[MITgcm-devel] verification/offline_exf_seaice revisited

[Jean-Michel Campin]

Hi Dimitris,

Reagarding the adjoint offline_exf_seaice experiment, I don't know much
except that the routines in code_ad are different from standard 
model/src version.

Regarding the forward offline_exf_seaice experiment, there is a 
"special" piece of code in pkg/exf "#ifdef ALLOW_BULK_OFFLINE"
which is certainly doing something, and indeed ALLOW_BULK_OFFLINE
is defined in offline_exf_seaice/code/EXF_OPTIONS.h

Cheers,
Jean-Michel

On Tue, Feb 14, 2012 at 04:38:57PM -0800, Menemenlis, Dimitris (3248) wrote:
> Patrick, Ian, Jean-Michel, Gael, and other offline_exf_seaice users and updaters, PLEASE HELP!
> 
> How does offline_exf_seaice work?
> 
> According to code/README, we (Patrick, Ian, and I) intended to have a time-varying SST specified using climsstfile.
> But if we set "climsstTauRelax = 0.0," in data.exf, this will turn off SST relaxation.
> This was also the case for the original "tauThetaClimRelax = 0.,"
> in data and for 2005-vintage code when we first devised this experiment:
> 
>       IF ( tauThetaClimRelax .EQ. 0.D0 ) THEN
>        doThetaClimRelax     = .FALSE.
>        lambdaThetaClimRelax = 0.D0
>       ELSE
>        doThetaClimRelax     = .TRUE.
>        lambdaThetaClimRelax = 1./tauThetaClimRelax
>       ENDIF
> 
> Shouldn't we instead specify
>  climsstTauRelax    =   deltaT,
> so as to have a one-time step adjustment of SST to the value specified in climsstfile ?
> 
> Also if we set "tempStepping=.FALSE.," in data, doesn't this shut down all temperature tendency terms?
> So I am mystified how offline_exf_seaice does (or did) what it is meant to do?
> 
> What am I missing?
> 
> Dimitris Menemenlis
> cell: 818-625-6498
> 

> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel