Changes in Service Elasticity of Travel Demand during Disaster: A New Indicator of Phase Transition

Abstract

The resilience studies often discussed phases of a system under disrupted conditions. However, these phases are arbitrary defined. In this study, we propose to use the elasticity as an indicator to identify the phase transition of transport system conditions under the disaster. The elasticity here refers to the service elasticity of travel demand, which indicates the degree of necessity of transport service. We use the multilevel log-log linear model to calculate the service elasticity and its temporal changes. We also use the change point detection algorithm to identify the timing of phase transition. Results show that the service elasticity becomes more elastic, indicating that the people may tend to stop making non-emergent travel in the first phase of the disruption, but after some time, people start to adapt with the disrupted condition (less elastic). The changes in the elasticity would help policymakers to decide the timing of phase transition of disaster management, e.g., when they can open the road for non-emergent travel needs.