[MITgcm-devel] changes in OBCs pkg

Jean-Michel Campin jmc at ocean.mit.edu
Wed Dec 7 10:51:46 EST 2005


Martin,

There is an alternative to the use of a 2D-exchange vertical slabs:
Assuming that the model read a global file (with the OB values),
the whole field is available, and we just have to figure out
where to take the values to fill-in the overlap.
It's similar to what is done in OBCS_INIT_FIXED when
the indices (OB_Ie,OB_Iw,OB_Jn & OB_Js) are set.

Jean-Michel

On Wed, Dec 07, 2005 at 04:32:57PM +0100, Martin Losch wrote:
> OK, I didn't think of the exchanges.
> In obcs_init_variables.F Patrick and I put a block with exchanges at 
> the end for exactly the same reason. It's not there for passive tracers 
> now (my job), but in fact it should be at the end of the 
> obcs_prescribe_read routine, except that there isn't a 2D-exchange for 
> vertical slabs, is there? I don't have a good suggestion for this 
> except that there should be exchanges for these fields.
> Martin
> 



More information about the MITgcm-devel mailing list