Modeling a spatiotemporal distribution of stranded people returning home on foot in the aftermath of a large-scale earthquake

On the day of the Tohoku-Pacific Ocean Earthquake (2011), all rail services in the Tokyo Metropolitan area were paralyzed, amid substantial confusion. Consequently, over 3 million persons were unable to return home on that day. Some, unable to contact their families, felt uneasy and set out to return home on foot. Main roads were seriously congested with cars and people, and the use of emergency vehicles was also obstructed. In this paper, we construct several models that describe decision-making and behavior of individuals attempting to reach home on foot in the wake of a devastating earthquake. The proposed models are calibrated using data taken from questionnaire surveys and person-trip surveys, addressed to occupants of the Tokyo Metropolitan area. We attempt to simulate the movement of individuals having decided to return home on foot and demonstrate the spatiotemporal distribution of those who might be exposed to city fires on their way home in the context of such an event.