Análise multifractal da velocidade do vento em Pernambuco

Abstract

The study of climate has great importance, given that a variation of climatic elements affect the economy of a certain region and life of the inhabitants. Climate variables temperature, humidity, atmospheric pressure, solar radiation, precipitation and wind can be affected by geophysical and environmental factors such as latitude, altitude, air mass, proximity to sea, sea currents and vegetation. Wind is the most complex climate element representing the natural phenomenon of turbulence, it is characterized by high temporal and spatial variability. Wind is generated by atmospheric air mass movement, and has influence on various environmental phenomena such as soil erosion, pollutant dispersal and transport of pollen and seeds. Knowing wind speed temporal and spatial distribution is crucial to evaluate the potential for generation of eolic energy. In this work we study long-term correlations in wind speed temporal series registered at twelve meteorological stations in the state of Pernambuco, Brazil. To this end we apply Multifractal Detrended Fluctuation Analysis (MF-DFA) on hourly wind speed data for the period 2008-2011. All the analyzed series exhibit multifractal properties with generalized Hurst exponents above 0.5 indicating persistent temporal dynamics for both, small and large fluctuations. We also calculate other multifractal measures Rényi exponent and singularity spectrum, and complexity parameters, position of maximum, width and asymmetry of multifractral spectrum. No correlation was detected between complexity parameters and the geographic parameters longitude, latitude and altitude of the station, except for asymmetry of multifractal spectrum: negative correlation with longitude for maximum wind speed and negative correlation with latitude for average wind speed. However for all stations the strength of multifractality (indicated by width of multifractal spectrum) is greater for maximum wind speed then for average wind speed. These results contribute to a better understanding of the nature of stochastic processes governing wind dynamics which is necessary for development of more accurate predictive models for wind speed temporal variability and diverse phenomena influenced by wind.