[MITgcm-support] noise again, coarse resolution

[Martin Losch]

Hi there,

I am trying to reproduce the results of
Pacanowski and Gnanadesikan 1998: Transient response in a z-level  
ocean model that resolves topography with partial-cells. MWR, Vol  
126, No 12, December 1998.
with the MITgcm (checkpoint58b_post). I tried to copy the  
configuration as closely as possible (running into all sorts of  
trouble, there is a typo in the equation for lambda and the density  
perturbation may be too large), but the model becomes unstable after  
50 days or so (visible in a peak in kinetic energy and a deltaX  
pattern in temperature). I tried free slip, no slip, JamartWetPoints,  
staggering the time step (but not CD-scheme), different advection  
schemes, increased friction by factor 10, timesteps between 3600 and  
900 seconds, but nothing improves the situation. I have attached my  
configuration (code and input directory, need to regenerate input  
files with matlab), which is very simple and takes about 60sec to  
run. Can anybody see what I am doing wrong?

Martin
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pg98.tgz
Type: application/octet-stream
Size: 3991 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20060502/c0c80eda/attachment.obj>