Characterization of Wind Power Forecast Error Using the WRF Model: Inner Mongolia Case Study

Abstract

In China, the wind power is expanding very quickly. We examine a specific real wind farm in China and analyze the WRF (Weather Research and Forecasting) model forecasts with different physical configurations and spatial resolutions. The model short-term wind speed predictions for February 2008 are compared with a real data collected in a met tower in the wind farm at 10-min sampling rate. In our case study, the WRF model run with the MYJ scheme outperforms the YSU scheme, on average. We focus on the model skill improvement for winds above 3m/sec that essential for wind farm companies. Our study indicates that wind speed forecast error at met tower location does not exceed 2 m/sec for 6 km resolution grid.