Agent-based Modeling and Disaster Management

Social simulation is usually used to analyze issues in social science and to study behaviors of people inspecific events. Unlike scientific experiments in other hard sciences, which can be tested in a closed environmentor in a laboratory, social simulation applies computation methods to examine social phenomena.In urban planning, understanding stakeholders—which include residents, businesses, factories, and localgovernments—is one of the important factorsin a successful project. In many cases, these stakeholders areheterogeneous individuals who may have different behaviors. Thus, to effectively solve issues in urban planning,planners need to understand stakeholders. Agent-based modeling (ABM) is widely used to analyze behaviors ofstakeholders under implementation of urban policies, especially in the events of natural disasters, which areconsidered as complex systems. In these analyses, spatial structures of affected areason which stakeholdersare located and interact are a crucial ground of ABM. Together with the development of geographic informationsystems (GIS), the database systems and analyses of ABM become more accurate and reliable, especially onphenomena with the complexity of spatial structures.This review article explains the development and definition of ABM and introduces software and toolkitsfor building an agent-based model, as well as reviews articles and research that use ABM to analyze the issuesin theoretical testing and urban planning. Schelling’s Segregation and Hotelling’s Law models are discussed asexamples of theoretical testing while robbery and driving behavior models are selected as the implications ofABM in urban planning. This article also focuses on the use of ABM on natural disaster policies and managementusing case studies of Japan and the United Kingdom.