[MITgcm-devel] calc_gw

[Jean-Michel Campin]

Hi,

I have a truly flux-form version of calc_gw that seems to 
work (no differences for cartesian-grid NH test).
I am going to check-in modified versions of calc_gw,
in several steps (so that we can track where the differences
come from):
step.1) no sidedrag (since it's wrong)
step.2) vertical grid factor (partial cell & non-uniform 
        vertical discretization)
step.3) horizontal grid factor

and will need to put back the correct sidedrag (step.4).

Jean-Michel