[MITgcm-support] Re: Problems with Leith

Chris Hill cnh at mit.edu
Mon Sep 13 21:23:21 EDT 2004


Note - as an alternative someone could implement a Smagorinsky style
dissipation. 
So far no "developers" have managed to do this.

Chris

> -----Original Message-----
> From: mitgcm-support-bounces at mitgcm.org 
> [mailto:mitgcm-support-bounces at mitgcm.org] On Behalf Of 
> Alistair Adcroft
> Sent: Monday, September 13, 2004 8:44 PM
> To: Raffaele Ferrari
> Cc: MITgcm-support at mitgcm.org
> Subject: [MITgcm-support] Re: Problems with Leith
> 
> Raf,
> 
> In our usual fashion we added Leith as a test of user 
> abilities. Included with Leith is the special feature that if 
> you are not using vectorInvariantMomentum=.true. (in PARM01) 
> then the model fails to tell you that you can't (yet) use 
> Leith with the conservative form of the momentum equations. 
> So far no users have passed the test. :-)
> 
> Switching to vector invariant equations should fix this behaviour.
> 
> A.
> 
> Quoting Raffaele Ferrari <rferrari at MIT.EDU>:
> 
> > Dear Alistair,
> > 
> > how is life in Princeton? Do you miss MIT and the long commute?
> > 
> > We (Baylor and I) have a question about the Leith viscosity. We are 
> > trying to switch to Leith the run of the high-res front for 
> the CPT. 
> > We compared a low res run with a constant biharmonic 
> viscosity with an 
> > equivalent run with Leith (biharmonic, vector-invariant 
> form). We set 
> > the Leith constant (ViscC4Leith) to be 0.1 and 1 (and 5e7 
> by mistake).
> > We find that the Leith runs are stable but with a lot of grid-scale 
> > noise, no matter what the dimensionells constant is. Are we doing 
> > something wrong?
> > 
> > Thanks,
> > Raf
> > 
> 
> 
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://dev.mitgcm.org/mailman/listinfo/mitgcm-support
> 




More information about the MITgcm-support mailing list