An integrated agent-based and hydrodynamic modelling framework for urban flood impact assessment using a risk priority index

Agent-based models (ABMs) are widely used to investigate flood impacts on transportation systems, but many do not adequately represent behavioural adaptation and information diffusion. This study presents an integrated framework coupling a hydrodynamic flood model with an ABM to evaluate flood impacts on transport networks and road users in a part of the River Thames, London. A novel Risk Priority Index (RPI) was developed to quantify behavioural responses by integrating water depth, demographic vulnerability, and flood-warning information. Social information exchange among agents was simulated using a Watts-Strogatz small-world network. The hydrodynamic model was calibrated and validated using the Sequential Uncertainty Fitting (SUFI-2) approach, achieving R 2 values of 0.91 and 0.89, respectively. Results show that flooding disrupted approximately 9% of the road network, causing trip disruption rates of 51% on weekdays and 45% on weekends. The proposed framework supports the identification of vulnerable populations and critical roads for flood risk management.