Explaining and Predicting Meteorology-Driven Variability in the Air-Quality Impacts of Biomass Burning Emission

Wildland fires are an increasing risk in California, with extreme fire events regularly threatening the environment, air quality, and human health. To increase forest resilience and reduce the threat of catastrophic wildland fire events, prescribed burning is increasingly recognized as an important forest management policy tool. However, understanding and predicting the air quality and human health impacts of wildland and prescribed fire events on downwind communities can be challenging due in part to the complexity of atmospheric dynamics associated with plume transport. Setting up hypothetical modeled burn scenarios throughout the year utilizing Weather Research and Forecasting coupled with Chemistry (WRF-Chem), we explore the resulting changes in population-weighted exposure for identical case study wildland fire emissions under different seasons and meteorological conditions. We examine the relationship between key meteorological indicators, categorize them by their effective impacts on downwind exposures, and explore their temporal variability using linear general additive models (GAM). Using a selected subset of the most influential meteorological variables we then demonstrate that population weighted exposure (PWE) for potential burn scenarios can be predicted for any given region based on observed or forecast meteorology, providing additional information on air quality risks associated with prescribed or wildland fires.