[MITgcm-support] vertical diffusivity

[Stanislav Martyanov]

Hello!

I think I've found a small bug in the vertical diffusivity calculation. But
I'm not sure if it's not just a feature..

In s/r calc_3d_diffusivity (line 116, MITgcm version 67x) for T and S we
have two different assignments:

       IF ( trIdentity.EQ.GAD_TEMPERATURE ) THEN
        DO k = 1,Nr
         DO j = 1-OLy,sNy+OLy
          DO i = 1-OLx,sNx+OLx
           KappaRTr(i,j,k) = KappaRTr(i,j,k)
#ifdef ALLOW_3D_DIFFKR
     &          + diffKr(i,j,k,bi,bj)
#else
     &          + diffKrNrT(k)
#endif
          ENDDO
         ENDDO
        ENDDO
       ELSEIF ( trIdentity.EQ.GAD_SALINITY) THEN
        DO k = 1,Nr
         DO j = 1-OLy, sNy+OLy
          DO i = 1-OLx, sNx+OLx
           KappaRTr(i,j,k) = KappaRTr(i,j,k)
#ifdef ALLOW_3D_DIFFKR
     &          + diffKr(i,j,k,bi,bj)
#else
     &          + diffKrNrS(k)
#endif

In other words, we assign either diffKrNrT, or diffKrNrS, depending on the
'trIdentity'. But later, and when GGL90 is activated, in the s/r
GGL90_CALC_DIFF we always subtract only diffKrNrS (line 52, 63).

            KappaRx(i,j,k) = KappaRx(i,j,k)
     &                  +( GGL90diffKr(i,j,k,bi,bj)
     &                     - diffKrNrS(k) )

As I understand, if we set diffKrT and diffKrS to different values in the
main 'data' file, then we will not be able to flawlessly cancel KappaRx
(which is equal to diffKrNrT in the case of temperature) and diffKrNrS in
the s/r GGL90_CALC_DIFF. The same holds for the ptracer case. Of course, it
all matters only if we consider different values for temperature, salinity
and ptracers' diffusivities set in the 'data' file.

Best regards,
Stanislav Martyanov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-support/attachments/20220119/4e7d468a/attachment.html>