Modelling distributions of significant wave height

The paper discusses the use of distribution functions in the description of the long-term behaviour of significant wave height data. It is shown by an example that the usual goodness-of-fit diagnostics do not ensure correct predictions of return values based on arbitrary statistical models. The possibility of defining populations of significant wave height and their associated sampling reference periods is investigated, and it is concluded that samples associated with the same reference period are subject to a significant statistical variability. A method for describing the long-term behaviour of H-s that accounts for this variability is proposed. (C) 2000 Elsevier Science B.V. All rights reserved.