Check use of wet point indexing vs ocean/land mask in iHAMOCC

[TomasTorsvik]

This came up as a question when applying fields defined only for wet points in BLOM inside iHAMOCC routines (PR #179).

In the BLOM part, the wet points are referenced by index counters isp, ifp and ilp. These are used also in iHAMOCC modules mo_bgcmean and mo_intfcblom. Else where iHAMOCC is using a masking field omask.

Do we have a consistent way of handling "wet point" fields in iHAMOCC that have been imported from BLOM? Should we apply omask for all these fields, or is there a good reason to keep the mix of wet points and masking as we do today?

Implementation for wet points indexing:

   do j=1,jj
      do l=1,isp(j)
         do i=max(1,ifp(j,l)),min(ii,ilp(j,l))
            ...
         enddo
      enddo
   enddo

[JorgSchwinger]

Inside "core" HAMOCC (i.e. in hamocc4bcm and below) we have only used omask to indicate wet points. I see no reason to change this.