An agent-based modeling approach for community resilience assessment accounting for system interdependencies: Application on education system

This paper introduces an agent-based modeling (ABM) approach to model a community as a system of interdependent systems for studying education systems resilience, one of the least-studied components of communities in the quantitative disaster literature. In this ABM approach, autonomous entities, called agents, simulate the components of a system, while internal interactions among them shape the system and external interactions among systems shape the community. To study the education system resilience subject to tornado hazard, a library of agents is proposed including school, household, electric power network, water supply network, and construction companies agents. The proposed ABM approach is applied on the virtual Centerville community. A Monte Carlo sampling analysis for various tornado intensities is conducted to account for and quantify the inherent uncertainties. Moreover, an education system resilience measure based on the education quality and quantity is proposed and computed for Centerville. The probabilities of the education system falling in different resilience levels are also computed for different tornado intensities. Taking advantage of the comprehensive quantitative model proposed, decisions for enhancing the education system resilience can be evaluated, which is demonstrated by assessing the effect of providing backup utilities for schools on the education system resilience.