[MITgcm-support] Obcs problems

[Rutkiewicz M.B.]

Hi Martin,
I have done just that after I sent my email, so it's great that you are suggesting the same thing! I have just shut off the rest of the boundary with a wall of about 2/3rds of the channel depth and it seems to work great.

Best wishes,

Maria

________________________________________
From: mitgcm-support-bounces at mitgcm.org [mitgcm-support-bounces at mitgcm.org] On Behalf Of Martin Losch [Martin.Losch at awi.de]
Sent: 02 July 2008 14:36
To: mitgcm-support at mitgcm.org
Subject: Re: [MITgcm-support] Obcs problems

On 1 Jul 2008, at 18:18, Nicolas Grisouard wrote:

> useOBCS = .TRUE. in /data.pkg/  needs =>  #define ALLOW_OBCS in /
> CPP_OPTIONS.h/  needs  =>   -DALLOW_OBCS in the Makefile.
which is done automatically by inserting the a line with "obcs" into
packages.conf, which is exactly, Maria's (correct) way of doing it.

Maria, your initial question:
OB_Jnorth=30*-1., is correct, but I assumes that the entire water
column belongs to the open boundary (and appropriate conditions are
provided). There is no capability in the code to restrict this to
the, say, 3 surface layers.

What do you want the model to do below the 3 surface layers along the
open boundary? You cannot just let it hang on ... You could insert
topograph (a wall), to close the domain below k=3.

Martin

_______________________________________________
MITgcm-support mailing list
MITgcm-support at mitgcm.org
http://mitgcm.org/mailman/listinfo/mitgcm-support