Linking Urban Morphology and Human Mobility Resilience to Pluvial Flooding: A Comparative Study of Natural and Stormwater Management Units in Shenzhen

Global climate change and rapid urbanization have intensified extreme precipitation events, making urban pluvial flood risk management particularly urgent. However, existing research exhibits theoretical and methodological limitations in reconciling pluvial flood resilience within high-density urban development, especially lacking a quantitative framework for human mobility resilience based on natural hydrological units. To address this, this study takes 48 natural catchment areas in Shenzhen as research units. By integrating mobile phone signaling data, remote sensing imagery, and multi-dimensional spatial form indicators, it constructs a post-disaster recovery curve based on population dynamics to quantify human mobility resilience. Using Principal Component Analysis (PCA), Ordinary Least Squares (OLS) regression, and the K-means clustering method, this research analyzes the mechanisms through which green infrastructure, road network structure, and three-dimensional building morphology influence resilience. The results show that three-dimensional building morphology is the core driver of post-disaster recovery capacity, with Floor Area Ratio (FAR), Standard Deviation of Building Height (SDBH), and Building Coverage Ratio (BCR) significantly promoting recovery, while Building Shape Coefficient (BSC) exerts an inhibitory effect. A key comparative analysis reveals that the model based on natural catchment areas has significantly better explanatory power than the model using stormwater management units, and the dominant factors differ: building morphology factors are more prominent in natural hydrological units, whereas road network structure factors are more significant in stormwater management units. This study confirms that natural geographic boundaries can more authentically reveal the intrinsic “morphology–resilience” relationship, providing an important theoretical and empirical basis for optimizing the planning and management units of sponge cities in high-density urban areas.