[MITgcm-devel] problem with 59c

[Dimitris Menemenlis]

Martin and Jinlun, thanks for help.  I am not sure that it is a zmin problem. 
There's something else going on ... and I have a feeling that when I find out I 
will feel very dumb for having wasted so much time on it.  The arctic CS510 
domain with closed boundaries is able to run stably and normally with everything 
almost exactly the same as in the global domain, with and without dynamics, with 
and without realfreshwaterflux.  Changing zmin to 4e8 in latest code did not fix 
problem.  So I am confused.  I will keep working on it and let you know what I 
find.  Dimitris