Is my bottom topography the reason for my blow up?

[cflanary]

I have my grid shown below of the central east Florida shelf. I'm trying to run 214 days and it runs 154 days and then becomes unstable and blows up in a few time steps. My 3d timestep is 150s and my 2d is 10s. When I check the restart file, the temperature and zeta values, at the Ci,Cj,Cw index point from my log file, are too large or too small.
I have some small areas of land in my grid, and the blow up point is in the southeast area of my grid, where it is shallow at the Bahamas.

I then tried a run with all the same forcing and timesteps with a flat bottom of 900m and no land mask, and it ran the whole 214 days.

Does this absolutely point to my steep bottom topography as the cause of the blow up?

Thank you,

Chris

[kate]

What are your rx0 and rx1 values? It should be in the ROMS output.

How many gridpoints do you have here? I can't quite see your grid, but it also doesn't look completely filled with grid lines. You might be better served by having more grid points in the east-west direction just to resolve the slopes better.

Personally, I would start by making my corners closer to 90 degrees and making sure the grid truly is orthogonal. Did you get the Seagrid mex files compiled?

[cflanary]

Hello Kate,

My rxo is 0.69 and my rx1 is 10.4. I tried to smooth with the smooth_bath Matlab file but it the values wouldn't go any lower. I have 78 in the I direction and 248 in the J and 20 vertical layers.

I did get the mex files compiled for seagrid.

[kate]

0.69 is at least twice as big as you want that thing to be. There's more than one way to lower it:
* Add more horizontal resolution
* Pick a deeper hmin
* Change your smoothing algorithm