Shielded but unprepared: quantifying the safe-development paradox in ice-jam flood risk management

Abstract

Ice-jam floods pose a critical threat to northern riverine communities due to their rapid onset, unpredictability, and potential to exceed historical open-water flood levels. While structural solutions like artificial ice breaking can mitigate immediate risks, long-term resilience relies heavily on household-level adaptation. However, many existing models fall short in capturing the complexity of human behavior, often depending on simplified or theoretical assumptions. This study enhances the Dynamic Ice-jam Flood Risk Assessment framework by integrating the Transtheoretical Model (TTM) of behavior change, using empirical data from Fort McMurray, Alberta. Shapley regression was applied to identify the relative influence of behavioral drivers—such as perceived risk, cost, self-efficacy, perceived response efficacy, prior flood experience, and social networks—on transitions through the TTM stages of change. These findings were incorporated into an agent-based model coupled with the RIVICE hydrodynamic model to simulate twelve policy scenarios blending structural interventions and behavioral strategies. Scenario simulations suggest that while structural measures can reduce overall flood risk by more than 70%, they also lead to a ∼60% decline in household-level adaptation efforts. Conversely, behavioral interventions tailored to the population’s readiness to act achieved household adoption rates of up to 18.5% of households (a 106% relative increase from the 9.0% baseline), albeit with smaller impacts on flood exposure. These results highlight the need for integrated policy approaches that balance physical risk reduction with strategies that sustain individual and collective adaptation. Subject to local recalibration, the framework offers a transferable pathway for cold-region flood adaptation policy.