Classical Extreme Value Model and Prediction of Extreme Winds

In this note, the propriety of using the Gumbel distribution for modeling of extreme wind speeds as opposed to use of the reverse Weibull distribution is discussed. It is shown that the assumption of bounded wind speeds and the subsequent implementation of the peaks over threshold method for estimating the required parameters from real data lead to contradictions, and that the Gumbel distribution at several sites serves more realistically as a model for the distribution of extreme wind speeds.