Exploring the Cooling Power of Trees Using Computer Vision

Rising temperatures are placing increasing pressure on many cities to adapt pub- lic spaces for heightened heat stress. While the cooling benefits of urban trees are well documented, there remains a limited understanding of how these effects vary across tree types, morphologies, and urban contexts. This study addresses this knowledge gap by employing computer vision techniques to analyze thermal imagery of urban trees in four climatically diverse cities: Amsterdam, Boston, Dubai, and Los Angeles. Our findings reveal significant variation in cooling performance attributable to genera-specific traits, water consumption properties and surrounding environmental factors. By initiating a comparative analysis of urban tree types and their thermal performance, this research provides a foundation for targeted, climate-responsive urban greening policies.