Optimum Averaging Times for Short-Term Wind Prediction in the Planetary Boundary Layer

Abstract

Root-mean-square error statistics were applied to tower wind data to find the optimum average times of the wind and azimuth to be used as predictors of a 2-min mean wind speed and direction, 0?10 min in the future. The data were divided into classes of stable and unstable boundary layers, and although the predictability was the same, the results showed the difference in the frequency distribution of the atmospheric wave structure for stable and unstable conditions. As opposed to the conclusion of previous studies, beyond the 1-min forecast interval, simple persistence is the best predictor of both the wind speed and direction at the three levels.