Responses of human activity and in-situ and satellite-observed air quality in U.S. cities to the COVID pandemic

The arrival of the COVID-19 pandemic in the United States prompted lock-down orders in some cities and states, but more muted responses in others, with timing and stringency varying widely. With more Americans working from home, patterns of vehicle activity and electricity use have shifted, but in ways that vary by time and region. Meanwhile, there has been an ongoing trend away from coal-fired power generation that accelerated in 2020. All of this has created a real-world experiment in how air pollution responds to changes in vehicle activity and power generation and associated emissions.

Here, we characterize spatial and temporal patterns of how human activity and air quality in U.S. cities have changed during the COVID pandemic. For vehicle activity, we quantify weekly and monthly changes based on government data for vehicle miles traveled and data from mobile apps. For electricity, emissions from power plants are taken from U.S. Department of Energy and U.S. Environmental Protection Agency data. For ground-level air quality, we examine measurements of nitrogen dioxide, ozone, and fine particulate matter at regulatory monitors. Anomalies in satellite-observed nitrogen dioxide are computed based on data from NASA's OMI instrument. We then examine how spatial and temporal patterns of trends in human activity correlate with the observed trends in air quality, and how changes in satellite-observed nitrogen dioxide correlate with ground-level nitrogen dioxide and two pollutants, ozone and particulate matter, that are influenced by nitrogen oxides. Preliminary results find consistency in the temporal patterns of reductions in vehicle activity measured by various data sources but differences in magnitudes. Nitrogen dioxide levels during the initial lockdowns fell more sharply than ozone and particulate matter, and then rebounded in subsequent months.