Flood-ABM: An agent-based model of differential flood effects on population groups and their decision-making processes

Abstract

Flooding is a global concern with wide-ranging impacts on communities, infrastructure, and ecosystems. Its effects are often unevenly distributed, influenced by complex interactions between environmental and social systems. In this paper, we present an agent-based model (ABM) that links hydrological and social systems through a differentiating lens. Our ABM captures individual and collective behaviors across pre-flood, during-flood, and post-flood phases, while differentiating population groups through a socio-economic index to examine how disparities in resources and vulnerability influence decisions related to preparedness, evacuation, coping, and adaptation. Our model innovatively integrates four decision-making theories—Protection Motivation Theory, the Theory of Planned Behavior, Cultural Risk Theory, and Social Capital Theory—in a complex framework that reflects diverse individuals strategies during floods. We find that agents facing intense flood threats respond rapidly with emergency measures—such as swift evacuations and immediate protective actions—even when constrained by resource limitations. In contrast, agents in less critical scenarios exhibit more measured responses, engaging in thorough pre-event planning and gradual evacuations that facilitate smoother adaptations post-flood. Economic impacts quantified by the model demonstrate that widespread business closures significantly reduce earnings, while increased spending on healthcare and evacuation drives up overall costs. During peak flooding, shelter agents suffer wealth declines due to the high costs of providing emergency services, and healthcare agent resources are strained by surging demand as health deteriorates among affected populations. Our results indicate that increased shelter capacity, expedited rescue responses, and enhanced healthcare provision collectively promote post-flood economic stabilization by reducing stranded individuals and mitigating peak-phase financial losses.