Multi-model study of mercury dispersion in the atmosphere: atmospheric processes and model evaluation

Abstract. Current understanding of mercury (Hg) behavior in the atmosphere contains significant gaps. Some key characteristics of Hg processes, including anthropogenic and geogenic emissions, atmospheric chemistry, and air–surface exchange, are still poorly known. This study provides a complex analysis of processes governing Hg fate in the atmosphere involving both measured data from ground-based sites and simulation results from chemical transport models. A variety of long-term measurements of gaseous elemental Hg (GEM) and reactive Hg (RM) concentration as well as Hg wet deposition flux have been compiled from different global and regional monitoring networks. Four contemporary global-scale transport models for Hg were used, both in their state-of-the-art configurations and for a number of numerical experiments to evaluate particular processes. Results of the model simulations were evaluated against measurements. As follows from the analysis, the interhemispheric GEM gradient is largely formed by the prevailing spatial distribution of anthropogenic emissions in the Northern Hemisphere. The contributions of natural and secondary emissions enhance the south-to-north gradient, but their effect is less significant. Atmospheric chemistry has a limited effect on the spatial distribution and temporal variation of GEM concentration in surface air. In contrast, RM air concentration and wet deposition are largely defined by oxidation chemistry. The Br oxidation mechanism can reproduce successfully the observed seasonal variation of the RM ∕ GEM ratio in the near-surface layer, but it predicts a wet deposition maximum in spring instead of in summer as observed at monitoring sites in North America and Europe. Model runs with OH chemistry correctly simulate both the periods of maximum and minimum values and the amplitude of observed seasonal variation but shift the maximum RM ∕ GEM ratios from spring to summer. O3 chemistry does not predict significant seasonal variation of Hg oxidation. Hence, the performance of the Hg oxidation mechanisms under study differs in the extent to which they can reproduce the various observed parameters. This variation implies possibility of more complex chemistry and multiple Hg oxidation pathways occurring concurrently in various parts of the atmosphere.