Pandemic wave trends in COVID-19 cases, mobility reduction, and climate parameters in major metropolitan areas in the United States.

Abstract

In this paper, we analyzed the association among trends in COVID-19 cases, climate, air quality, and mobility changes during the first and second waves of the pandemic in five major metropolitan counties in the United States: Maricopa in Arizona, Cook in Illinois, Los Angeles in California, Suffolk in Massachusetts, and New York County in New York. These areas represent a range of climate conditions, geographies, economies, and state-mandated social distancing restrictions. In the first wave of the pandemic, cases were correlated with humidity in Maricopa, and temperature in Maricopa and Los Angeles. In Suffolk and New York, cases were correlated with mobility changes in recreation, grocery, parks, and transit stations. Neither cases nor death counts were strongly correlated with air quality. Periodic fluctuations in mobility were observed for residential areas during weekends, resulting in stronger correlation coefficients when only weekday datasets were included in the analysis. We also analyzed case-mobility correlations when mobility days were lagged, and found that the strongest correlation in the first wave occurred between 12 and 14 lag days (optimal at 13 days). There was stronger but greater variability in correlation coefficients across metropolitan areas in the first pandemic wave than in the second wave, notably in recreation areas and parks. In the second wave, there was less variability in correlations over lagged time and geographic locations. Overall, we did not find conclusive evidence to support associations between lower cases and climate in all areas. Furthermore, the differences in cases-mobility correlation trends during the two pandemic waves are indicative of the effects of travel restrictions in the early phase of the pandemic and gradual return to travel routines in the later phase. This study highlights the utility of mobility data in understanding the dynamics of disease transmission. It also emphasizes the criticality of timeline and local context in interpreting transmission trends. Mobility data can capture community response to local travel restrictions at different phases of their implementation and provide insights on how these responses evolve over time alongside disease trends.