HIGH QUANTILE ANALYSIS OF EXTREME WINDS DATA

In contemporary wind engineering design, assessing and managing extreme wind conditions, particularly the distribution of annual maximum wind speeds, is pivotal for evaluating structural loads and ensuring the safety of various structures. While past research has extensively explored parametric models like the Gumbel, Weibull, and Generalized Pareto Distribution for modeling extreme wind distributions and the calculation of confidence intervals for distribution parameters using parametric and nonparametric bootstrapping techniques, there has been a notable dearth of studies focusing on nonparametric bootstrap methods for establishing confidence intervals for extreme quantiles. This study addresses this research gap by investigating the distribution of high quantiles associated with extreme wind speeds using data from maximum gust wind speed records in Türkiye's Isparta region. The primary objective is to create non-parametric confidence intervals for high quantiles, specifically those linked to extreme 50-year gust wind speeds, with the ultimate goal of enhancing the structural safety of wind turbines susceptible to extreme wind conditions. Multiple nonparametric bootstrap methods are applied, revealing that the accelerated Bias-corrected approach excels in terms of achieving reliable coverage probabilities, providing a robust framework for managing the risks posed by extreme wind events in wind energy infrastructure and structural engineering.