Intraseasonal and Seasonal Prediction of Wind Energy in Electrical Markets

Our organization produces forecasts of wind energy electrical generation in a large number of electrical interconnections in the United States, Canada and Europe. Using our data base of wind farm locations, turbine numbers and types, we are able to use reanalyzed winds from NOAA's Climate Forecast System Reanalysis to calculate the electrical power that would have been generated by the existing wind farm network for the last thirty years. The resulting data are helpful for long-term forecasting via two pathways. First, comparison with observed wind generation over the past few years allows for removal of model biases, so that seasonal scale forecasts of wind energy using output from the Climate Forecast System version 2 are possible. Second, the 30 years of simulated wind generation allows for calibration of an index-based forecast system, so that we can see how, electrical generation the existing wind infrastructure in a given electrical interchange would have correlated with atmospheric indices such as the NAO, MJO and ENSO. Forecasts of these indices can then be used to forecast wind generation within that interchange. We compare the skill of these approaches over a range of lead times.