The effects of bathymetry on the process of wind-wave generation in the Persian Gulf have been investigated using a long-term hindcasting analysis. A 5-yr (200408) hindcasting procedure is first performed (denoted as
origin) by using the Simulating Waves Nearshore (SWAN) model and COAMPS wind field data. Two alternative scenarios, in which wave breaking ( noBrek) and depth-induced refraction ( noRefc) are deactivated, are also generated. By comparing the results from alternative cases with those from the ordinary cases, a 5-yr total energy deviation (TED) is calculated and presented as seasonal contour maps depicting the sensitivity to bathymetry. The results show only 2% TED found in noBrek, but up to 20% TED found in noRefc, an order of magnitude larger than depth-induced breaking. Similarly, the seasonal effects of refraction on the nearshore wave-propagating directions can be investigated by comparison between originand noRefc. The seasonal histograms of mean wave angle are plotted and discussed for three selected nearshore sites around Qatar, located at the western, northern, and eastern sides of the peninsula. The largest relative occurrence percentage and the largest peak shift ranging in 1020 can be found at the northern side, while the eastern side results in a weaker and more random distribution in the winter (the strong shamal season) due to the leeside location. The effect of fetch-limited wind-wave generation is also present at the eastern site, as remotely generated waves propagating eastward over a long fetch toward this site offset southward-propagating waves generated by dominant winds from the north over a limited fetch.