Accurately predicting wind resources in forested terrain remains a challenge for wind energy development. Researchers have advanced computational fluid dynamics methods by developing a large-eddy simulation (LES) model that explicitly represents both realistic forest canopies and topography using high-resolution data from airborne laser scanning. The approach addresses critical gaps in current modeling practices, which typically rely on simplified surface roughness parameters that poorly capture forest-atmosphere interactions.
The three-phase study—verification, validation, and footprint analysis—compared explicit forest representation against traditional roughness-length models. Results confirm that when forest canopy structure is modeled directly, turbulence statistics align well with established literature benchmarks and vary predictably with tree density. Importantly, the study revealed that ground roughness effects become negligible once forest drag is properly represented, even for sparse vegetation. Conversely, models using only roughness length generated significant deviations in turbulence characteristics, partly because the approach cannot generate sufficient drag while maintaining grid resolution compatibility near ground level.
Validation against field measurements from southeastern Sweden demonstrated the model's ability to capture flow disturbances caused by different surface features across multiple wind directions. The coherence calculations matched or exceeded IEC standards, though slight discrepancies were attributed to reference velocity normalization methods.
A key finding emerged from the footprint analysis: wind at any point is most influenced by forest drag approximately 10 heights upstream, increasing to 50 heights upstream for turbulence effects. This insight directly informs computational domain sizing for forest simulations under neutral atmospheric conditions. Heterogeneous forest distributions produced higher drag than uniform conditions and better matched observations. Terrain features smaller than 100 meters significantly affected wind flow near ground level, though differences diminished at higher elevations.
These results provide wind resource engineers with practical guidance for siting assessments in complex terrain, enabling more accurate energy yield predictions and refined understanding of how forests modulate wind characteristics.
