Simple PC based significant-wave model

Modifications are made to the wave prediction method of the U.S. Army Corps of Engineers Shore Protection Manual (1984), to allow for spatially and temporally heterogeneous windfields, and to predict a time-series of outputs of end-of-fetch significant wave height and period. The model will run on a P.C. and is designed to provide a quicker and more versatile method of `significant wave' forecasting than the manual methods. Inputs are derived from surface atmospheric pressure charts. A hindcast is performed in the North-Atlantic to test end-of-fetch values against those measured by a wave-buoy. Predicted values of Hs are within 0.5 m of those recorded, and maximum values occur within 6 hrs. Maximum values of Ts are to within 1.5 s, and also occur within 6 hrs. Possible error sources include interpolation of the inputs and inherent overprediction of the Shore Protection Manual method.