Here is a sample of what is being done at the LEGI with the MITgcm code. 
All simulations are 2D, the typical scale of our virtual tanks is 1m. 
The code is always used in non-linear, non-hydrostatic configurations. 
The viscoity is set to 10^-6 m^2/s and the water is temperature-stratified. 
The thermal diffusivity is set to 1.43 x 10^-9 m^2/s so that Pr = 700. 
By doing this, our stratification is equivalent to a stratification with salt, 
as in the laboratory experiments we are reproducing here.

1) Grisouard_wave_attractor: simulation of an internal wave attractor. The water 
is 290mm deep and the tank is 453 mm long on its longer horizontal side. 
The angle of the sloping wall is atan(2). N = 2.2 rad/s (constant), 
omega_forcing = 1.34 rad/s.

Work about to be published: "Numerical simulation of a two-dimensional internal 
wave attractor", Grisouard N., Staquet C. & Pairaud I., submitted to J. Fl. Mech.

(A recent review about wave attractors: "Wave attractors: linear yet nonlinear", 
 Leo R. M. Maas, Intl. J. of Bifurcation and Chaos (2005), Vol. 15, No 9 (2005) 
 2757-2782)

2) Grisouard_Horn_et_al_2001: reproduction of the following article: 
"The degeneration of large-scale interfacial gravity waves in lakes" 
by Horn D. A., Imberger J. & Ivey G. N., J. Fl. Mech. (2001), 434:181

More classical experiment reproduced numerous times, starting with a tilted 
pycnocline in a two-layer fluid. No forcing is applied so the system tries to relax.
In this case, this leads to the formation of interfacial solitons.

---
Contact: Nicolas GRISOUARD  email: nicolas.grisouard@hmg.inpg.fr
PhD Student - Team ERES
Laboratoire des Ecoulements Geophysiques et Industriels, BP 53
38041 Grenoble cedex 9 France
tel : +33 (0)476 825 037 - fax : +33 (0)476 825 271
---