Inter-county mobility differences during the COVID-19 pandemic spread in Florida

Keywords: mobility change, COVID-19, Random Forest, geospatial modeling
Session Type: Virtual Poster Abstract
Day: Friday
Session Start / End Time: 2/25/2022 08:00 AM (Eastern Time (US & Canada)) - 2/25/2022 09:20 AM (Eastern Time (US & Canada))
Room: Virtual 38

Abstract

As one of the early hotspots in the COVID-19 pandemic, Florida, has experienced a high incidence of COVID-19, accounting for approximately 10% of the total COVID-19 cases in the U.S in summer 2020, when the case numbers reached a peak. In this presentation, we will discuss research on mobility patterns (specifically the number of inflow trips per census tract per person) for three Florida counties (Miami-Dade, Broward, and Palm Beach) during a three-month period in summer 2020, when COVID-19 case numbers were steeply rising. The inter-county differences in mobility were analyzed using a Random Forest regression model with a set of sociodemographic, travel-related, and built environment variables as well as COVID-19 new cases data. Mobility for 2019 (pre-pandemic) and May-July 2020 were analyzed for these three counties. Inflow trips showed a drastic decrease in mid-June 2020, when COVID-19 case numbers in the counties started to rise dramatically. Results returned by Random Forest model showed that the unique and local characteristics across the 3 different categories of variables for each county resulted in different sets of factor rankings for each county, and where COVID-19 case numbers had different impacts on mobility as well. The modeling approach is generalizable to other locations across the U.S. Examining a set of factors (e.g., employment-related, social behaviors, trips-related factors) that are significant explanatory variables is shown to be informative for public health stakeholders tasked with developing mitigation strategies to impede the spread of COVID-19 across local jurisdictions.