[Mitgcm-support] Help in Monterey Bay simulation

[mitgcm-support at dev.mitgcm.org]

Hi

I start to simulate the flow in the vicinity of Monterey Bay using MITGCM.
However, I don't think I get correct result. My test is using a coarse
grid resolution 40*80*21 layer(Zon*mer*depth). The run is forced by a
constant surface wind and realistic initial and boundary condition. The
bottom topography and geometry are also obtained from realistic data.
I got some disorder horizontal velocity in many layers, particularly near
the coastal geometry. Huge velocity field is generated near the boundary.
The horizontal velocity fields are illustrated on
http://fluid.stanford.edu/~yhtseng/research/monterey/index.html

The velocity field even points toward the boundary at some locations. I
suppose this problem is related to the boundary topography treatment.
However, I use partial cell and set hFacMin=0.15. This should smooth out
the boundary. Should I try other methods? I turn on "no_slip_bottom" and
set  bottomDragLinear=5.E-4, bottomDragQuadratic=5.E-3. This doesn't help.
I also turn both GM/Redi and KPP schemes on and try to increase the
turbulent viscosity. But no improvment either.

I am not very experienced with realistic geometry simulation. Please feel
free to give my some insight or suggestions if you know what's going on
here and how to fix this large velocity field near the topography. I do
appreciate your help!

Yu-heng