Effects of COVID-19 Mobility on Air Quality in the United States
Introduction
The novel coronavirus has upended many parts of American life, as many states reverse reopening plans and schools shift to hybrid and remote learning models for the fall term. As this public health crisis continues amid a larger climate crisis, we sought to explore the effect of reduced human activity on air quality.
Methodology & Findings
To measure human activity and air quality, we used Google Mobility Data and EPA (Environmental Protection Agency) Air Quality data, respectively. Google’s Community Mobility Reports describe individual and collective movement trends over time by geography and across different categories to analyze how people’s movement (e.g. trips to the grocery stores and parks) changed due to COVID-19. The EPA Air Quality data contains information about the change in pollutant quantity over time compared to some baseline quantity.
For statistical analysis, we used overall AQI (Air Quality Index) data over time, focusing on seven counties across the country. Fine particulate matter (PM2.5) was the most consistent pollutant across select cities.
Among the selected counties include Hawaii, New York, San Joaquin Valley, and San Juan. Hawaii was chosen with the hypothesis that there would be little AQI variation due to volcanic activity; New York was chosen due to reports of its initial sharp decrease in air pollution; and San Juan was chosen because it is the home of the Navajo nation and has a history of worsening air quality due to oil and gas pipeline development.
Moreover, Manhattan demonstrates “a kind of reverse environmental justice situation,” according to environmental science professor Daniel P. Schrag, who currently serves as the director of the Harvard University Center for the Environment.
“Most of the tall buildings are heated with fuel oil, especially on the wealthy east side of Manhattan, and the poorer neighborhoods in the outer parts of Queens and Brooklyn and Staten Island are actually cleaner,” Schrag wrote in an email. “It means that air quality in NYC is weakly anti-correlated with income.”
San Joaquin Valley was chosen later for its topography, where AQI levels will be more representative of local air pollution.
“A lot of the bad air quality in the San Joaquin Valley is due to the combination of trucks and fertilizer use on farms,” Schrag said.
The Google Mobility Data was plotted with the AQI data over time from February 16, 2020 compared to the 2019 median levels in each location (see the R Shiny App below).
Each location experienced a significant drop in mobility starting on March 12, 25 days after Feb 16. The counties containing the largest metropolitan areas, such as Boston in Suffolk County and New York City in New York County, saw the most drastic decreases in movement, with New York County peaking at 80% below the baseline on April 11. There are spikes in mobility in short intervals every week, which can be attributed to the fact that individuals were not spending much time at work on the weekends even before the pandemic. In most locations, there was no apparent correlation between the percent change in mobility and percent change in AQI.
Conclusions and Highlights
Because AQI is composed of several different pollutants, its fluctuations depend on many factors. Changes in mobility may reduce emissions from cars and public transportation, but other heavy emitters such as power plants and cargo transport also contribute to AQI. Despite a significant decrease in rush-hour traffic in metropolitan areas during the pandemic, industrial activities have not stopped, resulting in an increase in ground-level ozone from coal plants and oil refineries.
Similarly, agricultural activities contribute to AQI, especially in areas such as San Joaquin Valley that experience extensive fertilizer use from its nearby farms. While COVID-19 caused a temporary lull in individual emissions levels, countless other functions have continued as usual or even in increased volumes, for example PPE transport.
Additionally, yearly and seasonal fluctuations must be taken into consideration. As the year transitions from winter to summer, there are more sunny days. The heat and sunlight encourage ozone production, causing the formation of smog and poorly affecting air quality. As such, typical seasonal changes in atmospheric composition like these partially explain why counties with significant drops in mobility such as Suffolk County, MA, did not experience equal reductions in AQI.
It is crucial to understand the trends in air quality because it is not only an issue of climate change, but also of public health.
It is crucial to understand the trends in air quality because it is not only an issue of climate change, but also of public health. In a nation-wide study, Harvard researchers identified a clear link between long-term exposure to PM2.5 levels and COVID-19 death rates. This correlation exacerbates pre-existing inequalities, as minorities and families in poverty tend to live in the cheaper, more polluted inner city neighborhoods and are thus disproportionately affected by COVID-19.