Development of multi-scale indices of human mobility restriction during the COVID-19 based on air quality from local...

This study investigated whether variability in air quality, especially related to vehicular emissions, during the COVID-19 pandemic could indicate social distancing. Data from in situ measurements and satellite estimates were used. The study areas were São Paulo, Brazil, and Bologna, Italy. We focused our analysis on NO₂, a combustion-derived pollutant, because of its availability in surface stations and satellite tracking, and because it has a short atmospheric lifetime. The analyses included graphical, statistical, and wavelet transform-based approaches to understand NO₂ concentrations before and during the pandemic. After confirming the reduction in vehicular emissions during the pandemic, we created normalized indices to assess the social remoteness in 2020 in different locations, with a focus on São Paulo and Bologna. These indices were compared to existing indices based on cell phone mobility. The indices proposed in this study suffered high sensitivity to social distance compared to existing ones and helped to understand the actual application of social distance and contamination rates, considering the various dimensions of the problem.