We investigated the effect of the canopy description in a Reynolds-averaged Navier-Stokes method based on key flow results from a complex forested site. The canopy structure in RANS is represented trough the frontal area of canopy elements per unit volume, a variable required as input in canopy models. Previously difficult to estimate, this variable can now be easily recovered using aerial LiDAR scans. In this study, three approaches were tested which were all based on a novel method to extract the forest properties from the scans. A first approach used the fully spatial varying frontal area density. In a second approach, the vertical frontal area density variations were ignored, but the horizontally varying forest heights were kept represented. The third approach ignored any variations: the frontal area density was defined as a constant up to a fixed tree height over the whole domain. The results showed significant differences among the cases. The large-scale horizontal heterogeneities produced the largest effect on the variability of wind fields. Close to the surface, specifying more details about the canopy resulted in an increase of x – y area-averaged fields of velocity and turbulent kinetic energy.
Journal of Physics: Conference Series (online), 2014, Vol 524, Issue 1
Main Research Area:
5th International Conference on The Science of Making Torque from Wind 2014European Academy of Wind Energy : The Science of Making Torque from Wind