Airborne Emission Rate Measurements Validate Remote Sensing Observations and Emission Inventories of Western U.S. Wildfires

Carbonaceous emissions from wildfires are a dynamic mixture of gases and particles that have important impacts on air quality and climate. Emissions that feed atmospheric models are estimated using burned area and fire radiative power (FRP) methods that rely on satellite products. These approaches show wide variability and have large uncertainties, and their accuracy is challenging to evaluate due to limited aircraft and ground measurements. Here, we present a novel method to estimate fire plume-integrated total carbon and speciated emission rates using a unique combination of lidar remote sensing aerosol extinction profiles and in situ measured carbon constituents. We show strong agreement between these aircraft-derived emission rates of total carbon and a detailed burned area-based inventory that distributes carbon emissions in time using Geostationary Operational Environmental Satellite FRP observations (Fuel2Fire inventory, slope = 1.33 ± 0.04, <i>r</i>² = 0.93, and RMSE = 0.27). Other more commonly used inventories strongly correlate with aircraft-derived emissions but have wide-ranging over- and under-predictions. A strong correlation is found between carbon monoxide emissions estimated in situ with those derived from the TROPOspheric Monitoring Instrument (TROPOMI) for five wildfires with coincident sampling windows (slope = 0.99 ± 0.18; bias = 28.5%). Smoke emission coefficients (g MJ^-1) enable direct estimations of primary gas and aerosol emissions from satellite FRP observations, and we derive these values for many compounds emitted by temperate forest fuels, including several previously unreported species.