[MITgcm-devel] pTracer Pb: fixed

[Jean-Michel Campin]

Hi Patrick,

I found that you fixed the pTracer Pb in checkpoint51f_post
when you move  the initialisation of rFlx out of
S/R ptracers_integrate.F to thermodynamics.F

But the check-in message:

> o initialisation of rFlx extended to full array (required by TAF)
>   and shifted to thermodynamics
> o removed PTRACERS.h in ptracers routine
> o added surfacetendencyPtr to S/R parameter list pracers_forcing

did not tell that the forward code was wrong (I will not tell 
who is to blame), but Benjamin can confirm that it was. 

I found also an other problem, in PTRACERS_CONVECT:
      CALL CONVECTIVELY_MIXTRACER(
     I         bi,bj,k,weightA,weightB,
     U         pTracer(1-Olx,1-Oly,1,bi,bj,iTracer),
     I         myThid)
should be :
      CALL CONVECTIVELY_MIXTRACER(
     I         bi,bj,k,weightA,weightB,
     U         pTracer(1-Olx,1-Oly,1,1,1,iTracer),
     I         myThid)
I think we need to be very carefull in the agument list
because this kind of bug is in fact quiet hard to find.
(not always seen with a check-bounds compiler option like -C )
I will check in this fix right now.

See you,

Jean-Michel