Achieving Accurate Return Period Estimation of Significant Wave Height Using FAIR-Compliant Data

This paper presents the calculation of the return period for the significant wave height using three different methods: Block Maxima (BM), Peak Over Threshold (POT) and Point Process (PP). The data used for the calculation is in accordance with the FAIR (Findable, Accessible, Interoperable and Reusable) principle, with an example provided for the case study of Marina di Massa, Italy, for an hourly dataset of 40 years (1979-2018). The BM method is based on evaluating the maximum waves over a given period, while the POT method focuses on the peak values of the waves. The POT method is a widely used technique which can be used to estimate the return period of extreme waves. This method can determine the threshold values of the wave heights that represent extreme events. With this method, one can choose between graphical approach or an automatic threshold estimation approach to determine the threshold values of the wave heights representing extreme events. This study further includes the main characteristics of the POT method, provides insight into the two threshold estimation methods, while the Point Process approach is based on the identification of extreme wave events. All three methods are compared in terms of accuracy, and goodness-of-fit score (i.e negative log-likehood, AIC and BIC) and the results are then used to calculate the return period of the significant wave height for 25, 50 and 100 years. This analysis shows that POT is the most accurate and reliable method for calculating
the return period of the significant wave height. The BM and the PP approaches show marginally lower accuracies.