Subject: Re: appropriate namelist variables for other resolutions than t42.
From: Jeffrey Yin (jyin@atmos.washington.edu)
Date: Wed Aug 27 1997 - 17:25:09 MDT
Hello Wei--
I'm not sure if this is something that comes with the CCM3.2 or
if it's something that someone set up here at UW, but I have this program
called 'build.namelist' that sets the diffusion constants for different
resolutions as follows:
switch ( $CCMRES )
case T5:
set dif2 = 2.5e7
set dif4 = 1.e18
breaksw
case T21:
case T31:
case t31x15:
set dif2 = 2.5e5
set dif4 = 2.e16
breaksw
case t31x15x10:
set dif2 = 2.5e5
set dif4 = 2.e16
breaksw
case T42:
set dif2 = 2.5e5
set dif4 = 1.e16
breaksw
default:
exit 3
endsw
As far as timesteps go, DTIME = 1200. is probably safe; someone here ran
the model at T31x15 resolution with 10 levels with DTIME = 1800., but the
model blew up when I tried to run it at T31x15 resolution with 18 levels
with this timestep, so I had to go back to DTIME = 1200. I'm not sure
what EPS is either, but I haven't changed it, and haven't noticed any
problems.
I hope this helps. Good luck!
--Jeff--
Jeffrey Yin INTERNET: jyin@atmos.washington.edu
685-9524 UUCP: uw-beaver!atmos.washington.edu!jyin
Dept of Atmospheric Sciences
University of Washington, Box 351640
This archive was generated by hypermail 2b27 : Thu Jun 01 2000 - 09:11:24 MDT