[MITgcm-support] Question about RBCS gridding in rbcs_add_tendency.F

[Jean-Michel Campin]

Hi Katarina,

I don't know exactly what type of problem you found, but regarding the i,j loop ranges
in "rbcs_add_tendency.F", the current code is correct, i.e., we don't need to extend the
range to the full array size (1-OLx:sNx+OLx). Strictly speeaking, it would be enough for
Theta and Salt to update gTendency over 1:sNx but 1) this is not the case for momentum
tendency and 2) for some tracer diagnostics, we need 1 more point on each side of the tile.

Cheers,
Jean-Michel

On Tue, Jun 10, 2025 at 10:45:19AM -0400, Katarina Merk wrote:
> Dear all,
> 
> I???m encountering some unexpected behavior with the RBCS package. Specifically, the relaxation doesn't always seem to have the effect that I expect. 
> 
> While looking through the code, I came across a section in rbcs_add_tendency.F that did not make sense to me. The loops for the relaxations is currently written like:
> 
>      IF ( tracerNum.EQ.1 .AND. useRBCtemp ) THEN
>         irbc = MIN(maskLEN,tracerNum)
>         rec_tauRlx = rbcsVanishingFac/tauRelaxT
>         DO j=0,sNy+1
>          DO i=0,sNx+1
>           gTendency(i,j) = gTendency(i,j)
>      &       - RBC_mask(i,j,k,bi,bj,irbc)*rec_tauRlx
>      &        *( theta(i,j,k,bi,bj)- RBCtemp(i,j,k,bi,bj) )
> 
>          ENDDO
>         ENDDO
>       ENDIF
> 
> Should this instead use halo-aware indexing like:
> 
>       DO j = 1-OLy, sNy+OLy
>        DO i = 1-OLx, sNx+OLx
> Is the current gridding intentional, or could this be contributing to the relaxation issues I???m seeing?
> 
> Thanks for your time and help!
> 
> Best,
> Katarina

> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support