[MITgcm-devel] dgrid exchange

[Jean-Michel Campin]

Hi Jean,

d-grid exch is really like a-grid one, so it should be processed
in the same way.

I think we need to add:
> exch_uv_dgrid_3d_rl.F 
into file:
verification/OpenAD/code_oad/keepOriginal

Do you need also to change in the same way file:
> MITgcm_contrib/heimbach/OpenAD/code_regress/keepOriginal ?

Thanks,
Jean-Michel

On Tue, Jul 13, 2010 at 03:12:54AM -0500, Jean Utke wrote:
> Hi,
> 
> my AD  regression tests fail because of some missing exch
> routine.  I found Gael's  set of changes starting with:
> 
> Update of /u/gcmpack/MITgcm/model/src
> In directory faulks.csail.mit.edu:/u/u2/gforget/MITgcm/model/src
> 
> Modified Files:
> 	do_fields_blocking_exchanges.F 
> Log Message:
> fix adjoint of cd scheme for curvilinear grids etc., 
> which requires the adjoint of dgrid exchanges.
> 
> 
> I was under the impression after the last set of changes to 
> EXCH1/2 that the adjoints now are done for only a very 
> few exch routines... Is the inclusion of the dgrid_3d a
> departure from that?
> 
> Thanks, 
> 
> jean 
> 
> 
> -- 
> Jean Utke
> Argonne National Lab. / MCS (240-1156)
> 9700 S Cass Ave., Argonne, IL, 60439, USA
> phone +1 630 252 4552 / mobile  +1 630 363 5753
> 
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel