[MITgcm-devel] Unrealistic low SST with SEAICE_GROWTH_LEGACY undef

[Martin Losch]

I think that we should not, because then by default, potential problems are masked.

M.

On Feb 7, 2012, at 3:42 PM, Dimitris Menemenlis wrote:

> Martin, so shall we change the default SEAICE_gamma_t_frz to be different (smaller) that SEAICE_gamma_t
> (while the legacy code is still available)?
> 
> I agree with your second paragraph and that we should aim to get rid of as much as possible of the legacy
> code and to consolidate/reduce the available pkg/seaice and pkg/thsice options.
> 
> Dimitris Menemenlis
> 
> On Feb 7, 2012, at 5:54 AM, Martin Losch wrote:
> 
>> you are right: I am using SEAICE_gamma_t_frz=SEAICE_gamma_t, so I could try that. 
>> 
>> But my point is that with the non-legacy code this is not even necessary. The legacy code somehow produces very low temperatures, that then require some ad-hoc fixes (although I can think of a justification for having a very short gamma_t_frz). I think that we should be able to run the model without these fixes that cover up all sorts of problems (I am not saying that we should fix the legacy code, but that we should make sure that with the new code these things don't happen again. And I am afraid that they will, if we use some unclear parameterization as a rule).
> 
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel