[MITgcm-devel] question regarding merging the shelfice and icefront packages

[Jean-Michel Campin]

Hi Ian,

In reverse order:

1) useRealFreshWaterFlux & ALLOW_AUTODIFF in shelfice_forcing_surf.F
 I think I might have added this STOP (but did not know it would be perceived as,
 I quote, "a giant STOP") just because I was lazy to figure out how to avoid some 
 recomputations (+ not knowing if it would be ever used).
 So, could be safely removed (once recomp. are fixed).

2) In case I want to use pkg/shelfice without lateral melting (since it's still 
 not clear to me if this is important):
  a) addMass involves 3-D array (-> 3-D diags that might not be coded yet ?)
     whereas surface forcing is just 2-D.
  b) addMass does not connect to KPP (and I would argue that it make sense), 
     but surface forcing does.
So, it would be nice to keep the surfaceForcingT/S version/option.

Cheers,
Jean-Michel

On Thu, May 11, 2017 at 11:44:00PM +0000, Fenty, Ian G (329C) wrote:
> Dear ice shelf/ice front concerned readers,
> 
> I've been working on merging the shelf ice and ice front packages.  In shelfice, mass fluxes are always handed through EmPmR and T and S tendencies are handled via surfaceForcingT and S.    In contrast, the icefront package does not treat mass fluxes and T and S tendencies in the icefront package are handled by operating on gT and gS directly.  I should also point out that in shelfice the T and S tendencies are handled by directly operating on gT and gS when SHELFICEboundaryLayer=.TRUE.,
> 
> I am wondering is there any reason why one couldn't or shouldn't use addMass to handle all mass tendencies and directly operate on gT and gS to handle all T and S tendencies that arising from melting and freezing irrespective of k level?
> 
> Is there something special about applying tendencies to the surface level besides the convenience factor having access to surfaceForcingT and S and EmPmR?  
> 
> Finally, Ou and I noticed a giant STOP in shelfice_forcing_surf.F when 'useRealFreshWaterFlux=.TRUE.' and ALLOW_AUTODIFF is defined (lines 96-108) that prevent meltwater mass fluxes from contributing to EmPmR.  Is this STOP necessary or an artifact from an earlier experimental setup?
> 
> Cheers,
> Ian
>  
> 
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel