By late March 2020, nearly every state in the US had adopted some form of social distancing measures in order to slow the spread of COVID-19. Almost overnight, with schools and business closed, many people were left with nothing to do but stay home. As a result, cities witnessed a sharp decrease in car traffic. The air quality in these cities seemed to rapidly improve.

But how big of an impact did the coronavirus pandemic have on our air pollution? A new study published in the Bulletin of Atmospheric Science and Technology has quantified exactly that.

Using publicly-available data, they compared the measurements of nitrogen dioxide and PM_2.5 (fine particulate matter) measurements from April 2020 to data from 2015-2019. Nitrogen dioxide, which is emitted from exhaust produced by motor vehicles, is a good proxy for car traffic. PM_2.5 is particulate matter less than 2.5 micrometers in diameter, thinner than human hair. Its small size allows it to travel deep into your lungs and it causes many negative health effects. PM_2.5 is also emitted by vehicles, but its levels are more connected to commercial diesel trucks than passenger cars.

The researchers found that nitrogen dioxide levels decreased during the COVID-19 lockdowns, corresponding to the decreased car travel. However, in contrast to nitrogen dioxide, concentrations of PM_2.5 didn’t seem to be correlated with the decrease in human travel, and were actually more likely to be higher than the daily averages from the previous five years.

The researchers attribute this partly to the fact that commercial trucks didn’t experience the same decline in usage as passenger cars. Similarly, the other common emission sources of PM_2.5, such as power plants and residential heating, also didn’t experience a decline, and may also factor into these findings.