Wind-driven asymmetry in temperature gradients at the edges of a forest patch

Forest edges are expanding in fragmented and managed landscapes, and their microclimates influence understory conditions, regeneration, and habitat suitability. Edge-to-interior temperature gradients are often characterized only by distance from the edge, even though airflow can differ between windward and leeward edges. We developed a two-dimensional forest microclimate model that couples radiative transfer, canopy flow, and heat transfer on a finite-volume scheme. The model was driven by leaf area density and canopy height derived from unmanned aerial vehicle light detection and ranging (UAV-LiDAR). The flow solver was benchmarked against OpenFOAM, and near-surface air temperature was evaluated against multi-day observations in coniferous and broadleaf forest patches. The model reproduced benchmark canopy-flow patterns and near-surface air temperature in the coniferous patch (RMSE = 1.04 °C, R² = 0.87). It was further evaluated in a nearby broadleaf patch (RMSE = 1.73 °C, R² = 0.78). Simulations showed contrasting edge-to-interior cooling forms on the windward and leeward sides. Windward edges cooled more gradually because wind penetration maintained advective heat transport, whereas leeward edges cooled more sharply as weak flow shifted heat transfer toward diffusion. These asymmetric gradients were strongest in the understory and middle layer, with clear daytime variation. Our results indicate that edge influence depth cannot be assumed to depend only on distance from the forest edge. Forest microclimate monitoring and management in fragmented or retained patches should account for prevailing wind direction and edge orientation when locating transects, defining buffer zones, and evaluating thermal habitat conditions.