[MITgcm-devel] RE: Higher order dissip on the cube

[Dimitris Menemenlis]

>>  1) NaN's appearing in the monitor output

Alistair, I concur with Chris that the monitor NaNs occur for
single-CPU global_ocean.cs32x15 test but do not otherwise affect
results.

>> 2) An inconsistency between exch1 and ech2 when using higher
>> order dissipation

After fixing the the no-slip lateral boundary problem, the hi-res,
low-viscA4 cube configuration ran for an additional 70 days past the
Equatorial Seamount crash.  The new crash once again occurs near the
edge of a cube face, but not near a corner, violating w-cfl.  It's
possible but becoming increasingly unlikely that the edge location
of the crash is once again a coincidence.

Let me know when new exch2 code is checked in and I will test the
hi-res configuration one more time.

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