[Mitgcm-support] Re: .

[mitgcm-support at dev.mitgcm.org]

Hello Fabiana,

Sorry for not answering earlier.

Concerning the boundaries condition:
the default is "cyclic" BC both at the North-South and East-West boundaries;
imposing h(:,end)=0 (like in exp0) closes the domain at the Northern B and
because of cyclic condition, it is also closed at the Southern B.
So in fact, in exp0 the domain is closed on the 4 sides.

If you specify h(end,:)=5000, then the domain become really
cyclic in the E/W direction so you might expect a different behavior.

With no dissipation and a Crank-Nicholson time stepping:
 implicSurfPress=0.5,
 implicDiv2DFlow=0.5,
the external gravity waves should not be damped, even if the domain
is open (i.e. cyclic) in the E/W direction or N/S direction.
Is it the case ?

Jean-Michel