A comprehensive framework for quantifying diverse multi-hazard interactions

Résumé fourni par la source

Multi-hazard events pose complex and amplified risks to human life, infrastructure, and ecosystems, often surpassing those of isolated hazards due to cumulative, concurrent, or cascading interactions among multiple natural hazards. While single-hazard studies are well-advanced, evaluating multi-hazard interactions remains challenging due to complex interdependencies and variability across spatial and temporal scales. This study introduces a novel framework to evaluate four types of multi-hazard interactions: multivariate, preconditioned and triggering, spatially compounding, and temporally compounding events. The framework is designed to operate under both present-day and future climate scenarios. It involves three key steps: identifying extreme events, applying a nonparametric copula model to estimate joint return periods, and mapping multi-hazard intensities. Methodological variations address the unique characteristics of each multi-hazard event type. The framework is applied to four European case studies: Oslo (Norway), Múlaþing (Iceland), Essex (UK), and Nice (France), focusing on four pairs of multi-hazard events such as coastal and riverine flooding, rainfall-triggered landslides, intense wind and rainfall, and extreme heat and drought, respectively. Analyses consider current conditions a high-emission future scenario (RCP 8.5), utilizing diverse regional and global datasets. The findings quantify statistical dependencies between hazards and their joint probabilities, and project future increases in multi-hazard events under climate change, offering critical insights for adaptation and risk management strategies in the case study regions. Furthermore, this framework can be adapted to other regions to evaluate diverse multi-hazard interactions, supporting broader efforts in disaster resilience.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.