Pre- and Post-COVID-19 Reconfiguration of Human Activity–Function Spatial Relationships in Metro Station Areas: Evidence from Xi’an, China

Source abstract

Post-pandemic station-area studies have largely focused on activity levels, ridership, mobility volume, or aggregate recovery, with less attention to whether local human activity–function spatial relationships themselves persist or reorganize. This study examines five functionally contrasting station areas along Xi’an Metro Line 2 using observations from 2019 and 2023. Baidu Population Heat Index data represent observable population activity intensity, while Amap points of interest (POIs) characterize the spatial intensity of ten urban-function categories within 800 m station-area buffers divided into 100 m × 100 m grids. Bivariate Spatial Association (BSA), intertemporal transition analysis, and Local Bivariate Relationships (LBR) are combined to identify local relational states, transition pathways, and relationship forms across six weekday/weekend and intraday contexts. Overall, 61.31% of local association states retained the same categorical state across the two observations, including 28.90% that remained in the same significant state and 32.41% that remained non-significant, while 38.68% changed state. Weakening and Emergence were more common than Type Reconfiguration, and HH and LL states were more persistent than HL and LH mismatch states. Relational change also varied across facility categories and temporal contexts: corporate and office, transport, shopping, financial-service, and entertainment functions changed more frequently, whereas residential-community, catering, life-service, and medical functions generally showed greater continuity. LBR further revealed multiple linear, nonlinear, complex, and non-significant local relationship forms that differed across station contexts. The findings show that pre-/post-pandemic station-area change involved both relational continuity and localized reorganization. This relational perspective complements activity- or ridership-based assessment and supports screening and monitoring of station-area change without implying causal COVID-19 effects or direct measurement of urban resilience.