An intercomparison of mesoscale models at simple sites for wind
energy applications
Abstract. An intercomparison of model results from 25 different Numerical Weather Prediction (NWP) models is presented for the year 2011 at six sites in Northern Europe characterized by simple terrain. The model results and a detailed description of each model was submitted by 18 different modeling groups to a open call for data, and serves as a rare quantitative overview of the model uncertainties associated with state-of-the-art mesoscale models used for wind energy applications today. At three of the sites the model intercomparison was verified with observations from nearby meteorological masts. The intercomparison was based on statistical properties of the wind for a number of heights at each site. The results show better performance of the models and a smaller inter-model spread offshore and aloft (2–4 % mean wind speed bias above 40 meters), and greater errors and more spread for inland sites and closer to the surface (up to 7–9 % wind speed bias). For the distributions of wind speed, wind direction, and wind shear only small deviations exist between the observations and the average of the models, but a small shift of the average wind speed distribution towards high wind speeds at Cabauw, and an underrepresentation of strong shear cases was observed. Although the model setup options were studied to determine a 'best practice', no significant indicator was found.