Assessing canopy models across forest edges on flat terrain

Abstract

Computational modelling of atmospheric flows in forested areas is crucial for wind resource assessment, wind farm design and turbine siting, enabling the identification of low wind speed and high turbulence regions. However, when setting up a RANS simulation in a forested area, the question arises: what is the most appropriate canopy model? In this study, employing VENTOS-2D (RANS with k-ε turbulence model), we identified the most suitable of seven canopy models to simulate the mean flow characteristics (wind speed) and second-order flow statistics (turbulent kinetic energy and shear stress) of the flow over a forest edge at the Island of Falster, by comparing against anemometer measurements and LES results. All models exhibited similar profiles for the mean wind speed, with a normalised RMSE of approximately 0.300. Regarding turbulent kinetic energy, the Lopes model outperformed the others, achieving an RMSE of 0.002. Lastly, for the Reynolds shear stress, Liu and Svensson’s models showed the lowest RMSE values (1.756 and 1.609), while others were over 4× larger (RMSE > 7). In summary, the performance of each model varies with the variable in focus, which is related to the model’s derivation method. However, considering all three analysed variables, Liu emerged as the most effective.