[MITgcm-support] Re : Water at the bottom of NA below freezing point

[Martin Losch]

Just a quick comment for Jeff:
On 8 Mar 2008, at 19:48, Jeff Scott wrote:

> I've also tried different advection schemes (besides 2). 33 and  
> especially 77 are not usable, as they radically affect the model's  
> temperature structure, being rather diffuse. I've also tried the  
> Prather/SOM

When you use one of the multidimensional advection schemes (e.g., 33  
or 77), make sure that you have "staggerTimeStep=.true.", otherwise  
these scheme give very unexpected results during spinup. If you use  
staggerTimeStep=.true., then advscheme 2 and 77 (which are both 2nd  
order), give similar results in coarse resolution with much smaller  
overshoots (according to my experience). As far as I remember, Dror  
is using that kind of parameter combination, so his results are  
either due to the very small but finite overshoots, that are still  
present in 33 or 77, or something completely different.

another comment on numerical diffusivity: Thomas Zweigle has carried  
out a nice comparison of numerical diffusivities of the various  
schemes in a very high resolution (DNS) simulation. Thomas, maybe you  
can share a figure (a small jpeg or png please)?

Martin