[MITgcm-devel] tags and TAMC-compatibility

[Dimitris Menemenlis]

Jean-Michel, I am trying to restart a CS510 integration with updated  
code and I get the following peculiar behavior.

If I compile the code with Olx=Oly=5, as before, or with Olx=Oly=8 as  
is needed for advection scheme 7, the code blows up on first time step  
with "STOP in CALC_R_STAR : too SMALL rStarFacC !".

If I compile the code with Olx=Oly=4, it runs without crashing.

Could this behavior be related to your item 2) below?

For a CS32 configuration, however, the code runs OK for any value of  
Olx=Oly>=4, and it converges to exactly the same solution for values  
of Olx=Oly>=8, as you recommended that I use for the cube sphere with  
advection 7.

D.

Dimitris Menemenlis <menemenlis at jpl.nasa.gov>
Jet Propulsion Lab, California Institute of Technology
MS 300-323, 4800 Oak Grove Dr, Pasadena CA 91109-8099, USA
tel: 818-354-1656;  cell: 818-625-6498;  fax: 818-393-6720

On Jun 15, 2008, at 2:22 PM, Jean-Michel Campin wrote:

> 2) I tried advection scheme 7 but I am getting NaN at the 1rst
> time-step. I have a fixed for this, and could probably check-in
> this one (tomorrow morning ?) before you make the 2 tags,
> since it's not TAF sensitive.