[Mitgcm-support] aim on CS

[mitgcm-support at dev.mitgcm.org]

Chris,

I think we're ready for the big run.  J-M has convinced me
of needing to filter (U,V) rather than (U*,V*) with orography
so he has a modified version of forward_step.F for you.

The following data.shap works for hs94.cs-32x32x5 on xolas0 but
with -fast removed from Makefile.

% cat data.shap
# Packages
 &SHAP_PARM01
 nShapT=4,
 nShapUV=4,
 Shap_funct=20,
 &

which is currently running in
/d11/adcroft/pre38.may7/verification/hs94.cs-32x32x5/input

This has deltaT=600 <= 720 is unstable with this Shapiro.

There's also a modified shap_filt_uv_s2g.F in the above dir
but it shouldn't be needed. This was left over from my debugging
the SUN optimization problem. I expect removing -fast is all
you need to do.

A.