Abstract
Increasing frequency of extreme winter storms has resulted in costly damages and a disruptive impact on the northeastern United States. It is important to understand human mobility patterns during such storms for disaster preparation and relief operations. We investigated the effects of severe winter storms on human mobility during a 2015 blizzard using 2.69 million Twitter geolocations. We found that displacements of different trip distances and radii of gyration of individuals’ mobility were perturbed significantly. We further explored the characteristics of perturbed mobility during the storm, and demonstrated that individuals’ recurrent mobility does not have a higher degree of similarity with their perturbed mobility, when comparing with its similarity to non-perturbed mobility. These empirical findings on human mobility impacted by severe winter storms have potential long-term implications on emergency response planning and the development of strategies to improve resilience in severe winter storms.
Highlights
Recent developments in information technology have provided an unprecedented amount of crowdsourced spatial-temporal data to study human mobility [1,2,3,4,5]
Human mobility patterns under perturbed states, such as in natural disasters, require a deeper understanding in order to prepare for unfamiliar conditions in the future [6]
Previous research has found that natural disasters, e.g. hurricanes, floods and earthquakes, can cause significant impact on human mobility patterns [5,6,7,8,9,10]
Summary
Recent developments in information technology have provided an unprecedented amount of crowdsourced spatial-temporal data to study human mobility [1,2,3,4,5]. Findings about daily patterned human movements have fundamentally changed our understanding of human mobility at varying spatial scales. Human mobility patterns under perturbed states, such as in natural disasters, require a deeper understanding in order to prepare for unfamiliar conditions in the future [6]. Scholars in the disaster research area have identified scaling laws and evaluated the predictability of human mobility during and after extreme events using mobility patterns from non-perturbed states. Et al [7] used approximately one year of mobile phone data from 1.9 million users, and found that population movements following the Haiti earthquake had a high level of predictability, and destinations were correlated with normal-day mobility patterns and social support structure. A study by Wang and Taylor [5] showed that human mobility was significantly perturbed during Hurricane Sandy and exhibited high
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
