[MITgcm-devel] upcoming changes in seaice_growth.F

[Martin Losch]

Hi Patrick,

to begin with, I do not understand why there has to be a minimum ice area to begin with, except where area is used to compute hactual (which is now for some reason heffactual, another interesting evolution of names). As far as I understand, the entire discussion circles around this (mis-?) understanding that a minimum area needs to be enforced somehow.

Let me say it again: The only place where area needs to be capped is where is used in the denominator to compute the actual ice thickness. areaMin (formerly a22) was just not a good name for that (and I take all the blame for the name). There is no need, from both the physical and the numerical point of view, for the second constant "area_floor" and everything that is associated with it. Ian suggests "area_floor", etc, just to keep the modification that Gael (?) put, but make the inactive by default. If we all (including Gael) agree that it is not necessary, we can remove this part altogether.

I suggested to have new names and retire areaMin, in order avoid further confusion ... 

Martin
On Jun 15, 2011, at 4:00 PM, Patrick Heimbach wrote:

> 
> Hi Martin,
> 
> sounds ok to have two separate variables,
> but I don't understand, why
> 
> 1. area_floor = 0. by default (which defeats its purpose, doesn't it)
> 
> 2. the AREA = max(area_floor, AREA) code needs to be activated with a
> #define SEAICE_ENFORCE_FLOOR_ON_AREA
> 
> I would have assumed the opposite in both cases.
> Something important I am missing?
> 
> We are discussing about almost nothing (or rather 10e-5),
> so best to resolve asap to get on to things > 10e-5.
> 
> Cheers
> -p.