Because of the high variability of the wind resource and the nonlinear relation between wind speed and power, forecasting wind power is a complex task that is subject to the stochastic nature of the weather. At the same time, accurate forecasts are important for wind plant management and power system operations.

Argonne is utilizing advanced statistical and mathematical methods such as information theoretic learning and conditional kernel density estimation to improve the accuracy of wind power forecasts as well as estimate the forecast uncertainty.