Flood Emergency Mobility: Archetype-Based Modeling of Human Behavior

Abstract

Evacuation remains a critical challenge for emergency management, particularly in densely populated urban environments where human behavior strongly influences evacuation outcomes. Although substantial progress has been made in hazard modeling, early warning systems, and evacuation logistics, many models still rely on simplified assumptions of homogeneous compliance. Empirical evidence from past disasters shows that awareness of evacuation orders does not necessarily lead to timely or complete evacuation. This paper presents an archetype-based modeling framework to represent heterogeneous human behavior during flood emergencies, capturing coordinated, passive, and chaotic action patterns within an agent-based evacuation model. The framework is applied to a case study in Rotterdam, the Netherlands, considering three dike failure scenarios that generate flash flooding. By integrating behavioral archetypes with high-resolution geospatial data, flood dynamics, and urban infrastructure networks, the model can enable more realistic simulation of evacuation processes and emergent mobility patterns. The results highlight the significant influence of behavioral diversity and infrastructure exposure on evacuation efficiency, advancing human-centered approaches for crisis decision-making and urban governance.