Allow SHiELD_physics to use the full coupler fluxes. #53
Conversation
Kun's changes for SHiELD+MOM6 coupling
| @@ -1451,6 +1484,10 @@ subroutine GFS_physics_driver & | |||
| Diag%u1(:) = Statein%ugrs(:,1) | |||
| Diag%v1(:) = Statein%vgrs(:,1) | |||
| Sfcprop%qsfc(:) = qss(:) | |||
|
|
|||
| ! KGao | |||
| Diag%hflx(:) = hflx(:) | |||
There was a problem hiding this comment.
How do these differ from the existing diagnostics? Is this exposing the behavior of the coupler itself rather than the flux "seen" by the atmosphere?
There was a problem hiding this comment.
hflx and evap are two new diagnostics, which were to ensure the fields seen by the SHiELD physics are consistent with what in the coupler. They represent the temp and q fluxes (not the sensible and latent heat fluxes).
| @@ -6979,6 +7003,30 @@ subroutine gfdl_diag_register(Time, Sfcprop, Gfs_diag, Model, Cldprop, Atm_block | |||
| Diag(idx)%data(nb)%var2 => Sfcprop(nb)%uustar(:) | |||
| enddo | |||
|
|
|||
| ! KGao | |||
There was a problem hiding this comment.
Are these so that we can have arrays holding the flux directly from the coupler? If so, why not use the arrays in the Coupling% data structure?
There was a problem hiding this comment.
Yes, we let the Sfcprop% data structure holds surface layer variables (heat fluxes, ocean temp, z0/zt) passed from the coupler. It is a convenient choice as some fields are already defined. There should be a merge request to the atmos_driver repo to reflect this.
I am not sure of the role of Coupling% in SHiELD physics. I don't think Joseph and I have touched it for the SHiELD-MOM6 coupling.
There was a problem hiding this comment.
That is fine; I'm not sure whether Coupling% is actually hooked up to anything. Sfcprop is already being used and is convenient, but be aware that it may be altered in subtle ways elsewhere in the physics driver.
There was a problem hiding this comment.
The coupling arrays may be useful if we need to accumulate something over timesteps instead of passing instantaneous values.
SHiELD_physics/GFS_layer/GFS_typedefs.F90
Line 1911 in c93b388
| if (Model%sfc_gfdl) then | ||
| ! kgao - need a logic to ensure sfc_coupled is true when coupled with MOM6 | ||
| if (Model%sfc_coupled) then | ||
| ! a version of sfc_diff from coupling with MOM6 by kgao |
There was a problem hiding this comment.
Useful comment much appreciated.
Modification needed for the SHiEMOM model where the fluxes from the full coupler are used in the physics.
Developed by @gaokun227