[Mitgcm-support] RE: horizontal viscosity

[mitgcm-support at dev.mitgcm.org]

Hi Martin.,

 It is something that other people do. I don't see a problem with
checking it in to the main branch - although Jean-Michel may have an
opinion! The discrete formulation ( div(v(grad)) ) is such that it
should just work. If you do go for checking it in on the main branch
including code that plots the values at start, does initialisation
neatly so that someone can subsequently customize the function etc...,
would be nice. If that's more than you have time to do then an alternate
is to have a custom experiment that shows the bits of code you did hack
to make the changes you needed - of course then it is hard to keep the
custom hack code up to date.

Chris



> -----Original Message-----
> From: Martin Losch [mailto:mlosch at awi-bremerhaven.de] 
> Sent: Thursday, April 17, 2003 4:10 AM
> To: support at mitgcm.org
> Subject: horizontal viscosity
> 
> 
> Hi,
> 
> are there any plans to replace the so-far constant horizontal 
> viscosities, viscAh and viscA4, by 2D or 3D fields? So that one could 
> e.g., scale the coefficients with variable grid size, or implement 
> something like a Smagorinski-type viscosity? If I were to 
> code this up, 
> would it make sense to go for (global) 2D fields 
> Ah(i,j,bi,bj) or should 
> I immediately introduce 3D fields Ah(i,j,k,bi,bj)? Is this 
> anything of 
> interest to anyone else or is it just me (in which case I 
> would do as I 
> please and not check anything in (o: )?
> 
> Martin
> 
> -- 
> Martin Losch
> Alfred Wegener Institute for Polar and Marine Research
> Postfach 120161, 27515 Bremerhaven, Germany;
> Tel./Fax: ++49(0471)4831-1872/1797
> 
>