Bromine Chemistry in the Present-Day and Pre-Industrial Troposphere: Implications from Modeling and Satellite Observations

Abstract

This dissertation investigates the impact of bromine on tropospheric ozone, OH, and mercury in the preindustrial and present-day atmosphere through use of modeling and observations from satellite. We developed bromine simulation capabilities coupled to oxidant-aerosol chemistry in the GEOS-Chem chemical transport model (CTM). Standard gas-phase mechanisms for bromine chemistry were unable to reproduce recent estimates of tropospheric BrO from satellite. Agreement was improved significantly after imposing HBr+HOBr heterogeneous chemistry in the model. Under present-day conditions, we find that bromine decreases ozone by 6.5%, < 1 – 8 ppb, and global mean OH by 4%. Most ozone loss is due to HOBr production and photolysis, with additional contributions from \(NO_x\) and ozone loss through \(BrNO_3\) hydrolysis. Simulations of the pre-industrial atmosphere are important as baselines for ozone air quality and radiative forcing calculations. However, standard models for the pre-industrial overestimate ozone observations taken a century ago at Montsouris and cannot reproduce the observed aseasonality. We find that bromine chemistry significantly improves this agreement. However, bromine chemistry has negligible impact on the ozone radiative forcing, as concentrations of BrO remain similar. Despite the small change in BrO concentrations, lower ozone in the preindustrial leads to a 40% greater Br mixing ratios. We estimate that this change may have increased the lifetime of atmospheric Hg(0) against oxidation to Hg(II) by 70% since the pre-industrial, making atmospheric mercury a more global pollutant. Additionally, we develop a retrieval algorithm for stratospheric profiles of BrO number density from SCIAMACHY limb near-UV observations. Zonal means of our BrO profile retrievals throughout April 2008 show common features expected from stratospheric photochemistry and dynamics. We apply simulated \([BrO]/[Br_y]\) ratios to the BrO profile retrievals and estimate a stratospheric loading of \(23.5 \pm 6 ppt Br_y\). This supports the 23 ppt stratospheric \(Br_y\) assumed in the satellite-derived climatology of tropospheric BrO that we used to evaluate our GEOS-Chem simulation. Our results imply \(7 \pm 6 ppt\) Br from short-lived bromocarbons, at the higher end of the 3 – 8 ppt range suggested by observations.