Long-term wave climate risk analysis

Wave climate risk analyses must balance the information available from short data records with the application requirement of predicting rare events. The basis for long term risk analysis is typically the series of annual maximum events, rarely as long as 20 years. The triple annual maximum series, the 3 largest maximum monthly significant wave heights in each climate year, is used here to supplement the sparsity of the annual maximum data. Results from order statistics enable development of a maximum likelihood method for fitting a statistical distribution for the annual maximum event to the triple annual maximum data series and design events at the desired average recurrence intervals are then estimated as quantiles of the fitted distribution. The enhanced reliability of these estimates is confirmed by numerical experimentation.; Duration of sea state intensities is of consequence where there are risks due to high seas that persist for hours or days. Application of the triple annual maximum method to one hour significant wave height measurements is extended to runs of significant wave heights continuing over longer durations.; Intensity-duration-frequency curves from surface water hydrology are adopted for organization and presentation of the results of triple annual maximum analysis for sea state intensities of varying durations. With significant wave height as the intensity measure and average recurrence interval giving the frequency, a family of intensity-duration curves is constructed, from which magnitudes of estimated design events can be read with ease.; Confidence intervals provide a necessary sense of the uncertainty associated with design event estimates. A bootstrap procedure for estimating confidence intervals on triple annual maximum design event estimates is recommended, and its performance is validated by numerical experiments.; Wave climate risk analyses, comprising triple annual maximum distribution fits, intensity-duration-frequency curves, and bootstrapped confidence intervals on selected design events, are given for five National Oceanic and Atmospheric Administration National Data Buoy Center sites in United States waters.; These procedures and presentations, used routinely, are of value wherever risks due to sea state intensity must be addressed, Other geophysical risks are amenable as well to analysis by the same methods.