Mixing Length Closure Schemes for Forest Canopy Flows over Complex Terrain

While numerical models of canopy flow with a mixing length turbulence closure scheme have existed for a number of years, they have primarily been tested for homogeneous forest canopies or for idealised wind tunnel or flume experiments. Here we use observations from a recent field campaign of forest canopy flow over a complex site with heterogeneous canopy cover in order to test the assumptions made in the mixing length closure scheme. The observations are used to validate numerical model simulations of canopy flow over the field site. The suitability, and limitations, of using the model for representing complex, heterogeneous canopy flow are discussed. Finally the model is used as a tool to investigate flow separation in three-dimensional canopy flows.