change default resolution

[DeniseWorthen]

Description of changes

• Cleans up default_vars to reflect new default test resolution of c96-1deg ocean and ice.

• Reproduces existing tests by adding non-default configurations (eg c96-1/4deg ocean and ice) explicitly

• Removes many old settings in existing tests which are now defaults (eg. setting input_nml to the ccpp version).

• Adds s2s tests to replicate existing c384-1/4 deg tests for the control, benchmark, benchmark+waves and c384mx025 frac grid configurations. Also replicates c96mx025 restart test. These tests are b4b with existing baselines.

Three MOM6 default settings have changed to match MOM6 defaults for the 1deg model.

• MOM6 1deg does not have a river runoff file so MOM6_RIVER_RUNOFF is now set False

• MOM6_REPRO_LA and MOM6_THERMO_SPAN are now False (the default MOM6 setting). The non-default use of MOM6_THERMO_SPAN are removed since testing showed that using the default setting was b4b.

• A new MOM6 setting MOM6_USE_WAVES (default False) has been added in order to combine the wave and non-wave MOM input templates

Specific notes

Issues Fixed (include github issue #):

ufs-weather issue #221
ufs-s2s issue #218
Follow-on ufs-s2s PR #194

Are changes expected to change answers?

• No answer changes. Added tests repro existing tests

Specific changes:

• wave specific MOM input template is removed, cold nems.configure files are removed
• in addition to above added tests, the 1stepcold test and the (unused) bmark restart test is removed
• removes conversion from hours-> secs for CICE dumpfreq in edit_inputs since in CICE6 the restart dump frequency can no longer be set as "s".

Testing performed:

• machines: hera, orion, dell-p3
• added s2s tests are currently compared to the matching "cpld_" test.
• 35d RTs were tested by running both wave and non-wave for 1 date for the develop branch and for this branch with both the "cpld" version and the "s2s" versions. Baselines are not retained for these tests, but spot checking of the forecast files after 2 days showed all tests (develop and the two tests in the branch) reproduced each other.

Hashes used for testing:

• NEMS: f102e36
• CMEPS: 4d50adf
• FV3: 90bc1a4
• MOM6: 3ce81f4
• CICE: ac617cd
• WW3: 9c22b13
• FMS: f61416f
• stochastic_physics: 6eac84

Co-authored by:

[DeniseWorthen]

This PR is ready for final RT testing and commit. My understanding is that it will follow on PR #213.

[DeniseWorthen]

Tests are running on Orion, Dell-p3 and Hera.

[DeniseWorthen]

I need to re-run the RTs. I had mistakenly left three of the exisiting tests commented out (2threads, decomp and satmedmf).

[DeniseWorthen]

Ready for review.

[JessicaMeixner-NOAA]

There are a lot of changes, but they all look good to me. Thanks for all of this @DeniseWorthen I know this was no small task but we'll all appreciate the end result!

[SMoorthi-emc]

Just to add my 2 cents, I have been setting "stop_n=$FHMAX" in my script and I haven't seen any problem. The integration stopped at the right time and created the right history and restart files.

…

On Tue, Oct 20, 2020 at 11:04 PM Jun Wang ***@***.***> wrote: ***@***.**** commented on this pull request. ------------------------------ In parm/nems.configure.medcmeps_atm_ocn_ice.IN <#204 (comment)> : > @@ -42,7 +42,7 @@ ICE_attributes:: ProfileMemory = false OverwriteSlice = true mesh_ice = @[MESHICE] - stop_n = @[FHMAX] + stop_n = @[FV3_RESTART_H] I agree. If we do need to set ICE integration length, it would be "FHMAX-FHROT". In general, each subcomponent inherits the clock from top level component to control integration time. So I think the "The only function is to have the timing code inside of cice turn on and off" makes sense to me. But we can make changes later. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#204 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALLVRYSFK6Y3I2DUWNPDHYLSLZFVLANCNFSM4R7U2HPQ> .

-- Dr. Shrinivas Moorthi Research Meteorologist Modeling and Data Assimilation Branch Environmental Modeling Center / National Centers for Environmental Prediction 5830 University Research Court - (W/NP23), College Park MD 20740 USA Tel: (301)683-3718 e-mail: Shrinivas.Moorthi@noaa.gov Phone: (301) 683-3718 Fax: (301) 683-3718

[DeniseWorthen]

Thanks Moorthi. I was mistaken in my first statement about what stop_n does. It only controls the timers in CICE so it should work with pretty much any setting (other than 0 probably).