Using Measurements of PM2.5 to Assess the Applicability of Low-Cost Sensor Networks for Understanding the Impact of Wildfire Smoke

Wildland fires emit large quantities of gases and particulate matter that can be transported downwind, impacting the air quality of communities in the path of the smoke. In 2020, as wildfires spread through the Front Range of Colorado, particulate matter smaller than 2.5 μm (PM2.5) in the smoke impacted not only the health, but visibility for those in its path. There are a variety of instrumental techniques used to measure PM2.5. PurpleAir (PA) monitors are popular low-cost sensors (LCS) that are widely used in the United States and abroad, and include interactive online map tools for individual PA owners to make their data available to the public. However, PA monitors are known to overestimate PM2.5 in wildfire smoke. The Environmental Protection Agency (EPA) manages a sparse network of well-calibrated research grade PM2.5 observations and have made recommendations to correct PA PM2.5 for better agreement with EPA observations. Using the PA network of LCS positioned within a 4.5 km radius of EPA managed research-grade sensors, we explore the ability of the LCS network to quantify wildfire PM2.5. We use data from the 2020 East Troublesome and Cameron Peak Fires in Colorado to assess the EPA-recommended PA correction for wildfire smoke. For PA sensors with sufficient data, the EPA correction brought the hourly averaged PA PM2.5 data much closer to the official EPA PM2.5 observations. Results suggest PA monitors can be used reliably to understand air quality in regions impacted by wildfire smoke with the EPA correction applied, enabling the community-level LCS measurements to increase the spatial resolution of atmospheric measurements available for public health and welfare assessments.