A process-based modeling framework for pedestrian-scale urban thermal environments under complex shading conditions

Pedestrian-scale urban heat exposure is governed by highly dynamic radiative processes shaped by complex three-dimensional urban structures. However, most existing studies rely on static surface properties or geometric proxies, which fail to explicitly capture the spatiotemporal radiative mechanisms governing near-ground thermal environments. Here, UAV thermal infrared observations are integrated with ground measurements to develop the Composite Urban Thermal Exposure Index, a process-oriented framework for pedestrian-scale thermal analysis. Building on conventional land-cover and three-dimensional morphology indicators, CUTEI introduces dynamic radiative variables describing instantaneous solar exposure, neighborhood-scale radiative modulation, and radiative accumulation associated with thermal memory. Object-based hierarchical modeling shows that explicitly incorporating radiative processes reshapes the explanatory structure of land surface temperature. Dominant controls shift from material and geometric proxies to radiative process variables, improving the representation of street-scale thermal heterogeneity. Further analysis indicates that accumulated radiative history constrains temperature variability under identical instantaneous exposure conditions, highlighting the role of temporal integration in pedestrian-scale thermal formation.Process consistency is evaluated by comparing UAV-derived and ground-measured temperatures under complex shading conditions. Temperature differences across observation levels are systematically explained by CUTEI’s dynamic radiative indicators, reflecting different expressions of the same radiative process rather than measurement errors. Validation with independent multi-year samples confirms the robustness of the framework across observation contexts. Overall, CUTEI provides a physically consistent and interpretable framework for translating urban morphology and shading into dynamic radiative processes, offering a transferable quantitative basis for pedestrian-scale heat exposure assessment and heat adaptation planning.