Comparison Between Three Statistical Methods for the Extreme Value Analysis of Waves and the Projection of Return Periods

The primary aim of this investigation is to conduct a comparative analysis on the anticipated intervals at which significant wave height (Hs) will occur.The spectral partition technique was used to separate time series.Next, they utilized three established methods for extreme value analysis: (1) Initial Distribution: This method assumes a specific probability distribution for the data and estimates the return period for extreme Hs values based on that distribution.(2) Peak Over Threshold (POT): This approach identifies exceedances of a chosen threshold (a significant wave height) and analyzes those extreme events to estimate return periods.(3) Annual Maximums: Here, the highest Hs value for each year is extracted, and the return period is estimated based on this series of annual maxima.By analyzing extremes in both the combined data and each individual series, the researchers discovered that one series likely contributes more significantly to extreme Hs values within the overall dataset.This suggests that the separate series might represent different wave regimes with varying influences on extreme events.The study emphasizes the benefits of applying extreme value analysis (EVA) to independent wave data series.Furthermore, the peak over threshold statistical method exhibits heightened statistical robustness and improved reliability in predicting return periods using wave data.