Bivariate distributions of significant wave height and mean wave period of combined sea states

This work presents the results of the fit of three bivariate models to twelve years of significant wave height and mean zero-crossing period data of swell, wind sea components, and combined sea states from Australia. The Conditional Modelling Approach defines the joint distribution from a marginal distribution of significant wave height and a set of distributions of mean zero-crossing period conditional on significant wave height. The second model fits the Plackett model to the data, and the last one applies the Box-Cox transformations to the data with the aim of making it approximately normal to fit a bivariate normal distribution to the transformed data. The conditional model with a lognormal distribution for the significant wave height and lognormal distributions for the zero-crossing period gave the best fit for the total sea states and for the wind component. In case of the swell component the conditional model with a Weibull distribution to the significant wave height and a lognormal distribution to the mean zero-crossing period gave a relatively close fit to the data. (C) 2015 Elsevier Ltd. All rights reserved.