[Fwd: RE: [MITgcm-support] grid-scale noise in 1/4-deg model]

[Alistair Adcroft]

Dimitri,

My last email was missing a crucial detail!!!  Because you are using a
Gaussian grid you should use cosPower=2 along with the changes below.
Otherwise it will blow up.

Sorry about that ...

A.

-----Original Message-----
Yes, you need 
 #define COSINEMETH_III
 #define ISOTROPIC_COS_SCALING
in either CPP_OPTIONS.h or MOM_FLUXFORM_OPTIONS.h (if you use the former you
don't need it in GAD_OPTIONS.h). Better use CPP_OPTIONS.h so it's all in one
place.

Then my guess is you can push up viscA4 to around 1.E12 which should solve
you problems. I think 2.E12 will be beyond the edge of stability and you
might find 5.E11 to be sufficient but as I said before, find the upper limit
before choosing.

We will write a Smagorinsky scheme so that these issues are more easily
sorted out. Useful for the cube too.

A.
--
Dr Alistair Adcroft            http://www.mit.edu/~adcroft
MIT Climate Modeling Initiative        tel: (617) 253-5938
EAPS 54-1523,  77 Massachusetts Ave,  Cambridge,  MA,  USA