Assessing urban heat island hotspots through virtual and physical boundary analysis with space syntax

This study presents a three-dimensional framework for assessing the microclimate in urban canyons. The temperature data were collected using UAV-mounted sensors and calibrated against meteorological records. Kriging interpolation was employed to create high-resolution 3D thermal fields, while K-means clustering was used to identify hotspot groupings. Space syntax analysis was used to quantify the spatial integration and connectivity of these hotspots through Delaunay triangulation and network modeling. Findings reveal that even compact urban forms can generate pronounced hotspots, which are typically concentrated in dense, low-rise neighborhoods with narrow streets and scant vegetation, often occurring near the ground rather than at the canopy level. The proposed method integrates virtual boundaries with actual urban structures within an interactive visualization platform, allowing for the dynamic exploration of hotspot relationships. This framework offers a scalable, data-driven tool for assessing UHI at a fine-grained level and supports urban planning strategies aimed at achieving carbon neutrality.