Using mobility data to identify epidemiologically-relevant geographic regions

Keywords: COVID-19, spatial analysis, mobility, grouping, clustering
Session Type: Virtual Paper Abstract
Day: Friday
Session Start / End Time: 2/25/2022 11:20 AM (Eastern Time (US & Canada)) - 2/25/2022 12:40 PM (Eastern Time (US & Canada))
Room: Virtual 36

Abstract

This research offers an approach to group people into geographic regions that are epidemiologically-relevant and useful for monitoring and surveillance for analyzing infectious-disease outbreaks. I use cellular phone-based mobility data from to construct a bipartite social network (based block group residents' visits to points of interest), which is projected to create a network containing the estimated the number of interactions among people residing in different regions. In my initial analysis, I evaluate the proportion of interactions among residents of geographically-proximate regions, as this is a oft-encountered approach in geographic analysis. I evaluate the overall nature of these interactions at multiple stages throughout the COVID-19 pandemic, showing how contact patterns evolved in response to restrictions meant to reduce social contacts. In the second part of my analysis, I use the interaction network to identify strongly connected regions via community detection algorithms. These regions are epidemiologically-relevant for understanding infection transmission dynamics, as the number of interactions among each region's residents outweigh the number with residents of other regions. The usefulness of these regions are tested by evaluating variations in COVID-19 case rates throughout the pandemic.