Hey guys, when I was running a modelling with time step of 5 sec, it gave crash like a problem of negative depth. So I turned down time step to 1 sec, and it ran agian, but it would need much more time to finish running. Can I make the time step back to 5 sec, when it passes the point where it got crashed, so as to get the modelling sooner to be finished?
Try to update the bathymetry with a finer interpolation data. It might help to stabilize your model.
Mizu Kun, thanks for your sugesstion, I will consider to update the bathymetry then.
Are you using the horizontal momentum diffusion options?What kind of boundary conditions are in your model?