Multiscale multifractal analysis of near-surface wind speed time series

Abstract

Understanding the temporal/spatial behavior and dynamics of wind speed is of great importance for model building, wind prediction and sensor design. In the present study, we employ a recently proposed approach of multiscale multifractal analysis (MMA) to investigate the scaling properties of a group of wind speed time series recorded at ten locations with 1m interval in line. The results of MMA show that all wind speed time series not only exhibit strong multifractal properties, but also that these properties depend on the time scale, indicating the great necessity to study wind speed time series from multiple scales perspective. Subsequent analysis of shuffled and surrogate series reveals that the multifractality of wind speed time series is mainly stemming from the long-range correlation, while has less to do with broad probability density function. Furthermore, based on the results of Hurst surface analysis, we find that the scaling behaviors of all wind speed time series are similar for large fluctuations, while exhibit apparent distinction for small fluctuations.