Dynamic risk assessment of urban flood disasters based on functional area division—A case study in Shenzhen, China

The assessment of urban flood risk plays a vital role in disaster prevention and mitigation. This work aims to assess the dynamic risk of urban flood triggered by population movements through dividing urban functional zoning from the perspective of collective cognition. Firstly, the urban functional areas are identified using Points of Interest data and then the population movements mobile is detected based on functional areas using mobile signaling big data. Then, one-dimensional and two-dimensional hydrodynamic models are employed to simulate the 50-year flood scenario in Futian District, Shenzhen. Finally, a spatio-temporal dynamic assessment model for urban flood risk is constructed based on the extent of inundation, water depth, population density, and the disaster-bearing capacity of functional areas. The research findings are as follows: (1) Futian District's urban planning showcases harmonious integration of single-function and mixed-function areas. Utilizing the 50% perception standard efficiently identifies distinct functional types across diverse urban zones. The results are highly consistent with the actual situation. (2) During morning peak hours, the population exhibits a nuanced pattern of dispersal, concentration, and transition. Lunchtime witnesses multiple central clusters forming and gradually dispersing, while the evening peak witnesses population regrouping, covering broader geographical extents. Dynamic utilization of functional areas and mobile phone signaling data outperforms static population metrics, offering deeper insights into the complexities of human activity. (3) Between 12:00 and 13:00, lunchtime movements lead to a surge of 6 high-risk zones in the central area and 5 in the Meiling area. The dynamic flood risk assessment model, based on functional area delineation, effectively identifies disparities and fluctuations in flood risk across diverse functional areas during rainfall scenarios, ensuring heightened precision and accuracy in risk assessment.