Diagnosing Spatiotemporal Urban Flood Responses Under Shift Rainfall Variabilities

Conventional flood risk assessments primarily emphasize the spatial distribution of flooding, often overlooking its temporal characteristics. However, temporal flood distribution is a critical factor in real-time flood forecasting and early warning systems, particularly in flash flood scenarios. Moreover, rainfall movement significantly influences hydraulic behaviour, such as hydrograph response. This study investigates the impact of rainfall movement on the spatiotemporal dynamics of urban flooding using an innovative approach, the Percentage of Nodes in Flood Assessment (PNFA). To achieve this objective, 11,520 simulations were conducted using the Storm Water Management Model (SWMM) in Do Lo, Ha Noi, Viet Nam, employing sequential rainfall datasets under two distinct rainfall shifting scenarios. Rainfall movement was represented by shifting the time among sequential rainfall hyetographs recorded at three synthetic rain gauges. The shifting time varied from −1440 to 1440 minutes, allowing for a comprehensive assessment of rainfall shifts. Based on the percentage of nodes in flood, PNFA effectively evaluated flood behaviour and its sensitivity across both spatial and temporal scales under different rainfall patterns. The results indicate that uniform rainfall conditions led to the most severe flooding. A 32-minute partial shift in rainfall onset reduced the flooded area by 6%, the total flood volume by approximately 1,045 m 3 (~3.34%) and flood depth 13 cm (~63,2%) . Additionally, the peak flood time was delayed by 69 minutes, providing a crucial window for emergency response and evacuation. These findings underscore the importance of incorporating spatiotemporal rainfall variability into flood risk models to enhance urban flood mitigation strategies. This study sheds a light on nature-based solution (NBS) designing and assessing process by initially indicating the locality and size of the NBS.