Failure when change the grid in upwelling case

[ac412]

dear all,

When I try to compile my own climatology case, I created the grid file, clim file, initial file, bry file and forc file, but I always met the problem of 'blowing up' after 2-hours in the model time.

Then I try to test my topography in the upwelling case. So I keep everything the same with upwelling but add the grid file in *.in file and define the CURVGRID and MASKING, and also undef the ANA_GRID in *.h file. The compile is successful, but the calculation is blowing up within 2 time steps. Does it mean my topography is not smooth enough? Or something else could lead to the failure?

Many thanks~

[ptimko]

If the model is blowing up with 2 time steps it sounds like you're violating the CFL criteria. Try using a smaller time step.

[ac412]

If the model is blowing up with 2 time steps it sounds like you're violating the CFL criteria. Try using a smaller time step.

Thank you for your suggestion. However, it doesn't work. No matter I use the smaller time intervals (10s) or a fine resolution grid file, it blows all the time Q_Q. Besides, the rx0(=0.99) and rx1(=17) are too huge. How could I find a way to diminish them?

On the other hand, when I use the climatology force, the rx0(=0.2) and rx1(=6) are fine, but the results are still blowing up. Should I change the topography (I'm using the ETOPO1) ?

Thanks~~

[ptimko]

Something must be mismatched. I'm not sure I understand why your grid stiffness ratios are changing if you are using the same bathymetry.

For your own climatology case you are trying to set up. Have you check the coastline for single point seas? If it's blowing up near the boundary you may want to check if you have any land points within 3 or less grid cells from an open boundary. If you have water along the boundary you should probably have at least 3 grid cells of water points before you have any land points in that row or column.