[MITgcm-support] Uniform diffkrfile

[Geoffroy Aubry]

Dear MITgcm support list,

I am running an experiment using the 3D diffusivity file, and I've got 
some problems : I have some strong velocities when I use the diffkrfile.
So I tried to see what happened if using a uniform diffkrfile.

When I try to run an experiment with the following parameters :
  viscAh=0.1,
  viscAz=0.1,
  diffKhT=0.1,
  diffKzT=0.,
  diffKhS=0.1,
  diffKzS=0.,
and a diffkrfile with a uniform 1.E-5 m2.s-1, I get some instabilities 
very quickly.

If I try with :
  viscAh=0.1,
  viscAz=0.1,
  diffKhT=0.1,
  diffKzT=1.E-5,
  diffKhS=0.1,
  diffKzS=0.,
and no diffkrfile, the situation is stable

And if I try with
  viscAh=0.1,
  viscAz=0.1,
  diffKhT=0.1,
  diffKzT=1.E-5,
  diffKhS=0.1,
  diffKzS=1.E-5,
and no diffkrfile, the situation is no more stable.

I did try to change the code to see if it had something to do with the 
salinity (which is 35ppm everywhere in my experiment), but it doesn't 
seem to be the good option.
On this web page, you can find some more information about my 
configuration (pictures).
http://perso.crans.org/gaubry/dossier/stage/index.html#bug

The question is : do you know how to use the diffkrfile so that when it 
is uniform you get the same than when you change the diffusivity for the 
whole map? Once done, I suppose that I will be able to use it non uniform...
Thank you for your help,
-- 
Geoffroy Aubry