The Potential of Agent Based Models for Testing City Evacuation Strategies Under a Flood Event

Abstract

This paper explores the uses of Agent Based Models (ABM) and its potential to test large scale evacuation strategies in coastal cities under threat of an imminent flooding due to extreme hydro-meteorological events. The first part of the paper is an introduction to the field of complex adaptive systems (CAS) and the principles and uses of ABM in this field. It is also presented the benefits and limitations of such models. The second part of the paper focuses on the theory used to build the ABM. For this study, theories and frameworks of human behaviour and disaster psychology were used. To feed the ABM model qualitative and quantitative attributes or characteristics of human beings are abstracted from literature review, fieldwork and expert's knowledge. The third part of the paper shows the methodology used to build and implement the ABM model using Repast Symphony, a Java based modelling system. The results of the initial experiments implemented in a region of the city of Marbella, southern Spain, are presented and discussed. The preliminary results are promising to further enhance the development of the model and its implementation and testing at full city scale.