixing parameterization This PR adds Langmuir mixing parameterization (via CVMix) support in mpas-ocean: * adds the option in the CVMix repository to use the 'theory-wave' approximation of Li et al., 2017 to estimate the Langmuir enhancement factor from 10-meter wind, surface friction velocity and boundary layer depth; and * adds new entries in mpaso namelist allowing different options of Langmuir mixing parameterization. The ocean source code has previously been available, but this PR adds the script and coupling changes required to use it. By default, these changes are off but this PR will be climate changing when the new flags are used to turn on the new mixing. [BFB] [NML] [FCC]

jonbob · 2019-10-09T17:09:18Z

merged to next

jonbob · 2019-10-09T19:10:46Z

@mark-petersen - testing shows that 'none' in your last commit needs to be 'NONE' -- apparently it is case-sensitive. Can you add this change please?

jonbob · 2019-10-09T20:27:16Z

@mark-petersen - never mind, I'll just do it so we can be sure if works for testing tonight

…2952) Re-merge to pick up developer error

jonbob · 2019-10-09T21:49:19Z

re-merged to next. Now passes:

SMS.T62_oQU240.CMPASO-NYF.anvil_intel

Update mpas-source for Langmuir ocean vertical mixing parameterization This PR adds Langmuir mixing parameterization (via CVMix) support in mpas-ocean: * adds the option in the CVMix repository to use the 'theory-wave' approximation of Li et al., 2017 to estimate the Langmuir enhancement factor from 10-meter wind, surface friction velocity and boundary layer depth; and * adds new entries in mpaso namelist allowing different options of Langmuir mixing parameterization. The ocean source code has previously been available, but this PR adds the script and coupling changes required to use it. By default, these changes are off but this PR will be climate changing when the new flags are used to turn on the new mixing. [BFB] [NML] [FCC]

jonbob · 2019-10-10T14:52:05Z

merged to master and expected NML DIFF's blessed (but note we will expect more on cori once it's back up)

mark-petersen added mpas-ocean non-BFB PR makes roundoff changes to answers. labels May 29, 2019

mark-petersen requested review from vanroekel and jonbob May 29, 2019 16:57

rljacob assigned jonbob May 30, 2019

rljacob removed the request for review from jonbob May 30, 2019 17:17

mark-petersen force-pushed the mark-petersen/ocean/langmuir_mixing branch from 01d94b0 to 1520c23 Compare May 30, 2019 18:18

rljacob added this to the v2.0alpha milestone Jun 14, 2019

This was referenced Jun 21, 2019

Build stops when using python 3.7 #2895

Open

Update mpas-source to current ocean-BGC submodule. #3006

Merged

jonbob added NML CC PR is climate changing labels Jun 25, 2019

mark-petersen force-pushed the mark-petersen/ocean/langmuir_mixing branch from e84b461 to 6b288f8 Compare October 2, 2019 12:07

vanroekel approved these changes Oct 3, 2019

View reviewed changes

mark-petersen commented Oct 8, 2019

View reviewed changes

mark-petersen added BFB PR leaves answers BFB and removed CC PR is climate changing non-BFB PR makes roundoff changes to answers. labels Oct 8, 2019

jonbob and others added 5 commits October 8, 2019 08:36

Update bld scripts to reflect changes to mpas-ocean Registry

2dbdce3

Update defaults to turn on this new mixing parameterization

ab982d1

Fill in required new 10m-wind field in ocn_import_mct

c6eb821

Add missing call to mpas_dmpar_exch_halo_field for new 10m-wind field

315d834

Set Langmuir mixing off for this PR

4255941

mark-petersen force-pushed the mark-petersen/ocean/langmuir_mixing branch from 6b288f8 to 4255941 Compare October 8, 2019 15:39

Fix case-sensitive error

ce76f08

jonbob added a commit that referenced this pull request Oct 9, 2019

Re-merge branch 'mark-petersen/ocean/langmuir_mixing' into next (PR #…

1e38f9d

…2952) Re-merge to pick up developer error

jonbob merged commit ce76f08 into master Oct 10, 2019

jonbob deleted the mark-petersen/ocean/langmuir_mixing branch October 10, 2019 14:52

xylar mentioned this pull request Nov 8, 2019

Use cime_env where avaiable E3SM-Project/E3SM_test_scripts#13

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Update mpas-source for Langmuir ocean vertical mixing parameterization #2952

Update mpas-source for Langmuir ocean vertical mixing parameterization #2952

mark-petersen commented May 29, 2019 •

edited by rljacob

mark-petersen commented May 29, 2019

mark-petersen commented May 29, 2019

mark-petersen commented May 30, 2019

qingli411 commented May 30, 2019

mark-petersen commented May 30, 2019

vanroekel commented Jun 19, 2019

qingli411 commented Jun 19, 2019

jonbob commented Jun 25, 2019

vanroekel commented Jun 25, 2019

qingli411 commented Jun 25, 2019

mark-petersen commented Jun 25, 2019

qingli411 commented Jun 25, 2019

jonbob commented Jun 25, 2019 •

edited

mark-petersen commented Jun 25, 2019

qingli411 commented Jun 25, 2019

qingli411 commented Jun 25, 2019

qingli411 commented Jun 25, 2019

vanroekel commented Jun 26, 2019

jonbob commented Jul 1, 2019

mark-petersen commented Oct 2, 2019

mark-petersen commented Oct 2, 2019

vanroekel left a comment

mark-petersen Oct 8, 2019

mark-petersen Oct 8, 2019

mark-petersen commented Oct 8, 2019

mark-petersen commented Oct 8, 2019

jonbob commented Oct 9, 2019

jonbob commented Oct 9, 2019

jonbob commented Oct 9, 2019

jonbob commented Oct 9, 2019

jonbob commented Oct 10, 2019

Update mpas-source for Langmuir ocean vertical mixing parameterization #2952

Update mpas-source for Langmuir ocean vertical mixing parameterization #2952

Conversation

mark-petersen commented May 29, 2019 • edited by rljacob

mark-petersen commented May 29, 2019

mark-petersen commented May 29, 2019

mark-petersen commented May 30, 2019

qingli411 commented May 30, 2019

mark-petersen commented May 30, 2019

vanroekel commented Jun 19, 2019

qingli411 commented Jun 19, 2019

jonbob commented Jun 25, 2019

vanroekel commented Jun 25, 2019

qingli411 commented Jun 25, 2019

mark-petersen commented Jun 25, 2019

qingli411 commented Jun 25, 2019

jonbob commented Jun 25, 2019 • edited

mark-petersen commented Jun 25, 2019

qingli411 commented Jun 25, 2019

qingli411 commented Jun 25, 2019

qingli411 commented Jun 25, 2019

vanroekel commented Jun 26, 2019

jonbob commented Jul 1, 2019

mark-petersen commented Oct 2, 2019

mark-petersen commented Oct 2, 2019

vanroekel left a comment

Choose a reason for hiding this comment

mark-petersen Oct 8, 2019

Choose a reason for hiding this comment

mark-petersen Oct 8, 2019

Choose a reason for hiding this comment

mark-petersen commented Oct 8, 2019

mark-petersen commented Oct 8, 2019

jonbob commented Oct 9, 2019

jonbob commented Oct 9, 2019

jonbob commented Oct 9, 2019

jonbob commented Oct 9, 2019

jonbob commented Oct 10, 2019

mark-petersen commented May 29, 2019 •

edited by rljacob

jonbob commented Jun 25, 2019 •

edited