[MITgcm-devel] upcoming changes in seaice_growth.F

[Pierre Rampal]

Hi,

I have just one comment for now (see it below).

On May 25, 2011, at 3:40 PM, Gael Forget wrote:

> Hi Ian and co.,
> 
> below I report what Ian and I discussed at the ECCO2 meeting,
> which I just got started with seaice_growth.F rev.121 (see below).  
> 
> So, to complete the 'merging' process, we planned the following:
> 1) a few changes in the EVOLUTION branch
>        1.1) reduce the areaMin default
>        1.2) always include a_QSWbyATM_cover in QNET
>        1.3) use Ian's heff_star sqrt formula in heffActual
>        1.4) add the Winton (?) formula for flooding
> 	1.5) add a check for the areaMax range
> 2) remove code duplicates and restore the integrity of the two branches structure
>   (EVOLUTION: with Ian's codes // LEGACY: for backward compatibilty)
> 3) add somewhere in the code (SEAICE_OPTIONS.h or seaice_check.F may be)
>   the combination of CPP options that is advised for adjoint runs.
> 
> Once we complete that, I would also like to open up a discussion on the seaice_growth
> diagnostics, which I think can also use some review/clean up(?)/complements(?).
> 
> In seaice_growth.F rev.121, I did 1.2) and updated the global_ocean.cs32x15 results
> accordingly. I also got started with 2) and updated the 1D_ocean_ice_column
> SEAICE_OPTIONS.h accordingly (no change in results).
> 
> With regard to 1.2) I felt it was best not to do it in the LEGACY branch,
> since it would affect backward reproductibility. Makes sense, right?
> 
> With regard to 2) I left the FENTY_AREA_EXPANSION_CONTRACTION part for later,
> since 1.3) is a pre-requisite. Ian, you were going to check that using 0.05^2 rather 
> than 0.1^2 in the heff_star sqrt formula is fine in adjoint mode. Is it?
> 
> With regard to 1.4) I argued at the time that we may as well replace the
> flooding formula in the EVOLUTION branch (only). But may be we should
> keep both and add a CPP option. Opinions?
> 
> With regard to 1.1) Ian argued at the time that we don't need
> areaMin (old A22) to be a run time parameter. Opinions?
If I understand well, if areaMin is no longer a run time parameter, it will be impossible to change his value from data.seaice, is it right? If it is, I'm not sure that it is a good idea as it would be better to look first at the sensitivity of models results to a change in this value. To do that with convenience, it seems to me important to be able to change its value from data.seaice. Indeed, I would expect some important changes in air-sea fluxes in coarse resolution setup by changing this value. As you mention below, areaMin is currently set to 0.15, which is somehow very large. So, before having tested the effect of changing it from 0.15 to 10^-5  or any other value, I would keep it as a run time variable. 

> In any case we would leave the 0.15 default for the LEGACY branch.
> For the EVOLUTION branch we talked about 10^-5 if I remember right.
> Cheers,
> Gael
> 
Pierre


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