I think it could be better using a spectral element method based software, in order to model with sufficient accuracy the waves within the whole basin. You can try: SPECFEM3D, see http://geodynamics.org/cig/software/specfem3d/ or http://adsabs.harvard.edu/full/2002GeoJI.150..303K.
Another similar software will be soon available as GPL Opensource : SEM3D.