[Mitgcm-support] RE: sea-ice code

[mitgcm-support at dev.mitgcm.org]

All,

 What do you think about moving toward passing state and grid
information around via argument lists internally for these, increasingly
large, ancillary packages? I'm uncomfortable with the degree to which
these non-core packages are being so tightly coupled to the internal
organization of GRID.h, DYNAMICS.h etc.... 

 Also SEAICE_GRID.h has absolutely no explanation of what the grid
variables are! That's bad :-).....

Chris

-----Original Message-----
From: Dimitris Menemenlis [mailto:menemenlis at jpl.nasa.gov] 
Sent: Wednesday, December 11, 2002 4:00 PM
To: heimbach at mit.edu
Cc: Alistair Adcroft; Curtis Heisey; cnh at plume.mit.edu;
support at mitgcm.org
Subject: Re: sea-ice code


>> Dimitris,
>> 
>> let me know if you need help with combing the sea ice and
>> the exf bulk formulae, or if there's a part that I could do.
>> 
>> cheers
>> -p.

Patrick et al., does the following look reasonable:

#ifdef INCLUDE_EXTERNAL_FORCING_PACKAGE
        CALL EXF_GETFORCING( mytime, myiter, mythid )
#else
        CALL EXTERNAL_FIELDS_LOAD( mytime, myiter, mythid )
#endif
#ifdef ALLOW_SEAICE
        IF ( useSEAICE ) CALL SEAICE_MODEL( myTime, myIter, myThid )
#endif

where SEAICE_MODEL will provide three forcing options:

   1) load atmospheric state, evaporation, and runoff
      from files (present configuration)

   2) load atmospheric state, evaporation, and runoff using
      EXF_GETFORCING (sea-ice bulk formulae would still be used)

   3) get atmospheric state, runoff, and open-ocean
      bulk-formulae fluxes from EXF_GETFORCING

Written this way, the sea-ice code can accommodate future decisions to
modify/replace EXF_GETFORCING or to package the open-ocean bulk formulae
separately.  Checks that seaice_model will get what it needs and that
there
will be no file-loading duplications will be moved to seaice_check.F,
rather
than to forward_step.F where they presently reside.

In related news, Jinlun has repeated the verification/lab_sea, LSR tests
using
his POP-based code.  There are no edge problems.  So the LSR sea-ice
dynamic
solver problem may simply be one too many world-cup beers on my part,
while
porting the code in Hamburg ... Hic ;-)

D.

-- 
Dimitris Menemenlis                      menemenlis at jpl.nasa.gov
Jet Propulsion Lab, MS 300-323           tel: 818-354-1656
4800 Oak Grove Dr, Pasadena CA 91109     fax: 818-393-6720