Resumen
In this paper we analyzed the complexity of the time series of wind speed in Petrolina, Brazil at heights of 25 and 50 m. We applied the sample entropy (SampEn) method on wind speed temporal series for each month of 2010, in order to analyze the intra-annual variability of complexity for wind dynamics and its relation with the wind potential. The results showed that wind speed fluctuations are more regular when comparing consecutive data sequences over one hour periods. Although the average wind speed was higher at 50m, indicating higher wind potential, the values of SampEn were also higher between June and September indicating a greater complexity of wind dynamics and consequently lower efficiency in wind energy capture. In the months of February, October, and November, when the average wind speed exceeded the minimum value for generation of wind power (3.5 m s-1) at both heights, the average wind speed increased and the entropy decreased with height. This indicates a higher wind potential and more regularity of wind dynamics at a height of 50 m, both favorable for wind energy production. For the rest of the year the wind power generation was possible only at height of 50 m.