[MITgcm-devel] obcs_seaice_sponge.F

[Jean-Michel Campin]

Dimitris,

I found the Pb with multi-threading, but will go with 2 check-in
(in obcs_readparams.F), a quick one and one with more changes
related to threads.

And otherwise, I understand your point regarding input files, 
but in term of implementation, should be affordable, if using the
sponge option, to have few more 2-D mask and several 3-D arrays for 
the relaxation targets. It's just to fill theses with OBCS values
(and then EXCH + similar relaxation as done in rbcs).

Cheers,
Jean-Michel

On Wed, Sep 19, 2012 at 03:31:40PM +0000, Menemenlis, Dimitris (3248) wrote:
> Jean-Michel, I will wait for your multi-threaded fix before
> checking in obcs_seaice_sponge.F
> 
> The one-problem with 2-D/3-D rbcs relaxation fields is that they
> are impractical when you want to specify high-frequency boundary
> conditions for high resolution domains.  The input files would be big.
> For example, in Michael's regional 1-km PIG set-up, specifying daily
> rbcs over 40 years would require 23Nr-GB input files vs 112Nr-MB
> input files for obcs_sponge specification of same information.
> 
> >From an MITgcm user's perspective obcs_sponge is also convenient
> because it does not require any extra input files to be generated.
> I can see that obcs_sponge may not super-clean, for exotic obcs setups,
> complicated coastlines and bathymetries, and at corners, but we
> nevertheless use it a lot and it works well for simple-geometry hi-res
> regional setups, for example in applying 18-km obcs fields to a 1-km grid.
> 
> Dimitris Menemenlis
> 
> On Sep 19, 2012, at 7:56 AM, Jean-Michel Campin wrote:
> 
> > I think there are several issues, and the simplest way to fix
> > theses would just involve to  re-writte it in similar form as rbcs
> > (a 2-D mask + a 2-D/3-D field value to relax towards, both beeing exchanged), 
> > with this relaxation field set-up from the OBCS fields.
> > 
> > Now, it looks like my changes from yesterday are breaking the multi-threaded
> > tests. Will try to fix this soon, but might be better to wait until this
> > is fixed before make new check-in in obcs related code and experiments.
> 
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel