Over the last twenty years, numerous strategies and policies (e.g., 2030 Agenda for Sustainable Development) have emerged to promote flood risk reduction that is compatible with the conservation or restoration of river ecosystems, with the ultimate goal of achieving sustainable development. In this context, vulnerability analysis is considered a key aspect in flood risk reduction, and the most widespread methodology for its characterization is the construction of indices. However, such indices have been obtained so far through a fragmented approach, either because they do not consider all vulnerability dimensions (i.e., social, economic, ecosystem, physical, institutional and cultural) or components (i.e., exposure, sensitivity and resilience). Moreover, indices developed on a regional scale focusing on areas prone to flash floods are rarely validated, as the necessary information is often not available and flash floods do not occur simultaneously in all urban areas of a given region. This paper addresses the above two knowledge gaps and describes the construction at regional level of a flash flood Integrated Multidimensional Vulnerability Index (IMVI). Vulnerability here is characterized holistically, integrating in the index all the dimensions and components involved. Subsequently, an uncertainty and sensitivity analysis was performed to validate the IMVI. Lastly, vulnerability regional spatial patterns were identified through a Latent Class Cluster Analysis. The methodology implemented here allows one to identify vulnerability sources and their underlying causes, helping to improve flood risk management. Moreover, validation outputs enable to determine index uncertainty sources, thereby encouraging decision-makers to design more cost-effective vulnerability reduction strategies.

在过去的二十年中，出现了许多战略和政策（例如，2030 年可持续发展议程），以促进与河流生态系统保护或恢复相适应的洪水风险降低，最终目标是实现可持续发展。在这种情况下，脆弱性分析被认为是降低洪水风险的一个关键方面，最普遍的表征方法是指数的构建。然而，到目前为止，此类指数是通过零散的方法获得的，因为它们没有考虑所有脆弱性维度（即社会、经济、生态系统、物理、制度和文化）或组成部分（即暴露、敏感性和复原力）。而且，在区域范围内制定的侧重于易发生山洪的地区的指数很少得到验证，因为通常无法获得必要的信息，而且山洪并非在特定区域的所有城市地区同时发生。本文解决了上述两个知识空白，并描述了在区域层面上构建的山洪综合多维脆弱性指数（IMVI）。这里的脆弱性是整体表征的，在索引中集成了所有涉及的维度和组件。随后，进行了不确定性和敏感性分析以验证 IMVI。最后，通过潜在类聚类分析确定了脆弱性区域空间模式。此处实施的方法允许人们识别漏洞来源及其根本原因，帮助改善洪水风险管理。此外，验证输出能够确定指数不确定性来源，从而鼓励决策者设计更具成本效益的漏洞减少策略。