[MITgcm-devel] Re: sea ice obcs

[Dimitris Menemenlis]

Ian, I am copying reply to Michael and to devel list.

Jean-Michel and I have not yet discussed merging of your thermo code in the main 
development branch.  I don't know what would be best way to proceed.  As a 
starting point, could you prepare a list of the actual as opposed to the 
cosmetic changes that you have made to seaice_growth.F relative to some version 
of seaice_growth.F that resides on the main CVS branch.

The obcs-seaice code still needs work.  What I did last night is a starting 
point from which I am hoping that Michael (who requested this change for his 
Weddell Sea work) and others will take over.  In particular:

1. Although the obcs-seaice code is running and it seems to be doing what I 
think I asked it to do without breaking (too many) verification experiments, 
much more testing is required, including testing of the pickup capability and 
testing of the multidim and non-multidim advection schemes.

2. Right now OBCS are applied only to AREA and HEFF; the same needs to be done 
for HSNOW and  HSALT.  I left these two out on purpose so that Michael has a 
chance to get his hands a little bit dirty with the MITgcm code ;-)

3. For UICE and VICE, I set the southern and the western OBCS to zero.  I am not 
sure how general or how robust this strategy is and whether it will produce 
sensible results.

4. At the moment the obcs-seaice code only works with pkg/exf.  It does not work 
with obcs_external_fields_load.F.  I don't intend to fix this beyond issuing a 
warning in obcs_check.F

Dimitris

-- 
Dimitris Menemenlis <menemenlis at sbcglobal.net>
5056 Oakwood Ave, La Canada, CA 91011-2450
tel/fax: 818-790-6735;   cell: 818-625-6498

> Dimitris, It's so nice to see that this is happening.  I can cross it off of
> my infinitely long list.  By the way, I gave all of my thermo codes to
> Jean-Michel as I told you I would.  He said something about waiting for you
> to come before integrating them into the main branch latest version.  Well, I
> presume you came while I was away.  Did you guys work on that?  Succeed?  I
> tried to get my codes to work with the latest gcm version, mine was from May,
> but failed even after several days of effort. Ian