On the existence and characterization of extreme events in wind data

Abstract

We present evidence that quantitative definitions of extreme events in wind, so called gusts, may be flawed and unable to grasp true extreme situations of wind velocity in short time intervals. In particular, we show that the same statistics of particular pattern shapes in wind data are found in surrogate data generated from the original series of measurements. We apply a pattern recognition algorithm to a two-parameter “Mexican-hat” (duration and amplitude) and find the same probability of occurrence for both original measurements and surrogate data. The distribution of the corresponding values of the duration and amplitude present deviations that can be explained by the filtering of n-point correlations in the original series.