[MITgcm-devel] KPP, once again

[Martin Losch]

Hi there,

I am trying to make KPP work with my shelfice code. One of the main  
difficulties are that KPP always counts the maximum number of grid  
points from the top, but if some of the top layers are "not wet" that  
is shelfice, then this way of counting does not work, see kpp_init.F
>                   do k = 1, Nr
>                      nzmax(i,j,bi,bj) = nzmax(i,j,bi,bj)
>      &                                + INT(maskC(i,j,k,bi,bj))
>                   end do
Lateron in nzmax is passed a kmtj to kppmix where it does all sorts  
of harm. Do you think that replacing the above code with nzmax =  
klowc (which is the k-index of the lowest wet point computed in  
ini_masks_etc) is critical?

As it is now, nzmax = klowc-ktopc+1 de-facto (where ktopc is  
uppermost wet cell, not necessarily 1 in the case of shelfice).  
Unfortunately, ktopc is part of the shelfice-pkg and therefore I  
cannot really use it in KPP. Probably don't need to anyway.

Does anybody really know what's going on in KPPMIX? (o:

Martin