A simple and robust method for calculating return periods of ocean waves

Abstract

A new method is introduced for combining the long-term distribution of sea states with the short-term distribution of individual wave or crest heights, conditional on sea state. The method uses a Monte Carlo approach to simulate random realisations of the maximum wave or crest height in each sea state. A peaks-over-threshold analysis is conducted on the random maxima in each sea state in order to estimate the long-term distribution of individual wave or crest heights. The new method is significantly simpler than existing methods such as the equivalent storm approach, requires fewer assumptions and has similar computational times. The new method is applied to a 35 year dataset of wave buoy measurements and is shown to produce almost identical estimates of return values of individual crest heights to the equivalent storm method.