Coastal Wave-Height Statistics during Hurricane Ike

The statistical behavior of wave heights obtained from short-term (i.e., 30-min) wave records during Hurricane Ike was investigated. Spectral data from seven temporary pressure gauges and one National Data Buoy Center (NDBC) buoy moored off the coast of Texas, covering a 12-day period, were used. The wave heights, based on time series reconstructed from the spectra, showed a very good fit to the form of the Rayleigh distribution with Hrms as the parameter of the distribution, whereas the use of the spectrally derived parameter 8m0 led to overprediction of the wave heights. Incorporating the spectral band width (α) in the Rayleigh distribution reduced this overprediction. The empirical two-parameter Weibull distribution was also found to represent the wave-height distribution well, but with parameters different from a = 2.126 and b = 8.42. The Tayfun model, used as a representative of Rayleigh-like models, provided the best fit. When converting model-derived significant wave heights to other quantities, such as H1/10, H1/100, and Hmax, the standard Rayleigh coefficients are often used in practice. These led to considerable errors in the present analysis, which also shows that the use of correction factor or α can enhance the results. However, direct estimates of these quantities from a time series simulated using the model-derived spectrum may be an even better alternative to using the standard Rayleigh coefficients.