Impact characteristics and interaction effects of built environment on street space quality in megacities: A case study of Xi’an, China

The rapid process of urbanization has led to widespread issues of uneven street space quality, insufficient comfort, and inadequate governance in mega-cities. In-depth research into how built environment factors influence street space quality is of significant theoretical and practical importance. This study focuses on the central urban area of Xi’an, a mega-city in China, and constructs an index system for the built environment that covers four dimensions: spatial structure, interface landscape, functional vitality, and traffic conditions, comprising 19 variables. Utilizing multi-source data, such as street view images, points of interest (POIs), and architectural data, and employing the LightGBM machine learning model in conjunction with SHAP interpretability analysis, this paper reveals the non-linear impacts of built environment variables on street space quality and their interactions. The study finds that: (1) The influence of various built environment dimensions on street space quality varies significantly, with spatial structure being the most critical dimension, particularly the environmental openness index, street scale perception, and interface enclosure index; (2) The relationship between variables and street space quality is non-linear, with key threshold intervals identified. For example, the environmental openness index, green view index, and interface richness index show significant improvements in quality within specific ranges, whereas interface enclosure and travel convenience exhibit negative effects beyond certain thresholds; (3) There are significant and complex interactions between variables, with the synergistic or inhibitory effects of environmental openness, building height, and traffic facility density being particularly pronounced. This research offers a new theoretical perspective and methodological foundation for understanding the formation mechanisms of street space quality, providing valuable empirical evidence and optimization pathways for spatial governance and the development of high-quality streets in mega-cities.