Hurricane risk assessment in a multi-hazard context for Dominica in the Caribbean

Abstract

Abstract Hurricanes can trigger widespread landslides and flooding creating compound hazards and multiple risks for vulnerable populations. An example is the island of Dominica in the Caribbean, where the population lives predominantly along the coast close to sea level and is subject to storm surge, with steep topography rising behind, with a propensity for landslides and flash river flooding. The simultaneous occurrence of the multiple hazards amplifies their impacts and couples with physical and social vulnerabilities to threaten lives, livelihoods, and the environment. Neglecting compound hazards underestimates overall risk. Using a whole island macroscale, (level-I) analysis, susceptibility scenarios for hurricanes, triggered landslides, and floods were developed by incorporating physical process parameters. The susceptibilities were combined with vulnerability indicators to map spatial patterns of hurricane multi-risks in Dominica. The analysis adopted a coupled approach involving the frequency ratio (FR), analytic hierarchy process (AHP), and geographic information system (GIS). Detailed hazard modelling was done at selected sites (level-II), incorporating storm surge estimates, landslide runout simulations, and steady flow analysis for floods. High-resolution terrain data and simulation models, the Rapid Mass Movement Simulation (RAMMS) and the hydrologic engineering center’s river analysis system (HEC-RAS), were employed. Ground validation confirmed reasonable agreement between projected and observed scenarios across different spatial scales. Following the United Nations Office for disaster risk reduction (UNDRR) call for the inclusion of local, traditional, and indigenous knowledge, feedback, and expert opinion to improve understanding of disaster risk, 17 interviews with local experts and 4 participatory workshops with residents were conducted, and findings were incorporated into the analysis, so as to gain insights into risk perceptions. The study’s outcomes encompass projections and quantification of hurricane compound hazards, vulnerabilities, accumulated risks, and an understanding of local priorities. These findings will inform decision-making processes for risk mitigation choices and community actions by providing a new framework for multi-hazard risk assessment that is easy to implement in combining different data forms.