[MITgcm-support] problem higher order advections schemes

[Martin Losch]

Hi,

I have discovered another odd thing about the advection. I mentioned 
that earlier (a couple of months ago), but have put this problem aside:

In a zonally symmetric channel (linear EOS, salinity is passive), which 
is initially stratified, with a baroclinically unstable front, but 
zonally symmetric, the stratification is eroded when I use a higher 
order nonlinear scheme, DST3 with flux limiting and 
multiDimAdvection=.true. The zonal symmetry is NOT broken. When I use 
the standard 2nd order central differences the stratification is 
preserved, which to my understanding is what should happen.

I observe the erosion of the stratification also with adv scheme 77 
(4th order), with multiDimAdvection=.true.

I put the set-up in sea.mit.edu:/data/mlosch/channel, or 
sea.mit.edu:/data/mlosch/head/channel, in case anyone would like to 
have a look at this (pleeeeease!!!).
A zonal jet forces the system (wind-stress). The setup works with 
genmake, beware of the .genmakerc in the code-directory.

Martin

Martin Losch // mailto:mlosch at awi-bremerhaven.de
Alfred-Wegener-Institut für Polar- und Meeresforschung
Postfach 120161, 27515 Bremerhaven, Germany
Tel./Fax: ++49(471)4831-1872