University of Ottawa

Purpose

Coastal flood risk management in Canada has often relied on static, mapbased measures. However, natural disasters such as rapid riverine or coastal floods are highly dynamic, and complex interactions between the environment, assets, and humans arise throughout the course of such an extreme event. Agent-based models (ABM) can include spatiotemporal-based behaviours by assigning "agents" (in this case, humans) a set of rules to follow during an extreme event (e.g. floods). The development of an open-source ABM is proposed to facilitate effective scenario planning and flood risk management strategies. The capabilities of this ABM will include, but will not be limited to: assessing expected loss-of-life, identifying bottlenecks in road networks, and optimizing ideal mitigation options to reduce fatalities. Field surveys of real-time evacuation exercises will be conducted to capture the complex processes and behaviours that arise during a simulated evacuation and will be used as experimental data to validate the developed ABM software. The research findings will be shared through peer-reviewed publications and participating in conferences and workshops with the public.