A statistical model for determining impact of wildland fires on Particulate Matter (PM₂.₅) in Central California aided by satellite imagery of smoke.

As the climate in California warms and wildfires become larger and more severe, satellite-based observational tools are frequently used for studying impact of those fires on air quality. However little objective work has been done to quantify the skill these satellite observations of smoke plumes have in predicting impacts to PM2.5 concentrations at ground level monitors, especially those monitors used to determine attainment values for air quality under the Clean Air Act. Using PM2.5 monitoring data from a suite of monitors throughout the Central California area, we found a significant, but weak relationship between satellite-observed smoke plumes and PM2.5 concentrations measured at the surface. However, when combined with an autoregressive statistical model that uses weather and seasonal factors to identify thresholds for flagging unusual events at these sites, we found that the presence of smoke plumes could reliably identify periods of wildfire influence with 95% accuracy.