Taking the freeway: Inferring evacuee route selection from survey data

Abstract

Effective evacuation management plans can help reduce the negative impacts of disasters. Understanding evacuee travel behavior is critical for the design of evacuation plans. In this paper, we explore which factors contribute to evacuees selecting freeway vs. non-freeway evacuation routes. Freeways are of particular interest due to their ability to evacuate large volumes of traffic. This study used survey data collected for the Hampton Roads region of Virginia. Respondents were asked to provide their preferred route types in the event of a hypothetical Category 4 hurricane evacuation. A mixed (random parameters) logit model was proposed to determine factors that influence evacuees selecting between freeway and non-freeway route. The study found that several factors contribute to evacuees choosing a freeway over other routes. In the descending order of importance (i.e., marginal effects), these factors are: willingness to use the official recommended route, living in a single-family or duplex housing, expected travel time to reach the destination, being employed, and possessing prior evacuation experience. Conversely, a few factors had a negative effect on choosing a freeway. These factors are: willingness to evacuate two days prior to landfall and evacuating to a public shelter or a second home. The findings of this study can help emergency management and transportation agencies design effective traffic control plans to safely evacuate populations during a hurricane.

Keywords

• Hurricane evacuation
• Evacuee route choice
• Logit model
• Random parameter model