Weird Inf in the first time step, coupled with ROMS and SWAN

[donghua]

Hello, all

I have 3 versions of grids which have 22m resolution resolving surf zone.
The grids are slightly different. Their regions are a little bit different, but all include my region of interest.
When I run the grids coupling Swan and Roms, they show very weird Inf values within the grids.
All three grids have similar issues. Inf values pop up in different location for the three grids.

The result is in the file below:
https://drive.google.com/file/d/0B9c4DF ... sp=sharing

I check all my input. The input forcing files look OK.
When I recompiled and used my coawstM, I would have a slightly different blowing up result.

I've got stuck here for a long time. Can anyone give me some suggestion?

Thank you in advance.

Donghua

[jcwarner]

there could be a lot of different things going on here. We need more info to provide some guidance.
Can you run the grids with just swan and then just roms successfully?
then try to couple them and see what happens. Are the Inf values at the wet/dry interface, along an open boundary, ...
If you turn off some of the physics do they go away? Try a couple of things to narrow down the possibilities.
-john

[donghua]

Thank you, John

Yes, when I run ROMS-only or SWAN-only, both cases can run successfully. But when I run ROMS-SWAN coupling case, I will have Inf value in the sea. Using different versions of coawstM will give me different Inf-NaN maps.

I provide bulk forcing, initial forcing, and boundary forcing. But I tried to turn off all these forcing and only provide the grid, I will have the same issue: the Inf values will show up in the sea.

I am going to try the newest version of coawst code. I used the newest code before, but it gave me very high temperature near the coast, like 60 degree C. But I will still try it.

Any suggestion?

[jcwarner]

can you turn off any cpp options,
like SSWBBL, WEC, Sediment, etc.