Aspects of Applying the ACER Method to Ocean Wave Data

Abstract

This thesis gives a presentation of the classical Peaks Over Threshold (POT) method and the recently developed Average Conditional Exceedence Rates (ACER) method. Real life ocean significant wave height data sets, each containing 50 years of observations, have been analyzed using both these methods. When real life data are analyzed with the use of the ACER method, problems regarding finding an optimal set of parameters are often experienced. The practice that have been used to overcome these problems has demanded a large degree of knowledge and/or experience on the analysts behalf. The main focus in this thesis have thus been on finding a solution to these problems. It was found that if one of the parameters is fixed to a certain value, the problems are avoided and very reasonable results are obtained for all the data analyzed. The ACER method is found to perform better than the POT method.