From diagnosis to explanation: An integrated framework for examining spatially heterogeneous associations between temporary industrial-land use and neighborhood gentrification

As Chinese megacities transition to post-industrialism, the socio-spatial consequences of temporary industrial land reuse—a pivotal urban renewal strategy—remain under-theorized. This study integrates Multiscale Geographically Weighted Regression (MGWR), Geodetector, and Generalized Additive Models (GAMs) to examine associations between four functional types of temporary use and a Standardized Gentrification Index (SGI) capturing changes in educational and floating-population composition across 134 Beijing subdistricts. The results reveal differentiated and context-dependent patterns. Technology-oriented use density exhibits a stable and spatially widespread positive association with SGI. Cultural-use area is positively associated with SGI across most subdistricts, whereas cultural-use density is predominantly negative. Production-service area is significant only in a limited set of locations, while consumer-service measures show no statistically significant local MGWR coefficients. Cultural and technology clustering recur as the principal correlates of coefficient heterogeneity, whereas production clustering, transport-accessibility growth, and housing-price change matter in specific relationships. These findings suggest a state-mediated, technology-oriented pathway of demographic upgrading and a more conditional cultural pathway. Methodologically, the framework moves from mapping spatial non-stationarity to interpreting its contextual correlates, providing evidence for differentiated urban-renewal governance.