LONG TERM CORRELATIONS AND LACUNARITY OF WIND DIRECTION IN FERNANDO DE NORONHA

Abstract

The complexity of wind direction dynamics at Fernando de Noronha Island, Brazil is analyzed from the point of view of its lacunarity and persistence. We analyze hourly wind direction data recorded over the period 2003-2010. We apply lacunarity analysis that measures the distribution of gaps in dataset on daily wind direction time series in different directional sectors and estimate power-law exponent that describes the rate at which the lacunarity decreases with window length. We also apply Detrended Fluctuation Analysis (DFA) to study temporal correlations in wind direction increments. The results show that the value of lacunarity L(r) for all window sizes, as well the value of power-law exponent are the lowest for prevailing directional sector indicating less clustering and more uniform gaps in wind episodes. The value of DFA exponent α for increment series is greater than 0.5 indicating persistence in wind direction dynamics. These findings should prove useful for developing and validation of theoretical and computational models for wind direction dynamics in Fernando de Noronha, which is crucial for implementation of alternative energy solutions that guarantee sustainable development of this island.