A hybrid wind farm parameterization for mesoscale and climate models

To better understand the potential impacts of wind farms on weather and climate at the local to regional scale, a new hybrid wind farm parameterization is proposed here for mesoscale models, such as the Weather Research and Forecasting Model (WRF), or climate models, such as the Community Atmosphere Model (CAM). All previous wind farm parameterizations treat all the wind turbines in the same grid cell as identical (i.e., they all share the same upstream wind velocity) and ignore the effect of wind direction. By contrast, the new hybrid model considers each individual wind turbine, based on its position in the layout and on wind direction. The new parameterization is developed starting from large eddy simulations (LES) of existing wind farms, in which the local flow around each wind turbine is directly simulated at high spatial ( 3.5 m) and temporal ( 0.1 s) resolutions and the effects of subgrid-scale processes are modeled. Based on analytic and statistical relationships between the LES results and several geometric properties of the wind farm layout (such as blockage ratio and blocking distance), the new hybrid parameterization predicts the local upstream wind speed of each individual wind turbine in the same grid cell, and thus successfully account for the effects of layout and wind direction with little computational cost. With the newly predicted upstream velocity, the turbine-induced forces and added turbulence kinetic energy (TKE) in the atmosphere are derived analytically. The wind speed, wind speed deficit, and TKE profiles and power production obtained with the hybrid parameterization for the test case (the 48-turbine Lillgrund wind farm in Sweden) are in better agreement with the LES results than previous parameterizations. Future work includes the insertion of the hybrid parameterization into the WRF code to assess impacts on near-surface properties, such as temperature and heat and momentum fluxes, in the region surrounding the wind farm.