zero significant wave height at two corner grid cells in ROMS-SWAN coupled model

[xwu]

Hello,

I am not sure whether anyone has seen or reported a similar problem. I am running a ROMS-SWAN coupled realistic model. The model grid has three open boundaries (south, west and north), and the eastern boundary is closed. SWAN model receives realistic incoming wave forcing from a boundary condition (BC) file. This BC file provides directional wave spectra along the three open boundaries, including at the two grid cells at the southwest and northwest corner. The coupled model outputs the bulk wave parameter (including Hsig) during the 1-month study period.

Then I compared the output Hsig with the wave height reconstructed from the wave spectra provided in the SWAN BC file. I find that, along the three open boundaries, the output Hsig matches well with the wave height from the BC file, except at the two grid cells at the southwest and northwest corner. At these two corners, the output Hsig is all zero at each time step while the BC file provides non-zero wave heights. I do not know why zero Hsig occurs at these two corners.

My colleague is running an idealized ROMS-SWAN coupled model. He also noticed that zero Hsig occurs at the two corner grid cells.

thanks,
Xiaodong Wu

[jcwarner]

can you post this message here:
https://github.com/jcwarner-usgs/COAWST/issues
along with your swan input file.
thanks
j