Updates for wave coupling in NUOPC #23
Conversation
- Adding wave information to mech_forcing type to pass from ice_ocean_boundary to wave types - This is only set-up to read surface Stokes drift and guess the wavelength as something reasonable for now to demonstrate that it works. This needs to be set-up properly before merging this into the main repository.
|
If we just update MOM6 with ufs-s2s-model, the regression tests all pass. See this branch: https://github.com/JessicaMeixner-NOAA/ufs-s2s-model/tree/feature/mom6onlycplwav The logs for the ufs-s2s-model with the wave-coupling updates are running, the baselines had to be remade due to recent changes in WW3 which changed the wave results, but the tests are still running. |
config_src/nuopc_driver/mom_cap.F90
Outdated
| @@ -606,6 +606,21 @@ subroutine InitializeAdvertise(gcomp, importState, exportState, clock, rc) | |||
| Ice_ocean_boundary%lrunoff = 0.0 | |||
| Ice_ocean_boundary%frunoff = 0.0 | |||
|
|
|||
| Ice_ocean_boundary%use_waves=ocean_state%use_waves | |||
There was a problem hiding this comment.
I don't see where Ice_ocean_boundary%use_waves is declared but it must be, otherwise the implicit none should catch it.
It seems it is only used in mom_cap_methods to allocate and fill the import state. You could probably do this w/o having to define ice_ocean_boundary%use_waves if you checked instead for the ustkb,vstkb being allocated in ice_ocean_boundary.
There was a problem hiding this comment.
The use_waves was added the Ice_ocean_boundary type. You are correct that we can check if ustkb is allocated instead. I am running the test with this final update and then will push that plus the logs up.
|
Here are the passed logs for all tests, pre- use_waves -> if associated update: Here are logs with the last update showing for both a test with and without waves it still passes: |
Merge in latest dev/gfdl updates
Retaing dimensional rescaling for several diagnostics: - Apply dimensional rescaling of fluxes%net_mass_src to and the net_mass_src argument to get_net_mass_forcing to [R Z T-1]. - Rescaled the patm argument of convert_state_to_ocean_type to [R L2 T-2] and press_to_Z to [Z T2 R-1 L-2]; these are not actually exercised yet, so this has a very limited scope, although three other local variables were also dimensionally rescaled. - Revised the line breaks in two calls to register_restart to place the units and conversion factos on the same line, following a widespread code pattern. - Used the scale argument in calls to global_area_integral or global_area_mean for 6 diagnostics, so that 3 other primary variables can be calculated in scaled units and rescaled via factors specified in the register_restart calls, following a widespread code pattern. All answers and output are bitwise identical.
These updates add on to @breichl updates for wave-ocean coupling with FMS. This adds the necessary fields to the IOB structure in the NUOPC cap.
This is related to Issue #2