Revealing important nocturnal and day‐to‐day variations in fire smoke emissions through a multiplatform inversion

Abstract We couple airborne, ground‐based, and satellite observations; conduct regional simulations; and develop and apply an inversion technique to constrain hourly smoke emissions from the Rim Fire, the third largest observed in California, USA. Emissions constrained with multiplatform data show notable nocturnal enhancements (sometimes over a factor of 20), correlate better with daily burned area data, and are a factor of 2–4 higher than a priori estimates, highlighting the need for improved characterization of diurnal profiles and day‐to‐day variability when modeling extreme fires. Constraining only with satellite data results in smaller enhancements mainly due to missing retrievals near the emissions source, suggesting that top‐down emission estimates for these events could be underestimated and a multiplatform approach is required to resolve them. Predictions driven by emissions constrained with multiplatform data present significant variations in downwind air quality and in aerosol feedback on meteorology, emphasizing the need for improved emissions estimates during exceptional events.