A collection of Economics in Hydrology simulations
Capturing flood-risk dynamics with a coupled agent-based and hydraulic modelling framework
The article presents a coupled agent-based and hydraulic modeling framework to study flood-risk dynamics in settled floodplains. The framework combines:
- Agent-Based Model (ABM): Simulates households and government agents. Households decide—based on protection motivation theory—whether to take no action, implement flood protection, or file complaints to the government. The government decides whether to reinforce levees, balancing cost-benefit analysis and complaints from citizens.
- Hydraulic Modeling: Simulates flood events and levee breaches based on hydrological and topographical data, using the LISFLOOD-FP model.
Key findings and contributions:
- Individual (household) decisions can significantly influence overall community flood risk.
- The model captures levee effect (structural protection can inadvertently increase risk by encouraging more urbanization in floodplains) and adaptation effect (increased risk awareness and adoption of individual mitigation measures reduce damage over time).
- Community and government interactions are dynamic, with feedbacks such as governmental actions responding to collective complaints, and individual actions (like installing barriers) reducing perceived risk.
- Application to a synthetic case of the Po River floodplain in Italy demonstrates how differing parameters and risk perceptions shape risk over a 50-year simulation.
- The resulting framework serves as an explanatory tool for exploring the complex interplay between hydrological processes and social decision making in socio-hydrological flood-risk systems.
Significance:
This framework allows assessment of both spatial and temporal flood-risk patterns by considering both human and physical system evolution, rather than treating vulnerability as static. It has important implications for flood-risk policy and management as it highlights the need to consider human behavioral feedbacks in risk assessment, not just physical flood models or static vulnerability indices.