Modeling and Simulation of Vehicle and Pedestrian Movements in Mudflow Emergency Evacuations

Abstract During emergencies, urban systems—such as transport, production, and services—are disrupted, and users’ behavior changes significantly. An effective response requires preplanned evacuation strategies suited to each emergency. A key challenge in modeling is capturing vehicle–pedestrian interactions under the chaotic conditions that arise during a disaster. This study presents a hybrid methodology to model vehicle–pedestrian interactions, in which most traffic movements are simulated at a microlevel, but arcs with high physical interaction between vehicles and pedestrians are handled using an ad hoc mixed-traffic modeling approach to properly represent pedestrian crossings in front of slowly moving vehicles. The proposed approach realistically models vehicle and pedestrian movements in emergency events under different evacuation scenarios. It is applied in a case study of mudflows in the Salado River basin affecting the Chilean city of Chañaral at the river’s mouth. The proposed methodology aims to improve other approaches for evacuation planning that do not produce sufficient detail in the vehicle–pedestrian interaction within a microsimulation engine to enable effective decision-making. This aspect becomes relevant in emergencies, in which microscopic interactions among entities can create bottlenecks during evacuation that are difficult to capture without sufficient modeling detail. It incorporates multimodal flows and evacuee behavior, and models mixed flows of vehicles and pedestrians sharing the roadways of certain streets on which agglomerations occur. The approach is implemented through the application programming interface (API) of the commercial microscopic traffic simulation software Paramics. The approach allows testing different evacuation strategies, including shelter placement and backup bus services, to accelerate the evacuation process. The different evacuation strategies are analyzed to test and compare their effectiveness in evacuating the total population. In addition, we performed a sensitivity analysis of evacuation times with respect to changes in the population to be mobilized and the modal split between cars and pedestrians during the evacuation.