Review article: Stocktaking of methods for assessing dynamic vulnerability in the context of flood hazard research

Abstract. Dynamic vulnerability, driven by changing social, economic, physical, and environmental characteristics, is critical to understanding flood risk. Despite its importance, existing flood risk assessment research often overlooks the mechanisms that drive dynamic vulnerability and the interactions between underlying characteristics. In this study, we systematically review methods used to assess dynamic vulnerability in the context of floods and compile their findings about the drivers and effects of the dynamics in a dataset. We identify 28 relevant studies and group them into four categories of vulnerability dynamics: single-event, consecutive events, co-occurring events, and underlying dynamics. We find that most studies rely on indicator-based, statistical, or qualitative methods, with a notable under-representation of damage curves and process-based modeling approaches such as agent-based models. Demographics, economic characteristics, and awareness of flood risks are vulnerability dimensions most frequently assessed, whereas governance, health, crime, and conflict are rarely addressed. Data sources vary widely, with interviews and surveys dominating studies on consecutive events and single-event dynamics. In contrast, studies on underlying dynamics and co-occurring event dynamics use a much wider array of data sources (e.g., cadastral data, maps, or modeled data). This review highlights methodological gaps, including the limited analysis of causal relationships and the lack of integrated approaches for multi-hazard contexts. Advancing flood risk research requires holistic assessments, integration of diverse dimensions, and the development of dynamic modeling techniques to capture evolving vulnerability processes.