Agent-based models of human response to natural hazards: systematic review of tsunami evacuation

Abstract

This systematic review provides a comprehensive overview of tsunami evacuation models. The review covers scientific studies from the last decade (2012–2021) and is explicitly focused on models using an agent-based approach. The PRISMA methodology was used to analyze 171 selected papers, resulting in over 53 studies included in the detailed full-text analysis. This review is divided into two main parts: (1) a descriptive analysis of the presented models (focused on the modeling tools, validation, and software platform used, etc.), and (2) model analysis (e.g., model purpose, types of agents, input and output data, and modeled area). Special attention was given to the features of these models specifically associated with an agent-based approach. The results lead to the conclusion that the research domain of agent-based tsunami evacuation models is quite narrow and specialized, with a high degree of variability in the model attributes and properties. At the same time, the application of agent-specific methodologies, protocols, organizational paradigms, or standards is sparse. Supplementary Information The online version contains supplementary material available at 10.1007/s11069-022-05643-x.

Crowd navigation for dynamic hazard avoidance in evacuation using emotional reciprocal velocity obstacles

“…We apply the value of V c i from ( 8) in the ERVO model as shown in (1). This approach enables the ERVO model to consider not only the velocities of existing agents as in the RVO model [43], but also the emotional impact caused by panic emotion in multi-hazard situations.…”

“…Crowd evacuation is a complex and challenging problem, particularly in scenarios involving dynamically moving hazards [1]- [5]. The ability to navigate safely and efficiently while avoiding obstacles is critical for successful evacuation [1], [2], [6]- [9].…”

“…Crowd evacuation is a complex and challenging problem, particularly in scenarios involving dynamically moving hazards [1]- [5]. The ability to navigate safely and efficiently while avoiding obstacles is critical for successful evacuation [1], [2], [6]- [9]. This paper proposes a novel application of the emotional reciprocal velocity obstacles (ERVO) method for obstacle avoidance in dynamic hazard scenarios.…”

System identifiability in a time-evolving agent-based model

“…However, existing studies have predominantly focused on compartmental epidemic models described by ordinary differential equations. Identifiability for highly stochastic and highdimensional agent-based models, which are increasingly used in real-world applications [31][32][33][34][35][36][37], has not been systematically evaluated.…”

Agent-Based Modeling of Tsunami Evacuation at Figueirinha Beach, Setubal, Portugal

“…ABMs have been applied to various domains, where they have been utilized to simulate the behaviors and interactions of individuals in an environment [18,19]. For example, ABMs have been used to simulate the dynamics of evacuation from volcanoes [20] and coastal communities in response to tsunami events [21,22]. Some studies were based on past events [23], while others were aimed at risk analysis [24].…”