Investigating the Spatiotemporal Response of Urban Functions to Fine-Grained Resident Activities with a Novel Analytical Framework and Baidu Heatmap

Studying the response of urban functions to residents’ spatiotemporal activity patterns is essential for understanding urban functions and guiding resource allocation. Unlike previous studies constrained by fixed intervals and static functional spaces, this study has developed an analytical framework to examine urban functional responses to residents’ activity patterns under dynamic spatiotemporal combinations. Tensor decomposition was employed to identify key temporal activity patterns of residents and dynamic urban functional patterns, while a Random Forest model was used to evaluate the contributions of five POI (Points of Interest) groups—Transportation, Organizations, Leisure, Habitation, and Basic Facilities—derived from a reclassification of 17 original POI categories, and the Elasticity Index (EI) quantifies functional responsiveness to activity changes. Results indicated that (1) four temporal patterns (sleeping, commuting, daytime, and leisure) and four spatial function types (the basic living area, the residential areas with mixed functions, residential areas with commercial functions and bustling business districts) characterized Beijing’s urban dynamics; (2) the five types of urban function varied with spatiotemporal context, with basic living POIs dominating daytime activities in residential zones and transportation POIs prevailing during commuting in mixed-use areas; (3) EI revealed significant spatial heterogeneity in adaptive capacities to activity transitions, which helped to accurately identify the key areas for improving urban functions. These findings provide new methodological insights and scientific evidence for resilient urban planning and resource optimization, supporting data-driven decision-making for spatial planning, infrastructure allocation, and emergency response management.