Investigating the Statistical Relationship between Satellite Data and PM2.5 Monitoring Data to Assess Smoke Exposures from the 2025 LA Wildfires

Abstract The January 2025 Los Angeles (LA) fires burned over 50,000 acres, destroyed over 18,000 structures, and killed 31 people, marking them as one of the worst natural disasters in California’s history. In addition to the visible damage they had wreaked, the fires impacted surface air quality and exposed residents of the LA region to numerous hazardous air pollutants from burning human-made structures and materials. Given the severity of the fires and their location within the most populated county in the U.S., it is critical to understand the extent of ground-level smoke exposure. More specifically for health assessment, it can be necessary to assess the increased exposure to air pollutants associated with smoke events. Here, we investigate the relationship between satellite data and PM2.5 monitoring data to assess smoke exposures from the 2025 LA wildfires throughout the region. Specifically, we use the satellite-based NOAA Hazard Mapping System (HMS) smoke product and surface PM2.5 data from both regulatory monitors and low-cost PurpleAir sensors. We assess the statistical significance of mean PM2.5 concentrations on days with observed HMS smoke plumes aloft compared to the mean PM2.5 of days with no smoke plume. Additionally, we assess for potential smoke-impacted days missed by the HMS satellite smoke product. We estimate that the smoke from these fires led to between 7.21 and 7.44 million people exposed to elevated PM2.5, with approximately 2.18 million people exposed to at least 1 day with a daily mean PM2.5 concentration above 55.5 μg/m3, the U.S. Environmental Protection Agency’s Air Quality Index threshold for unhealthy air. Further, we demonstrate the importance of careful statistical treatment of data, including accounting for multiple hypothesis testing, when using multiple near-real-time data products for smoke exposure assessment. For the LA fires specifically, our work suggests that combining all the HMS smoke product data, including light density polygons, with surface stations can provide valuable information on the spatial extent of ground-level smoke exposure.