Novel Extreme Value Estimation Procedures: Application to Extreme Wind Data

Stochastic modeling of long-term and extreme value estimation of wind and sea conditions for probabilistic reliability assessments of wave energy devices

“…A threshold of 18 m/s is considered such that 199 independent maximum storm values (minimal separation: 30 h) result. This threshold fulfills the average number of extreme values proposed by Gross et al (1994). Fig.…”

“…Simiu et al, 1995). Gross et al (1994) showed that an optimal threshold should correspond to a mean exceedance rate of 5-15 values per year. A minimum temporal separation between different recorded maximum storm values make sure that the extreme values do not result from the same wind storm.…”

Bootstrap simulations for evaluating the uncertainty associated with peaks‐over‐threshold estimates of extreme wind velocity

“…Here, data series (n ¼ 900, T ¼ 30 years) were simulated from a Gumbel distribution ( ¼ 30 mph, ¼ 6.5), similar to that used by Gross et al (1994). The POT sample size (k) was varied from 30 to 300 (i.e.…”

Bivariate estimation of extreme wind speeds

“…In several works [3][4][5][6][7][8] the extreme value type III distribution for the maxima, also called RW distribution has been considered as an adequate model of EWS based on epochal and peaks over threshold (POT) approaches, and it has been suggested that its application would result in more reasonable estimates of the distribution of yearly maximum wind speed and of wind-induced failure probabilities and wind load factors than the corresponding estimates based on the G distribution.…”