Mapping the exposure to solar radiation on urban street canyon using street view images, deep learning, and spatial syntax

Solar radiation in urban canyons significantly impacts resident health and microclimates. Traditional measurements often oversimplify complex street geometries, failing to capture dynamic urban realities. Using central Shanghai as a case study, this research develops a high-precision framework for assessing solar radiation exposure by integrating street view images, deep learning, and spatial syntax. The results indicate that Pudong New Area and Baoshan District experience the highest levels of solar radiation in summer, while the city center exhibits relatively lower radiation intensity. The amount of radiation demonstrates clear monthly fluctuations with seasonal variation, gradually increasing from May to August and declining from September to October. Areas with high solar radiation and high spatial accessibility show a marked spatial concentration during summer, particularly in July and August. Overall, this study establishes a high-precision, scalable framework for measuring solar radiation, enabling accurate identification of high-risk exposure areas. It underscores the importance of enhancing greening and shading measures, especially in areas with both high radiation and high accessibility, and highlights the impact of seasonal changes on the urban thermal environment. These findings provide scientific support for urban planners and managers in improving urban livability and optimizing spatial layout.