Optimizing flood response: A decision support tool for protecting critical societal services utilizing flood level projections

Floods are the most common natural hazards causing disasters, with more than 3200 recorded events during the period 2000-2019. Due to projected and already observed increases of climate change phenomena, the frequency and magnitude of extreme weather events are likely to increase, which will subsequently increase the occurrence of major flooding events. This will lead to challenges for the actors tasked with rescuing and protecting people and property from the floods. These actors need to determine which properties and infrastructure to protect to maintain critical societal services or to mitigate cascading hazards. This paper investigates the potential of a decision support tool for allocating resources for the protection of critical societal services during floods. It contributes to the literature on flood protection management by presenting a novel approach to the problem. The tool is centered around an integer programming model with input data based on flood level projections. A mix between quantitative and qualitative data collection methods was applied, involving professionals and experts from the domain studied. The allocations determined by the model in 320 instances of the problem were compared to reactive base case allocations. The results indicate that the model outperforms a reactive allocation approach, even when acting on uncertain information. The tool showed good potential in its ability to increase the task effectiveness and enhance the user performance. One of the conclusions is that proactive protection planning is better than reactive, since it enables protection of the most valuable critical services.