Assessing the Impact of Transportation Diversity on Postdisaster Intraurban Mobility

Abstract

Transportation infrastructure enables mobility in urban communities and impacts the functionality of other infrastructure and services. Natural hazards can cause failures in a transportation system, which can affect mobility and other economic activities in a community. Diversity is recognized as an important factor of resilience in transportation infrastructure, but empirical work linking the two is limited. In this study, the impact of transportation diversity on mobility in New York City after Hurricane Sandy is explored. Transportation diversity, defined as the availability and distribution of modes in a community, is measured by employing a recently developed approach at the zip code level using transportation system GIS data. The geotagged Twitter data of one month before and after Hurricane Sandy in New York City are used to understand mobility patterns before and after this extreme event. The primary locations of individuals are found, and their mobility patterns are subsequently determined by measuring travel distance, the radius of gyration, and mobility entropy one week and one month before and after the hurricane. Individuals are grouped in quartiles based on the transportation diversity of their primary locations, which is determined by two different methods that overcome the lack of transportation system information in call detail records and social media data. The results indicate that the distance, radius, and entropy of all individuals significantly decreased after Sandy. The comparison of the significance of change in mobility metrics by diversity quartiles revealed that one-week distance and radius metrics in low transportation diversity quartiles were changed and statistically significant. Thus, individuals with primary locations in zip codes with higher transportation diversity generally had a higher maintained distance and radius one week after the hurricane. The comparison of diversity quartiles in the one-month analysis showed that the radius of low transportation diversity quartiles was impacted more than high transportation diversity quartiles and was statistically significant. Further, the comparison of results of one week and one month after the hurricane indicated that distance, radius of gyration, and entropy improved after a month as the transportation system was recovering. The findings establish an empirical link between transportation diversity and intraurban mobility in the wake of natural disasters such as Hurricane Sandy and confirm that transportation diversity influences individual postdisaster mobility. In addition, the results contribute to the mobility resilience literature by deepening our understanding of the underlying drivers of changes in human mobility following extreme events. Further, the adopted approach supports pinpointing areas with low transportation diversity, which can enable more targeted infrastructure and urban resilience management.