running momentum flux correction experiments

[njn01]

I am planning on running momentum flux correction experiments with
CCSM4, so I would like to add a wind stress perturbation (time-
dependent for one run and an annual mean for another) to the ocean, in
addition to the wind stress produced from the coupler. The idea is
the total wind stress will be mean(obs) + anomalies(model). It seems
the best way to do this is within the forcing_ws.F90 file, and that I
would have to add another case that is a combination of "analytical"
and "annual/monthly". Is this the best way to proceed?

 

[njn01]

I believe the order of the subroutine calls to setup the
surface forcing is

1. call ocn_import_mct (called from ocn_comp_mct.F90)

note that this will in turn call rotate_wind_stress, thus
setting SMF and SMFT

2. call pop_set_coupled_forcing (called from ocn_comp_mct.F90)

this call does not change SMF or SMFT

3. call set_surface_forcing (called from step_mod.F90)

4. call set_ws (called from set_surface_forcing)

note that set_ws WILL OVERWRITE the SMF or SMFT fields, if
ws_data_type is something other than 'none' in the forcing_ws_nml.

So, you need to make sure not to overwrite what the coupler already
sent, but add to them. You can simply use different array names. Also,
please make sure that the wind stress fields that you are planning to
add are already on the model grid, ROTATED, and in cgs units. In addition, as you said,
you need to add another case, e.g., "annual/monthly" and specify something
other than 'none' for ws_data_type. Also, this combination has not been
tested here at NCAR!