Evacuation studies have grown in importance over the years as a number of recent emergencies, natural and man-made, have raised the general level of awareness about public responses to the threat or actual occurrence of disasters. An accurate prediction of the rates of evacuation and estimate of the time required to clear a risk area are important planning tools that can mitigate the consequences of an emergency situation.
Traditional evacuation models are predicated on the assumption that everyone would seek the quickest or shortest route to safety, given a life-threatening situation. Observations, however, show that a large percentage of the population does not seek the quickest route to safety.
Parents may move toward dangers to pick up their children from schools. Persons at work may go back home to pick up dependent family members, pets, and personal effects before evacuation begins in earnest. Incorrect assumptions of evacuee behaviors could lead to measures that negatively impact the traffic flow during evacuation.
One effective method to evaluate different evacuation strategies is the use of simulation. Most established simulation models, however, are not built to take the underlying drivers’ social behavior into considerations. In this study, we develop a computerized tool for modeling evacuation dynamics with household consolidation, and then incorporate it into a traffic-simulation software platform. This tool will allow a percentage of the population to consolidate as a family before they evacuate.
After that, a study is conducted to explore the consolidation by household in a network under various demand levels. A mathematical model is presented to capture the underlying relationships among the network components. Next, the traffic volumes entering and leaving the network are investigated to highlight some recommendations about the appropriate implementation of contraflow or staged evacuation strategies.
To help decision makers have a better understanding of the evacuation traffic patterns, this study also examined the influences from spatio-temporal information such as the information dissemination delay, the evacuees’ preparedness time, the numbers and locations of shelters in a network, and demographical information like the number of vehicles in a family.
The proposed research will allow planners to study more realistically the effects of evacuation strategies. The results of studying such household by consolidation behavior are (1) evacuation times are significantly longer compared to the assumption of evacuees taking the shortest route away from danger in low/average demands; (2) with heavy demand, low consolidation rates can produce long evacuation times due to the rapid development of congestion at the network exits; (3) with heavy demand, high consolidation rates could delay the turning point to reverse the inbound lanes to outbound in a contraflow operation; (4) the sequencing of converting inbound lanes to outbound in a contraflow operation should start at the outermost links and work inward, due to extra bi-directional traffic on the network engaged in consolidation activities; (5) information delays and evacuees’ preparedness as a family, coupled with the family consolidation behavior, are important parameters to the evacuation performance; (6) information on demographics and geography also has an important impact on the network evacuation efficiency and evacuees’ social behaviors; more specifically, the evacuation performance is very sensitive to the number of shelters in the network.
Source: University of Maryland
Author: Liu, Ke