Impact of Wildfires on O 3 and Air Quality Across the United States for 2019–2024 Using Generalized Additive Models

Abstract Wildfire smoke contains various air pollutants and can be transported and dispersed over long distances, deteriorating air quality across a wide area. In 2019–2024, numerous large‐scale wildfires occurred across the United States (US) and surrounding regions. We collected daily PM 2.5 and maximum daily 8‐hr average O 3 (MDA8) data from ∼800 US air quality monitoring sites (April–October, 2019–2024) and incorporated model‐based meteorological data (MERRA‐2 and HYSPLIT) and satellite‐derived parameters (HMS and TROPOMI) for use in a Generalized Additive Model (GAM). Smoke days were identified using HMS and PM 2.5 data, and GAMs were developed individually at each site for the MDA8 using non‐smoke days. Our results demonstrate that incorporating interaction smooth terms improves GAM performance, particularly for high‐altitude mountainous sites where simple models struggle with MDA8 predictions. We then applied the same model to the smoke days to quantify the smoke O 3 contribution to the MDA8 (SMO). We estimated an overall mean SMO of 3.8 ppb across the US, adjusted to 2.8 ppb after bias correction via Empirical Distribution Matching (EDM) for a more conservative estimate. Exceedance days (>70 ppb) with significant SMO across the US totaled 5,393 (∼33%) and 4,099 (∼25%) in the uncorrected GAM and GAM‐EDM, respectively. A key advantage of our approach is its reliance on satellite and model‐based data, ensuring nationwide coverage and consistency. Our method can be broadly applied across the US to improve the understanding of smoke impacts on O 3 and support regulatory decision making.