Quantifying the role of bromine in the atmospheric oxidation of mercury (Hg)

The cycles of the toxic element mercury (Hg) and bromine (Br) are inextricably linked, since Br radicals are a major oxidant for elemental mercury, Hg(0). Global dispersion of Hg occurs through transport of Hg(0), due to its atmospheric lifetime of 6 months. Upon oxidation of Hg(0) to soluble divalent mercury, Hg(II), deposition will occur on timescales of approximately 1 week. There are many uncertainties associated with atmospheric Hg chemistry, leading to uncertain predictions of its fate and impacts on ecosystems. Here we assess the role of Br in the oxidation of Hg(0) using a chemistry-climate model WACCM and quantify sources of uncertainty due to different factors. Oxidation of Hg(0) by Br is found to dominate near the surface in the Southern Hemisphere midlatitudes, as well as throughout the upper troposphere and lower stratosphere. Elsewhere in the troposphere, the reaction of Hg(0) by hydroxyl (OH) radicals is the primary oxidation pathway. However, these results are highly dependent on the model’s lower troposphere bromine concentrations. Comparing different model versions of GEOS-Chem and WACCM, the chemical lifetime of Hg(0) can vary by a factor of more than 20 in the Southern Hemisphere midlatitudes due to differences in simulated Br concentrations. The models show much closer agreement in their simulated OH concentrations, highlighting the higher uncertainties in Br chemistry. We also explored the uncertainty in Hg reaction rates using global sensitivity analysis in a box model representing WACCM chemistry. Uncertainties in the OH-driven oxidation reactions of Hg(0) dominate uncertainties in the Hg(0) lifetime in the Northern Hemisphere, while the reaction rate of Br with Hg(0) is the key uncertainty over much of the Southern Ocean. We identify ~10 reactions out of the full chemical mechanism of 72 Hg-related reactions that contribute almost all of the variability in outputs, indicating the potential for constructing a simplified mechanism for Hg chemistry. Overall, our results emphasize that predictions of Hg deposition are highly impacted by uncertainties in lower troposphere Br radical concentrations, suggesting that more observational constraints on Br are necessary to improve the accuracy of Hg models.