stress_mag passed to MOM6 not filled with proper values

[zhaobin74]

Currently in MOM6 plug, stress_mag, one of the components of ocean boundary forcing, is allocated but filled with zeros. The logic in MOM6 is that if a field is allocated, it will use its value, hence zeros are used. This apparently is not correct if certain mixing schemes use friction velocity for their purpose. @adarmenov first pointed out the output of ustar is zero everywhere, precisely due to this.

The fix is easy: do not allocate it such that MOM6 will compute it using wind stress (TAUX and TAUY).

[Dooruk]

Could we clarify the impact of this? Do both kpp and epbl use this? I know @sinakhani is testing an alternative MLD (mixing?) scheme, will this issue impact that as well?

[sinakhani]

I am running a test to diagnose "ustar" from diag table to see what ocean uses as ustar. Based on my understanding, if we set (USTAR_FORCING_VAR = "") in MOM_input/override, MOM6 uses wind stress to calculate ustar. I will update this comment when I diagnose ustar.

[zhaobin74]

I am running a test to diagnose "ustar" from diag table to see what ocean uses as ustar. Based on my understanding, if we set (USTAR_FORCING_VAR = "") in MOM_input/override, MOM6 uses wind stress to calculate ustar. I will update this comment when I diagnose ustar.

I have never heard about USTAR_FORCING_VAR. A quick search says the only place USTAR_FORCING_VAR showed up in MOM6 source code is in config_src/drivers/solo_driver/MOM_surface_forcing.F90 (line 1802). But we don't use solo_driver; we use FMS_cap

[zhaobin74]

Could we clarify the impact of this? Do both kpp and epbl use this? I know @sinakhani is testing an alternative MLD (mixing?) scheme, will this issue impact that as well?

good question. My runs are using epbl so it is affected. I am not sure about KPP. I don't have time to dig deeper.

[sinakhani]

Ok, based on the configuration we use in our setup, we are importing zero stress magnitude ( sqrt(TAUX^2 + TAUY^2) ) from MOM6 Plug into the ocean. This affects our upper ocean mixing/circulation. As discussed, we need to fix it. I have an update on (MOM6_GEOSPlug.F90) to send a correct stress_mag into MOM6. Here I have uploaded a file that shows u* calculated in GEOS versus u* outputted from MOM6 (I think they are in good/fair agreement). I will submit a PR for this to be reviewed by the ocean team. Thank you for the comments by @adarmenov @zhaobin74 @Dooruk @mfmehari @amolod . Any further comments are very welcome.