This talk is to illustrate how we can use ordinary transportation data as well as emerging datasets to advance research in disasters and resilience. I will discuss two case studies. In case study 1, we use an ensemble of datasets including social media data, subway ridership and taxi data to empirically test the long-held assumption in disaster research and practice: warning responses (for notice events) are a linear, sequential process that involves a psychological stage followed by a behavioral stage. This hypothesis is widely adopted by modelers, explicitly or implicitly and underlies many of the simulation models used for emergency response and disaster planning. Our research shows that the classic sequential-response assumption is not universally true. In case study 2, we use subway ridership data to capture an intrinsic property of resilience, that resilience is a dynamic process. By adopting a life-cycle perspective from pre- to during- and post-disaster, we illustrate that ordinary transportation data can be used to capture resilience as a process at a high spatial and temporal level (community level). This study advances research in infrastructure and community resilience by bridging the gap between our conceptual understanding of resilience as a process and its existing, empirical quantification as a static attribute.
Back to Long Programs