Meso-scale modeling of a forested landscape

Meso-scale models are increasingly used for estimating wind resources for wind turbine siting. In this study, we investigate how the Weather Research and Forecasting (WRF) model performs using standard model settings in two different planetary boundary layer schemes for a forested landscape and how this performance is changed when enhancing the roughness by a factor four in one of the schemes. The model simulations were evaluated using data from a 138 m tall mast in southeastern Sweden, where an experiment with six sonic anemometers and standard meteorological instrumentation was performed 2010-2012. The land cover around the mast is dominated by forest and for the most common wind direction, the forest extends more than 200 km from the mast. The two low-roughness simulations showed differences both in terms of estimated wind resource and wind shear. The simulation with enhanced roughness results in an improved correlation with measured data for near-neutral situations in the observed height range, whereas the correlation is deteriorated relative to the standard setup for stable atmospheric stratifications for heights above approximately 80 m. The inclusion of the displacement height in the post-processing of the results is also discussed.