[MITgcm-devel] open boundaries for sea ice

[Martin Losch]

Hi Dimitris,
the simple answer to 1.: because it was done so in gad_advection.F,  
which was the starting point of seaice_advection.F
I guess the reason is that with this multidim advection, as  
documented in the header:
> C The algorithm is as follows:
> C \begin{itemize}
> C \item{$\theta^{(n+1/2)} = \theta^{(n)}
> C      - \Delta t \partial_x (u\theta^{(n)}) + \theta^{(n)}  
> \partial_x u$}
> C \item{$\theta^{(n+2/2)} = \theta^{(n+1/2)}
> C      - \Delta t \partial_y (v\theta^{(n+1/2)}) + \theta^{(n)}  
> \partial_y v$}
> C \item{$G_\theta = ( \theta^{(n+2/2)} - \theta^{(n)} )/\Delta t$}
> C \end{itemize}
so you update heff/area twice and then after each update you'll have  
to reapply the boundary conditions.
2. No, I don't, but they should be simple to adapt from  
obsc_apply_tloc.F: copy the file and replace all OB?t with OB?heff  
etc. OK, you have just done it while I am writing. Unfair.

Don't forget snow (and ice-salinity).

Martin

PS. While you are hacking seaice_advection.F why not include os7m?  
just kidding, but it should be simple.

On 10 Oct 2007, at 22:02, Dimitris Menemenlis wrote:

> Martin, I am getting more and more pressure from one of your former  
> AWI colleagues to implement OBCS for pkg/seaice.  In  
> seaice_advection.F I notice:
>
>> CML#ifdef ALLOW_OBCS
>> CMLC-     Apply open boundary conditions
>> CML          IF ( useOBCS ) THEN
>> CML           IF (tracerIdentity.EQ.GAD_HEFF) THEN
>> CML            CALL OBCS_APPLY_HEFF( bi, bj, k, localTij, myThid )
>> CML           ELSEIF (tracerIdentity.EQ.GAD_AREA) THEN
>> CML            CALL OBCS_APPLY_AREA( bi, bj, k, localTij, myThid )
>> CML           ENDIF
>> CML          ENDIF
>> CML#endif /* ALLOW_OBCS */
>
> Two questions:
>
> 1. Why do you need to do above twice?  Wouldn't once, after both  
> the X and Y direction loops have completed suffice?
>
> 2. Do you already have stubs available for OBCS_APPLY_HEFF and  
> OBCS_APPLY_AREA?
>
> D.
>
> -- 
> Dimitris Menemenlis <menemenlis at sbcglobal.net>
> 5056 Oakwood Ave, La Canada, CA 91011-2450
> tel/fax: 818-790-6735;   cell: 818-625-6498
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel