Development of a Mass Evacuation Decision Support Tool

Abstract

This study presents a sequential modelling framework of a mass evacuation decision support (MEDS) tool and develops countermeasures to improve mass evacuation processes. The study develops a large-scale traffic evacuation microsimulation model to design, test and evaluate contrasting evacuation scenarios, and evacuation improvement strategies, alternatively countermeasures. One of the notable contributions of this study is that it combines a flood risk and a traffic microsimulation model to assess evacuation operations under floods of different extremes. The microsimulation model uses evacuation demands obtained from a Halifax regional transport network model and considers a dynamic traffic assignment process to capture time-dependent traffic congestion propagation in the network. The study extends the microsimulation model by the inclusion of a transit network to meet the transportation needs of the transit-dependent population during evacuation. The study develops a Mixed Integer Linear Programming-based optimization model to identify marshal point locations and transit routes for a multimodal evacuation. The study leverages the simulation model to analyze further complexity, including vehicle collision-related traffic disruptions during evacuation. A combined Bayes theory and Monte Carlo simulation approach is adopted to determine the spatial and temporal distribution of collision hotspots and their occurrence during evacuation. The application of the developed modules has been instrumental in the design and implementation of countermeasures. The study designs, implements, and evaluates two strategic level countermeasures, namely staged and bus-based evacuation. The study innovates a fuzzy logic-based prioritization process for a staged evacuation based on the zonal vulnerability index. The study develops a Bayesian Belief Network-based vulnerability assessment model that provides a combined zonal vulnerability index considering geophysical, social and mobility characteristics. The study formulates a Knapsack optimization problem following a dynamic programming solution approach to allocate buses to evacuees. The uniqueness of the countermeasures developed in this research is that they ascertain a vulnerability-based prioritization of evacuees for evacuation. The countermeasures offer promising results in terms of evacuation time and network congestion improvements within the traffic microsimulation model. The MEDS tool will be helpful for emergency engineers and planners to understand potential impacts of wide-ranging magnitudes associated with a mass evacuation and plan mitigations proactively.