[MITgcm-support] rbcs advice

[Menemenlis, Dimitris (3248)]

Ah get it.  One last suggestion.  Probably also not useful for your problem, but just in case.

Chris, Gael, Ayan, and I have set-up and integrated a 1-48th global simulation that includes tidal forcing.
We save hourly output of full 3D U/V/T/S/Eta, so these could be used as boundary conditions for
"replaying" some small portions of domain.  You would not need sponge layer because the hourly
obcs boundary conditions would naturally take care of most radiation issues.

On Mar 4, 2014, at 6:38 AM, Jody Klymak <jklymak at uvic.ca> wrote:

> Sure, I've used obcs_sponge a lot as well. 
> 
> The problem for baroclinic flows is that the sponge is usually a good fraction of a mode-1 wavelength, so you really want incoming waves to be able to propagate through the sponge.  Using rbcs works pretty well for this, but comes at significant i/o expense.  
> 
> It would be nice if there was a version of obcs_sponge that read in the whole sponge, and didn't just use the boundary condition as the sponge forcing. 
> 
> Thanks,   Jody