search

COSIMA / cice5

Clone of The Los Alamos sea ice model (CICE) with ACCESS drivers. See https://github.com/CICE-Consortium/CICE-svn-trunk/tree/cice-5.1.2
4 stars 13 forks source link

Provide namelist option surf_vel_scale in surface_flux_nml to use abso… #30

Closed aekiss closed 5 years ago

aekiss commented 5 years ago

…lute (rather than relative) wind for all surface fluxes. https://github.com/OceansAus/access-om2/issues/137

@russfiedler, @nichannah does this look like it will work? I haven't checked that it will compile...

russfiedler commented 5 years ago

@aekiss No, the namelist variable is private to the surface_flux module. I wouldn't think it would compile. I think it would be better to make the change here https://github.com/OceansAus/cice5/blob/6f8f1edf1b775b17067a11f6b4642d7cfdb6ca9c/drivers/auscom/surface_flux_mod.F90#L445-L446

Just make an if block to calculate udif as either the difference or -u_atm or -v_atm. Another way would be to multiply U_surf by a 0/1 mask scalar as appropriate.

aekiss commented 5 years ago

Ah yes that was dumb, I was wondering about the scope of the namelist variable.

I'd initially suggested we modify those lines in surface_flux_mod.F90 (445 to 446 in 6f8f1ed, 440-441 in master) but you suggested we set the surface velocities to zero in the atmospheric flux calculation instead https://github.com/OceansAus/cice5/blob/master/drivers/auscom/cpl_forcing_handler.F90#L976-L977 as this affects all the turbulent fluxes not just the wind stress.

russfiedler commented 5 years ago

Oops, I'll have to delete that comment from the record!

Actually, I think it's much better, in the MOM/CICE setup, to supply the correct surface velocities to the surface_flux module and let it decide what it wants to do with them. The same with CICE. If it wants the ocean velocity then it can choose whether to use it or not. MOM/SIS is a bit different since it uses the data_override facility to do this sort of thing.

aekiss commented 5 years ago

Don't worry, it will disappear from Slack soon enough... unlike my bad coding here. Does this update look any better?

russfiedler commented 5 years ago

Looks good to me. Having a quick test run (or at least compile) example would be nice. One for the TWG...

aekiss commented 5 years ago

This didn't compile (can't have an IF inside a WHERE) so to avoid the IF I've reimplemented it in terms of a surface velocity scaling factor surf_vel_scale instead of a logical. This now compiles and I'll try out a test run. Any comments on this approach?

russfiedler commented 5 years ago

I should have spotted that. You need to break the WHERE construct apart and have two branches.

I think a clearer approach would be to to use the logical but set the value of surf_vel_scale in the code. It's trivial but makes it obvious to the user what needs to be set. I don't think there would ever be a reason for fractional values.

aekiss commented 5 years ago

OK how's this? It compiles happily. I've also updated the documentation on w_atm - I think I got that right.

russfiedler commented 5 years ago

Looks good. Yeah, I think the documentation for w_atm was incorrect. The original should have said relative not absolute.

aekiss commented 5 years ago

Yes, that's what I thought. OK I'll give it a test run when I get a spare moment.

aekiss commented 5 years ago

I've completed test runs - it seems fine. @russfiedler or @nichannah could you merge it please?