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.

本研究提出了一套用于评估城市峡谷微气候的三维分析框架。研究采用搭载无人机（Unmanned Aerial Vehicle, UAV）的传感器采集温度数据，并以气象观测记录为基准对数据进行校准。本研究运用克里金插值（Kriging Interpolation）方法生成高分辨率三维热场，同时通过K均值聚类（K-means Clustering）算法识别热热点集群。此外，借助空间句法（Space Syntax）分析，通过德劳内三角剖分（Delaunay Triangulation）与网络建模手段，量化了这些热热点的空间整合度与连通性。研究结果显示，即便紧凑的城市形态也会形成显著的热热点，这类热热点通常集中分布于街道狭窄、植被稀少的高密度低矮社区，且多贴近地面而非树冠层。本研究提出的方法将虚拟边界与实际城市建筑结构整合至交互式可视化平台中，支持对热热点关联关系的动态探索。该框架可提供一套可扩展、以数据为驱动的精细化城市热岛效应（Urban Heat Island, UHI）评估工具，并为旨在实现碳中和的城市规划策略提供技术支撑。