Search for a command to run...
In July 2022, two catastrophic flash flood events occurred consecutively in the mountainous regions of Sichuan Province, southwest China, causing significant casualties and infrastructural damage. The first event, on 12 July, took place in the Heishui River watershed of Pingwu County and resulted in 18 fatalities. Four days later, on 16 July, extreme rainfall triggered a devastating flash flood in Baishi Township, Beichuan County, leading to 18 fatalities. This study presents a comparative analysis of these two events by integrating post-disaster field surveys, hydrological analysis, and hydrodynamic simulations to reconstruct the disaster processes and evaluate the performance of monitoring, early warning, and emergency response systems. Both disasters were directly triggered by intense, short-duration rainfall in upstream catchments. However, the Baishi event exhibited more complex mechanisms, including debris flow deposition at the confluence of the Baishui and Qingpian Rivers, a "blockage-break" cycle at a critical bridge, and channel constriction caused by pre-existing landslides. These factors synergistically amplified the flood peak, leading to rapid overbank flow and severe inundation of the township within minutes. The comparative analysis reveals systemic vulnerabilities common to both cases, including an insufficient density of rainfall monitoring stations in upper catchments, delays in disseminating warnings to community-level responders and residents, and inadequate lead time for organised evacuations. By contrasting the triggering mechanisms, disaster amplification factors, and response challenges of these two consecutive events, this study emphasises the necessity of enhancing comprehensive risk prevention. Key recommendations include optimising the layout of monitoring networks in remote upstream areas, developing dynamic and localised early warning protocols with clear actionable thresholds, and refining community-based evacuation plans with safe routes. The findings demonstrate the value of using post-event survey data to validate models in ungauged catchments and contribute to developing more resilient flash flood risk management strategies in complex mountainous terrains under intensifying climatic extremes.