[MITgcm-devel] RE: 510x50 cubed sphere

[Alistair Adcroft]

You still need viscosity unfortunatley and because of the cube you're forced
to use Laplacian and not bi-harmonic.

Re: Sulu Sea, try adding this to PARM01

 hFacMin=0.1,
 hFacInf=0.1,
 hFacSup=5.,
# z - begin
#   nonlinFreeSurf=3,
# z - end
# z* - begin
    select_rStar=2,
    nonlinFreeSurf=4,
# * - end

and watch with wonder at the Sulu Sea and then send JMC a bunch of flowers!

A.
--
Dr Alistair Adcroft            http://www.mit.edu/~adcroft
MIT Climate Modeling Initiative        tel: (617) 253-5938
EAPS 54-1523,  77 Massachusetts Ave,  Cambridge,  MA,  USA

-----Original Message-----
From: Dimitris Menemenlis [mailto:menemenlis at jpl.nasa.gov] 
Sent: Monday, December 15, 2003 1:40 AM
To: Alistair Adcroft
Cc: 'Christopher N. Hill'
Subject: RE: 510x50 cubed sphere


Alistair, am I supposed to run the cube-sphere configuration with no
explicit horizontal diffusivity nor viscosity whatsoever?  I tried turning
on biharmonic, as we do for 1/4-deg integration, and it caused trouble at
cube corners.

With 20-s time steps, the 510x510x50 configuration crashes after three and
half days.  I am attaching a figure that shows sea-surface height shortly
before the crash.  A copy of the STDOUT file is in
/ocean/data4/dimitri/STDOUT.0000

Problem is in the Sulu Sea between the Philippines and Malaysia.  Seems to
be vertical CFL violation.  I will try to fix this tomorrow by smoothing
bathymetry in that region.  Any other suggestions?

D.

-- 
Dimitris Menemenlis <menemenlis at jpl.nasa.gov>
Jet Propulsion Lab, California Insitute of Technology
MS 300-323, 4800 Oak Grove Dr, Pasadena CA 91109-8099
tel: 818-354-1656;  fax: 818-393-6720