Workflow to accurately model vegetation canopy effects in building energy simulation

This work develops an efficient and accurate workflow for integrating vegetation canopy effects into building energy simulations through unmanned aerial vehicle (UAV)-based video capture and automated reconstruction techniques. Unlike traditional methods that rely on static vegetation representations, the present approach utilizes a dual-perspective UAV video strategy to simultaneously capture the building exterior and its surrounding vegetation from an interior-facing viewpoint.

This method enables precise shading analysis while significantly reducing computational costs compared to full-scale microclimate simulations. The high-resolution UAV data is processed using advanced photogrammetry and 2D Gaussian Splatting to reconstruct a detailed 3D building model with optimized vegetation meshes that accurately preserve canopy geometry. The 2D Gaussians refine the meshes’ representation by optimizing mapping calculations. These meshes are then incorporated as shading elements within energy simulation platforms such as Grasshopper/EnergyPlus, thereby enhancing simulation accuracy over conventional coarse approximations.

By dynamically integrating real-world vegetation geometry, the present workflow yields context-aware and seasonally adaptable results, bridging the gap between high-fidelity 3D reconstruction and practical energy analysis. This scalable and automated approach offers a promising avenue for urban energy modeling and the optimization of passive solar design in built environments.