[MITgcm-devel] seaice+thSIce changes

[Martin Losch]

Hi again,
I have decided to go for a compromise: I checked in everything that  
does not, by default, change the results of pkg/seaice (and certainly  
not those connected to pkg/thsice).

So neither the evap*(1-area) fix, nor Jinluns fixes are not included  
(which we probably want to have in growth.F as default). I sent  
Dimitris "my" growth.F (with theses fixes) for him to try out.

Martin

On 1 Nov 2006, at 09:23, Martin Losch wrote:

> Hi Jean-Michel,
>
> now things start to get difficult  for me as I have implemented  
> snow advection for the seaice-pkg (without thsice), but as this  
> would change everybodies results (together with the evap and snow  
> fixes in growth), I held it off until Patrick has confirmed that it  
> does not destroy his ecco run.
>
> Therefore it would save me a lot of work, if we could decide,  
> whether I am allowed check in my pkg/seaice soon.
>
> Martin
>
> On 1 Nov 2006, at 04:23, Jean-Michel Campin wrote:
>
>> Martin,
>>
>> I started to make changes in seaice advection (pkg/seaice)
>> of thSIce variables, but realized that more EXCH calls
>> are needed to implement correctly the advection of Qice
>> (instead of Qice*hEff as it is now).
>> I left this piece of code within #ifdef THSICE_NEW_ADVECT
>> (undef in pkg/seaice/seaice_advdiff.F).
>>
>> Also, I am going to try to enforce a minimum ice volume (or hEff)
>> together with a minimum ice Area after the advection.
>> (lines > C jmc: new version: > C jmc: comment out the old version:
>> in pkg/seaice/seaice_advdiff.F ); it might prevent
>> some random blow-up (as the one you reported earlier ?).
>>
>> I also moved around the diffusion (mainly because it's
>> a pain to change 3 S/R, seaice_calc_rhs, seaice_advdiff &
>> seaice_advection instead of only 2, every time I had to add an
>> argument to SEAICE_ADVECTION), disable it for hSnow & Qice
>> (since we still need to use the 1rst Order upwind scheme ...),
>> but should be OK (I did short tests, both with & without thSIce,
>> and nothing has changed).
>>
>> Jean-Michel
>> _______________________________________________
>> MITgcm-devel mailing list
>> MITgcm-devel at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-devel
>
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel