Abstract. The assessment of wind conditions in complex terrain requires the use of Computational Fluid Dynamics (CFD) simulations incorporating an accurate parameterization of forest canopy effects and variable thermal stability effects. This study aims to investigate how incorporating the presence of trees can improve flow predictions. A three-dimensional steady Reynolds-averaged Navier-Stokes (RANS) equations model is set up using OpenFOAM to simulate the flow over a complex terrain site comprising two parallel ridges located near Perdigão, Portugal. A 7.5 km × 7.5 km terrain of the Perdigao site is constructed from elevation data centered around a 100 m met-mast located on the northeast ridge. A 30-min averaged stationary period corresponding to near-neutral conditions on a single met-mast tower is simulated. The impact of incorporating different source terms is studied such as forest canopy, Coriolis forces as well as also buoyancy forces. The prediction capability of the models is analyzed for different groups of towers on the South-West ridge, inside the valley, and on the North-East ridge based on the flow topology. The inclusion of a canopy model is shown to improve predictions close to the ground for most of the towers, while reducing prediction accuracy on top of the ridges, illustrating the need to represent terrain heterogeneity.