Homeostatic representation for risk decision making: a novel multi-method simulation approach for evacuation under volcanic eruption

All decision-making processes are complex and include a high number of variables. Particularly, decisions related to disaster risk reduction despite having an inherent uncertainty, depend on many factors associated with the inhabitants in danger, risk disaster management authorities, previous experiences, among others. Although considerable research has been carried out to understand the evacuation processes undergoing among different types of natural hazards, few of them consider psychosocial variables under a dynamic approach. This research proposes a novel way of representing decision making using the risk homeostasis theory approach. We developed a multi-method simulation model to gain an understanding of how individuals react and what are the decision-making processes undergoing a volcanic eruption. For this, we developed a system dynamics (SD) model that captures the psychosocial decision-making process of individuals facing a volcanic eruption. This theory proposes that individuals will make decisions to balance their levels of perception and acceptance of risk and that, in addition, they depend on certain motivational and cognitive variables. Once the evacuation decision making was captured by the SD model, we represented the evacuation process through agent-based simulation based on the relationship between evacuation probability and the difference between risk perception and risk acceptance levels. For experimentation and validation methods, we worked with the 2008 Chaitén (southern Chile) volcanic eruption and further population evacuation as a use case scenario.