Quantifying the added value of impact-based warnings for flash flood monitoring using innovative multi-source impact data

Flash-floods have devastating impacts on societies. Early warning systems play a crucial role in mitigating these impacts by enabling preventive actions. Traditionally, warning systems have been hazard-based, relying on predefined thresholds for physical parameters such as discharge in the case of floods. However, there is growing interest in impact-based warnings (IBW), which integrate hazard information with exposure and vulnerability to communicate potential impacts. IBW are expected to improve communication with the public, prioritize affected areas, and reduce false alarms, but their operational benefits are not sufficiently validated. This study focuses on IBW for flash floods, comparing their performance with traditional hazard-based warnings (HBW). We used a validation framework based on diverse datasets, including legislative decrees attesting to natural disasters, insurance claim records, and fire and rescue service operation logs. The analysis spans a 13-year continuous period (2010–2022) over the French Mediterranean region, and two different spatial resolutions (river reach and municipality scales). The IBW were based on thresholds of the number of flooded buildings, while the HBW are based on discharge return period thresholds. Our results demonstrate a clear added value of IBW over HBW, particularly at finer spatial scales. On the river reach scale, IBW reduced the number of false alarms by a factor of 2 to 3 compared to HBW. This highlights the potential of IBW to enhance decision-making processes by providing more precise and actionable warnings. The findings underscore the need for systematic impact data collection and the integration of diverse datasets to further refine and validate IBW.