The combination of human mobility data with other datasets can allow a better understanding of unprecedented situations, such as the behavior of populations affected by multiple simultaneous disasters. Natural disasters occurring during COVID-19 are one such circumstance; historical evacuation and shelter-seeking behaviors may be fundamentally altered by the presence of an ongoing pandemic.
The continued effects of COVID-19 pandemic and resource limitations in many public departments necessitate a deeper investigation of the evolution of individual mobility behaviors when disaster strikes. This understanding can allow for allocation of resources where they are most needed in both time and space.
Using information on shelter locations and shelter stays, researchers at CrisisReady have been able to examine differences in shelter-seeking behavior across time during the hurricane season. In combination with investigation of simultaneous mobility patterns, this allows CrisisReady to provide analyses that can influence logistics for disaster response.
Hurricane Laura, which made landfall on August 29, 2020, resulted in dramatically different shelter-seeking behavior from the affected population than Hurricane Ike, despite also being a major hurricane in the same region. Shelter use was substantially lower than in 2008, and shelters further away from the affected region did not report as large an increase in utilization. The use of multiple data streams helped provide a clearer picture; mobility data showed large population movements out of impacted areas ( blog post ), despite the relative lack of shelter use, consistent with reports of increased use of hotels, private vehicles, and other nontraditional evacuation destinations.