A framework for investigating multi-hazard interactions to develop a decision-support system for disaster risk management 

Multi-hazard events pose significant threats to human lives and assets, often exceeding the risks associated with single hazards due to simultaneous, cascading, or cumulative occurrences of multiple interacting natural hazards. The society and environment in various European regions, susceptible to many climatic extremes, are anticipated to be profoundly impacted in the coming decades due to the increasing frequency and severity of multi-hazard events linked to changing climatic conditions. While most natural hazard studies have predominantly focused on single hazards or multi-layer single hazards, the quantitative assessment of multi-hazard interactions remains in its early stage of development. Investigating diverse types of multi-hazard events is particularly challenging due to complex interactions between hazard drivers and the spatial and temporal heterogeneity of multiple hazard occurrences. This study aims to introduce an approach for investigating four distinct types of multi-hazard interactions: preconditioned and triggering, multivariate, temporally compounding, and spatially compounding events, under present-day and future climate change scenarios. The research is conducted as part of a Horizon Europe project MEDiate (Multi-hazard and risk informed system for Enhanced local and regional Disaster risk management), which seeks to "develop a decision-support system (DSS) for disaster risk management by considering multiple interacting natural hazards and cascading impacts." The framework is implemented on four interactive multi-hazard pairs—compounding coastal and riverine flooding, extreme heat and drought, extreme wind and precipitation, and extreme precipitation and landslides—in four European testbeds: Oslo (Norway), Nice (France), Essex (UK), and Múlaþing (Iceland), respectively. The proposed multi-hazard interaction framework aims to estimate the probability of occurrence of multiple hazards over various time and space scales, associated with the four types of multi-hazard events. The framework involves two key steps. First, it identifies extreme events for individual hazard indicators (e.g., peak river flow, surge, near-surface wind speed, precipitation, air temperature) at different time intervals (e.g., daily, monthly, quarterly) and locations within the testbed regions. Second, a nonparametric bivariate copula-based approach is employed to estimate joint return periods for various combinations of hazard indicators associated with different types of multi-hazard events. The analysis is conducted for both present-day conditions and the 2050 RCP 8.5 climate change scenario, by using several freely available regional and global observation and modelled datasets related to different indicators of multi-hazard events. The findings of this study illustrate the degree of statistical dependence between various combinations of interactive hazards in space and time, quantifying joint probabilities of multi-hazard events. Furthermore, it demonstrates how these probabilities are likely to change in the future due to the impacts of climate change. This research emphasises the importance of considering diverse scenarios of multi-hazard events in formulating future climate change adaptation responses. The findings of this study will inform the DSS being created in the MEDiate project by developing accurate multi-hazard scenarios to estimate the potential effects of different disaster risk mitigation and adaptation strategies. The results could also contribute valuable insights for developing multi-hazard risk management policies elsewhere globally, where susceptibility to multi-hazard events is increasing.