Impact of tropical and boreal biomass burning on atmospheric composition

Abstract

Atmospheric trace gases and aerosols emitted from biomass burning significantly influence atmospheric composition globally and locally on short-term and climatological time scales. Global chemical transport models (CTMs) can be used to predict the transport and evolution of biomass burning emissions but have high uncertainties for reasons including computational constraints on model complexity and uncertainties in biomass burning emissions inventories. Chemical data assimilation systems can be used to reduce the impacts of emissions uncertainty and model deficiencies in representing sub-grid scale processes by constraining CTM analyses with satellite atmospheric composition observations. Chemical re-analyses produce best-estimates of the real atmospheric composition through the application of chemical data assimilation. In this dissertation I evaluate how tropical biomass burning emissions impact variability in tropical tropospheric ozone concentrations and improve the representation of tropical and boreal biomass burning emissions in global CTMs. I ask how variability in tropical tropospheric ozone is related to biomass burning emissions, how well global models capture emissions from biomass burning globally, and what the contribution of biomass burning emissions is to global background air quality. First, I use a global chemical reanalysis from the Real-time Air Quality Modeling System (RAQMS) to show that variability in tropical convection and Indonesian biomass burning emissions contribute to observed El Ni˜no Southern Oscillation (ENSO) variability in tropical tropospheric ozone. Next, I show through comparisons with satellite, ground based, and airborne measurements that the biomass burning emissions inventory used in the Unified Forecast System/Real-time Air Quality Modeling System (UFS-RAQMS) global model significantly underestimates CO emissions from Siberian and Indonesian biomass burning in July-September 2019. I demonstrate that assimilation of satellite carbon monoxide (CO) retrievals significantly reduces this bias. I then present results from an iterative finite difference mass balance approach designed to adjust the CO biomass burning emissions. Finally, I show that the adjusted CO biomass burning emissions inventory improves agreement between UFS-RAQMS CO and observations and increases background tropospheric ozone concentrations. Due to projected increases in biomass burning in a changing climate, improved predictions of biomass burning emissions are necessary for providing accurate air quality forecasts.