A new mechanism for atmospheric mercury redox chemistry

. Mercury (Hg) is emitted to the atmosphere mainly as volatile elemental Hg 0 . Oxidation to water-soluble Hg II plays a major role in Hg deposition to ecosystems. Here, we implement a new mechanism for atmospheric Hg 0 / Hg II redox chemistry in the GEOS-Chem global model and examine the implications for the global atmospheric Hg budget and deposition patterns. Our simulation includes a new coupling of GEOS-Chem to an ocean general circulation model (MIT-gcm), enabling a global 3-D representation of atmosphere– ocean Hg 0 / Hg II cycling. We ﬁnd that atomic bromine (Br) of marine organobromine origin is

Abstract. Mercury (Hg) is emitted to the atmosphere mainly as volatile elemental Hg 0 . Oxidation to water-soluble Hg II plays a major role in Hg deposition to ecosystems. Here, we implement a new mechanism for atmospheric Hg 0 / Hg II redox chemistry in the GEOS-Chem global model and examine the implications for the global atmospheric Hg budget and deposition patterns. Our simulation includes a new coupling of GEOS-Chem to an ocean general circulation model (MITgcm), enabling a global 3-D representation of atmosphereocean Hg 0 / Hg II cycling. We find that atomic bromine (Br) of marine organobromine origin is the main atmospheric Hg 0 oxidant and that second-stage HgBr oxidation is mainly by the NO 2 and HO 2 radicals. The resulting chemical lifetime of tropospheric Hg 0 against oxidation is 2.7 months, shorter than in previous models. Fast Hg II atmospheric reduction must occur in order to match the ∼ 6-month lifetime of Hg against deposition implied by the observed atmospheric variability of total gaseous mercury (TGM ≡ Hg 0 + Hg II (g)). We implement this reduction in GEOS-Chem as photolysis of aqueous-phase Hg II -organic complexes in aerosols and clouds, resulting in a TGM lifetime of 5.2 months against deposition and matching both mean observed TGM and its variability. Model sensitivity analysis shows that the interhemispheric gradient of TGM, previously used to infer a longer Hg lifetime against deposition, is misleading because Southern Hemisphere Hg mainly originates from oceanic emissions rather than transport from the Northern Hemisphere.
The model reproduces the observed seasonal TGM variation at northern midlatitudes (maximum in February, minimum in September) driven by chemistry and oceanic evasion, but it does not reproduce the lack of seasonality observed at southern hemispheric marine sites. Aircraft observations in the lowermost stratosphere show a strong TGM-ozone relationship indicative of fast Hg 0 oxidation, but we show that this relationship provides only a weak test of Hg chemistry because it is also influenced by mixing. The model reproduces observed Hg wet deposition fluxes over North America, Europe, and China with little bias (0-30 %). It reproduces qualitatively the observed maximum in US deposition around the Gulf of Mexico, reflecting a combination of deep convection and availability of NO 2 and HO 2 radicals for second-stage HgBr oxidation. However, the magnitude of this maximum is underestimated. The relatively low observed Hg wet deposition over rural China is attributed to fast Hg II reduction in the presence of high organic aerosol concentrations. We find that 80 % of Hg II deposition is to the global oceans, reflecting the marine origin of Br and low concentrations of organic aerosols for Hg II reduction. Most of that deposition takes place to the tropical oceans due to the availability of HO 2 and NO 2 for second-stage HgBr oxidation.

Introduction
Atmospheric mercury (Hg) cycles between two stable redox forms, elemental (Hg 0 ) and divalent (Hg II ). Most Hg emissions are as gaseous elemental Hg 0 , which is relatively inert and sparingly soluble in water. Hg 0 is oxidized in the atmosphere to Hg II by loss of its two 6s 2 electrons. Hg II salts are water-soluble, partition into the aerosol, and are efficiently removed from the atmosphere by wet and dry deposition, but they can also be reduced back to Hg 0 . Understanding atmospheric Hg redox chemistry is critical for determining where Hg will be deposited globally. Here we propose an updated Hg 0 / Hg II redox mechanism, incorporating recent kinetic data and other observational constraints, for use in atmospheric models. We implement the mechanism in the GEOS-Chem global model (Bey et al., 2001; and examine the implications for the global atmospheric Hg budget and deposition patterns. Previous work assumed gas-phase OH and ozone to be the dominant Hg 0 oxidants (Bergan and Rodhe, 2001;Dastoor and Larocque, 2004;Selin et al., 2007;Bullock et al., 2008;Travnikov and Ilyin, 2009;De Simone et al., 2014;Gencarelli et al., 2014). It is now well established that the corresponding HgOH and HgO products are too thermally unstable to enable oxidation to Hg II under atmospheric conditions Goodsite et al., 2004;Calvert and Lindberg, 2005;Hynes et al., 2009;Jones et al., 2016). OH and ozone could still potentially be important oxidants on aerosols (Ariya et al., 2015). The importance of Br atoms for Hg 0 oxidation was first recognized to explain atmospheric mercury depletion events occurring in the Arctic boundary layer in spring (Schroeder et al., 1998), where sea salt photochemistry provides a large Br source (Fan and Jacob, 1992;Steffen et al., 2008;Simpson et al., 2015). Oxidation of Hg 0 to Hg II by Br is a two-stage exothermic mechanism whereby the second-stage conversion of HgBr to Hg II can be carried out by a number of radical oxidants (Goodsite et al., 2004;. Holmes et al. (2006) first suggested Br atoms could be the main Hg 0 oxidant on a global scale, with Br originating principally from photochemical decomposition of bromoform emitted by phytoplankton (Yang et al., 2005). More recent observations of background tropospheric BrO point to ubiquitous Br radical chemistry in the troposphere (Theys et al., 2011;Wang et al., 2015). Recent aircraft observations of Hg II and BrO over the southeastern US are consistent with Br atoms acting as the main Hg 0 oxidant (Gratz et al., 2015;Shah et al., 2016).
Atmospheric Hg II can be deposited or alternatively reduced back to Hg 0 . Competition between these two processes determines the global patterns of Hg deposition as well as the regional fate of Hg II emitted directly by combustion. Atmospheric reduction mechanisms for Hg II are poorly understood, but photoreduction in aquatic systems has been widely observed (Costa and Liss, 1999;Amyot et al., 2000;Mason et al., 2001). Fast in-plume reduction of Hg II emitted by coal-fired power plants was first reported by  and Lohman et al. (2006), but more recent field observations suggest that on average only 5 % (range 0-55 %) of emitted Hg II is reduced in the plume (Deeds et al., 2013;Landis et al., 2014). Recent aircraft observations  and emission inventories  suggest that previous reports of in-plume reduction of Hg II may reflect in part an overestimate of Hg II emissions.
The speciation of atmospheric Hg II is unknown (Jaffe et al., 2014;Gustin et al., 2015;Jones et al., 2016). It is generally assumed from chemical equilibrium considerations that the main Hg II species are HgCl 2 in the gas phase and Hgchloride complexes in the aqueous phase (Hedgecock and Pirrone, 2001;Selin et al., 2007;Holmes et al., 2009). Hgchloride complexes are relatively resistant to photoreduction (Allard and Arsenie, 1991;Pehkonen and Lin, 1998). Hg II also strongly binds to organic ligands, including to reduced sulfur complexes and carboxyl groups Haitzer et al., 2002;Ravichandran, 2004;Zheng and Hintelmann, 2009). Photoreduction of Hg II bound to dissolved organic carbon (DOC) and other organic matter has been widely reported in aquatic environments (Amyot et al., 1994;Xiao et al., 1995;O'Driscoll et al., 2006;Whalin and Mason, 2006) and could also possibly take place in organic aerosols (OA). Bash et al. (2014) found that including incloud aqueous photoreduction via organic acids based on the mechanism proposed by Si and Ariya (2008) improved the simulation of Hg wet deposition in their regional air quality model. Observed photochemically driven shifts in the abundance of naturally occurring Hg isotopes in the atmosphere and precipitation support the occurrence of aqueous-phase photoreduction involving Hg II -organic complexes (Gratz et al., 2010;Sonke, 2011;Sonke et al., 2015).
The GEOS-Chem model is widely used for global studies of atmospheric chemistry and has been applied in particular to simulations of atmospheric Hg and biogeochemical cycling with a focus on interpreting observations (e.g., Weiss-Penzias et al., 2015;Zhang et al., 2016). The Hg chemical mechanism in the current standard version of GEOS-Chem (www.geos-chem.org) is based on . Our updated mechanism includes more recent information and is applied here with an improved GEOS-Chem simulation of halogen chemistry (Schmidt et al., 2016). As part of this work, we also introduce a new coupling of GEOS-Chem to a global 3-D ocean model (Y. X.  to better interpret observed seasonal variations of atmospheric Hg in the context of both atmospheric chemistry and oceanic drivers of air-sea exchange . Table 1 lists the new chemical mechanism implemented in GEOS-Chem. Table 2 lists reactions proposed in the litera- Henry's law coefficients (M atm −1 ) relating gas-phase and aqueous-phase concentrations are 1.28 × 10 −1 exp(2482(1/T − 1/298)) for Hg 0 (Sanemasa, 1975), 1.1 × 10 −2 exp(2400(1/T − 1/298)) for O 3 (Jacob, 1986), 6.6 × 10 2 exp(5900(1/T − 1/298)) for HOCl (Huthwelker et al., 1995), and 1.4 × 10 6 for Hg II (HgCl 2 ; Lindqvist and Rodhe, 1985). The concentration of OH(aq) is estimated as given in the text. Lifetimes of Hg I species are sufficiently short that steady-state can be assumed at all times.  Lin and Pehkonen (1998); (9) Wang and Pehkonen (2004); (10) Lin and Pehkonen (1997); (11) Buxton et al. (1988). c This is an effective rate constant for intermediate pressures most relevant for atmospheric oxidation and uses a higher level of theory than Goodsite et al. (2004).  .8, 9.10, 7.55, 6.99, 6.11} × 10 −11 cm 3 molecule −1 s −1 . e We assume that the rate coefficient determined for HgBr + HO 2 holds more generally for Y ≡ HO 2 , OH, Cl, BrO, and ClO based on the similarity in their elementary reactions with HgBr and estimates from Wang et al. (2014). f Abstraction alone competes with oxidation to Hg II for HgCl as the rate of HgCl thermal dissociation is negligibly slow (Holmes et al., 2009). g HgCl + Y rate coefficients are assumed to be the same as HgBr + Y . ClHg-Y and BrHg-Y have similar bond energies . h Oxidation of Hg 0 (aq) takes place in clouds only, with concentrations of Hg 0 (aq) and aqueous-phase oxidants determined by Henry's law equilibrium (footnote a). Aerosol liquid water contents under non-cloud conditions are too low for these reactions to be significant. Photoreduction of Hg II (aq) takes place in both aqueous aerosols and clouds, with gas-aerosol partitioning of Hg II as given by Amos et al. (2012) outside of clouds and Henry's law equilibrium for HgCl 2 in cloud (footnote a). The aerosol is assumed aqueous if RH > 35 %. i OCl − has similar kinetics to HOCl but is negligible for typical cloud and aerosol pH given the HOCl / OCl − pK a = 7.53 (Harris, 2002). j Parameterization for photoreduction of aqueous-phase Hg II -organic complexes. Here j NO 2 (s −1 ) is the local photolysis rate constant for NO 2 intended to be representative of the near-UV actinic flux, [OA] (µg m −3 STP) is the mass concentration of organic aerosol under standard conditions of temperature and pressure (p = 1 atm, T = 273 K), and α = 5.2 × 10 −2 m 3 STP µg −1 is a coefficient adjusted in GEOS-Chem to match observed TGM concentrations (see text).

Chemical mechanism
ture but subsequently estimated to be too slow to be of atmospheric relevance. Oxidation of Hg 0 to Hg II in the gas phase involves a two-stage process (Reactions R1, R4) with competing Reactions (R2, R3): Hg I X + Y → Hg 0 + XY (R3) where X is the first-stage Hg 0 oxidant, Y is the second-stage Hg I oxidant, and M is a molecule of air (Goodsite et al., 2004;Donohoue et al., 2006;. The Hg I intermediate has a lifetime of less than a minute anywhere in the troposphere, so local steady state can be assumed. Goodsite et al. (2004) found that HgBr is sufficiently stable for the Br atom to be an effective first-stage Hg 0 oxidant  The Hg I compounds are weakly bound and thermally dissociate too fast to allow for second step of oxidation to Hg II . f Peleg et al. (2015) find a strong correlation between observed nighttime gaseous Hg II and NO 3 radical concentrations in Jerusalem. There is theoretical evidence against NO 3 initiation of Hg 0 oxidation , and NO 3 may instead serve as the second-stage oxidant of Hg I . This would be important only in warm urban environments where nighttime NO 3 is high. g Found to be negligibly slow. h The formation of compounds with structural formulae BrHgO and ClHgO requires a prohibitively high activation energy. Note that e applies to compounds with structural formulae HgBrO, HgOBr, HgClO, and HgOCl. i The direct formation of HgO (s) from this reaction is energetically unfavorable (Calvert and Lindberg, 2005;Tossell, 2006). It has been postulated that intermediate products like HgO 3 could lead to the formation of stable (HgO) n oligomers or HgO (s) via decomposition to OHgOO (g) (Subir et al., 2011), but these must be stabilized through heterogeneous reactions on atmospheric aerosols (Calvert and Lindberg, 2005), and the associated mechanism is unlikely to be significant in the atmosphere (Hynes et al., 2009). A gas-phase reaction has been reported in chamber studies (Hall, 1995;Pal and Ariya, 2004;Sumner et al., 2005;Snider et al., 2008;Rutter et al., 2012), but this is likely to have been influenced by the walls of the chamber (as seen in Pal and  and presence of secondary organic aerosols (in Rutter et al., 2012). j Atmospheric concentrations of I and IO (Dix et al., 2013;Prados-Roman et al., 2015;Volkamer et al., 2015) are too low for these reactions to be significant. k Rates are too slow to be of relevance. l Concentrations of HgSO 3 under typical atmospheric SO 2 levels are expected to be very low. and proposed Y ≡ OH and Br as effective radicals to carry out the second stage of oxidation to Hg II . The Cl atom can also oxidize Hg 0 to produce HgCl (Balabanov and Peterson, 2003;Donohoue et al., 2005).  found that a broad range of radical oxidants could oxidize HgBr and HgCl including Y ≡ NO 2 and HO 2 , the most abundant atmospheric radicals, as well as Y ≡ BrO, ClO, and Cl. Other proposed first-stage gas-phase Hg 0 oxidants appear to be unim-portant, including OH, O 3 , I, I 2 , Br 2 , Cl 2 , BrO, ClO, HCl, HO 2 , and NO 3 (Table 2). Hg 0 is sparingly soluble in water, but it has been suggested that fast aqueous-phase oxidation could take place in cloud droplets (Munthe and McElroy, 1992;Pehkonen, 1997, 1998;Whalin et al., 2007). We include these processes in our mechanism with O 3 (aq), HOCl(aq), and OH(aq) as oxidants.
Strong complexes between Hg II and organic acids may allow electrons to be transferred to Hg II during photolysis . Gårdfeldt and Jonsson (2003) found that direct photolysis of Hg II -oxalate resulted in Hg II reduction, and Si and Ariya (2008) reported the same for Hg II -dicarboxylates. O'Driscoll et al. (2004) reported a linear increase in the efficiency of Hg II photoreduction in freshwater with increasing DOC concentrations. We assume here that Hg II photoreduction is dependent on the local concentration of OA and on the NO 2 photolysis frequency (j NO 2 ) taken as a proxy for the UV actinic flux; we further assume the aerosol to be aqueous only at relative humidity greater than 35 %. The scaling factor α for the reaction (Table 1) is adjusted in GEOS-Chem such that simulated total gaseous mercury (TGM ≡ Hg 0 + Hg II (g)) concentrations match the global mean observed at land stations.

General description
We start from the standard version v9-02 of the GEOS-Chem Hg model (www.geos-chem.org; Amos et al., 2012;Song et al., 2015) and implement a new 2010 inventory for speciated anthropogenic emissions  including contributions from commercial products (Horowitz et al., 2014). We then add original updates, described below, for atmospheric chemistry and atmosphere-ocean coupling. The standard v9-02 model includes cycling of Hg 0 and Hg II between the atmosphere, land, and the surface mixed layer of the ocean Soerensen et al., 2010). The atmospheric model transports Hg 0 and Hg II as separate tracers. Gas-particle partitioning of Hg II is parameterized following Amos et al. (2012) as a thermodynamic equilibrium function of local temperature and mass concentration of fine particulate matter (PM 2.5 ). This parameterization is based on observed relationships between the gas / particle Hg II concentration ratio and PM 2.5 , and it does not resolve effects from particle composition or surface area. Holmes et al. (2010) presented the last detailed global atmospheric budget analysis of Hg in GEOS-Chem, and we use it as a point of comparison for this study. Their model version included Hg 0 oxidation by Br atoms, with Br concentrations specified in the troposphere and stratosphere from the p-TOMCAT and NASA Global Modeling Initiative models, respectively (Yang et al., 2005;Strahan et al., 2007). The standard v9-02 model uses Br concentration fields from the GEOS-Chem tropospheric bromine simulation by Parrella et al. (2012) and features other model updates relative to Holmes et al. (2010), including the gas-particle Hg II equilibrium partitioning described above, a corrected washout algorithm , and improvements to the 2-D surface ocean model (Soerensen et al., 2010). In this work we update the Br concentration fields to a more recent version of GEOS-Chem (Schmidt et al., 2016), as discussed further in Sect. 3.2.
We conduct a 3-year simulation for 2009-2011 driven by GEOS-5 assimilated meteorological data from the NASA Global Modeling and Assimilation Office (GMAO) with native horizontal resolution of 1/2 • × 2/3 • and 72 vertical levels extending up to the mesosphere. The GEOS-Chem simulation is conducted at 4 • × 5 • horizontal resolution by regridding the GEOS-5 meteorological data. We use anthropogenic Hg emissions for the year 2010 from Zhang et al. (2016). The spatial distribution of soil Hg is determined following the method of , with updated soil respiration emissions (910 Mg Hg yr −1 ) for consistency with the mechanistic terrestrial model developed by Smith-Downey et al. (2010). Emissions from snow and re-emission of deposited Hg follow . The model is initialized with a 15-year simulation to equilibrate the stratosphere. We present model results as averages for the three simulation years (2009)(2010)(2011).

Atmospheric chemistry
Hg 0 / Hg II redox chemistry in the standard GEOS-Chem v9-02 model includes Hg 0 oxidation by Br atoms and Hg II incloud photoreduction as described by Holmes et al. (2010), with Br concentration fields from Parrella et al. (2012). Shah et al. (2016) updated that chemistry following  to include second-stage oxidation of HgBr by HO 2 , NO 2 , and BrO, as well as new kinetics for HgBr dissociation. Here we further expand and update the chemistry using the mechanism described in Table 1. Br and Cl radical concentrations are taken from a new GEOS-Chem simulation of tropospheric oxidant-aerosol chemistry by Schmidt et al. (2016). Unlike Parrella et al. (2012), this simulation does not include a bromine radical source from sea salt debromination as recent evidence suggests that BrO concentrations in the marine boundary layer (MBL) are generally sub-ppt (Gómez Martín et al., 2013;Wang et al., 2015). Simulated concentrations of Br in the free troposphere are, on the other hand, a factor of 2 higher than Parrella et al. (2012), reflecting more extensive heterogeneous chemistry. Schmidt et al. (2016) show that their simulation provides a better simulation of aircraft and satellite observations of tropospheric BrO. Stratospheric concentrations of Br and Cl species are from the GEOS Chemistry Climate Model (Liang et al., 2010) and Global Modeling Initiative (Considine et al., 2008;Murray et al., 2012), respectively.
We use the global 3-D monthly archive of oxidant and radical concentrations from the GEOS-Chem simulation of Schmidt et al. (2016) that included detailed gas and aerosol chemistry. We apply diurnal scaling following    (Martin et al., 2006;Hudman et al., 2007;Singh et al., 2007;Lin and McElroy, 2010;Mao et al., 2010;Travis et al., 2016). GEOS-Chem NO 2 columns have also been evaluated against satellite observations over China (Lin et al., 2012), North America (e.g., Lamsal et al., 2014), and Africa (Marais et al., 2012). Fast Hg 0 oxidation in the polar spring boundary layer is simulated by specifying high BrO concentrations when conditions for temperature, sea ice cover, sunlight, and atmospheric stability are met . Monthly mean OA concentrations are archived from a separate v9-02 GEOS-Chem simulation including primary emissions from combustion and secondary production from biogenic and anthropogenic hydrocarbons (Pye et al., 2010). An evaluation of modeled OA against aircraft observations globally is presented in Heald et al. (2011), which showed no systematic bias in remote environments but an underestimate of median concentrations in polluted regions. Aqueous-phase concentrations of Hg 0 , O 3 , and HOCl are determined from the gas-phase partial pressures and Henry's law coefficients (Table 1). We estimate aqueous in-cloud OH concentrations following Jacob et al. (2005) as [OH(aq)] = β [OH(g)], with β = 1 × 10 −19 M cm 3 molecule −1 .

Atmosphere-ocean coupling
GEOS-Chem v9-02 uses a 2-D surface-slab ocean model with no lateral transport and with fixed subsurface ocean concentration boundary conditions (Soerensen et al., 2010). Oceanic circulation and mixing between the surface mixed layer and deeper ocean strongly impact Hg 0 concentrations in the surface ocean through the supply of reducible Hg II , driving changes in oceanic evasion (Soerensen et al., , 2014. Here we present a new two-way coupling of the GEOS-Chem atmospheric Hg simulation to the MITgcm 3-D oceanic general circulation model with embedded ecology (Y. X. . First, the GEOS-Chem Hg model with the 2-D slab ocean is initialized over 15 years of repeated present-day emissions and meteorological data (15 years is required to equilibrate the stratosphere). Starting from these initial conditions, we conduct the GEOS-Chem simulation for the desired period (here 3 years: 2009-2011) and archive monthly surface air Hg 0 concentrations and total Hg II deposition fluxes. We then initialize the Y. X.  ocean model for 20 years with these archived surface boundary conditions to equilibrate the upper several hundred meters of the ocean. The monthly mean surface ocean Hg 0 concentrations from the final year of the simulation are then input to GEOS-Chem, replacing the 2-D slab ocean model concentrations. The 15-year GEOS-Chem simulation is repeated with this new input. This process is iterated until convergence of results is achieved. We find that two iterations are sufficient.
4 Global budget of atmospheric mercury 4.1 Budget and lifetimes Figure 1 (top panel) shows the simulated vertical and latitudinal distributions of annual zonal mean Hg 0 and Hg II mixing ratios. The fraction of tropospheric Hg II in the aerosol phase ranges from less than 10 to more than 90 % . Modeled Hg 0 decreases rapidly in the stratosphere, while Hg II increases with altitude and dominates total Hg in the stratosphere. This vertical structure is driven by chemistry Holmes et al., 2010), and these general vertical trends in the two species are consistent with limited available observations (Murphy et al., 2006;Talbot et al., 2007;Lyman and Jaffe, 2012). The exact partitioning of total Hg in the stratosphere between Hg 0 and Hg II is uncertain, and GEOS-Chem predicts a higher Hg II fraction relative to other models . Photodissociation of gas-phase Hg II halides may be possible at ultraviolet wavelengths (Maya, 1977), but whether this is important in the mid-to-upper stratosphere requires further investigation. The stratosphere in the model accounts for 12 % of total Hg atmospheric mass and is in a very different redox regime from the troposphere. Here we focus on tropospheric budgets, which are most relevant to Hg deposition. The stratosphere is discussed further in Sect. 4.4. Figure 2 shows the global tropospheric Hg budget from our GEOS-Chem simulation, including the rates of the major Hg 0 / Hg II cycling pathways. The global tropospheric Hg reservoir is 3900 Mg, including 3500 Mg as Hg 0 and 400 Mg as Hg II , smaller than the 4500 Mg reservoir in Holmes et al. (2010). In both cases, the reservoir was adjusted through the Hg II photoreduction rate coefficient to match the observed annual mean TGM at long-term monitoring sites, mainly located at northern midlatitudes (Fig. 3). Holmes et al. (2010) used observations for 2000-2008 averaging 1.87 ± 1.00 ng m −3 STP (p = 1 atm, T = 273 K; n = 39 sites), whereas we use observations for 2007-2013 averaging 1.46 ± 0.25 ng m −3 STP (n = 37). Observed atmospheric concentrations in North America and Europe have declined on average by 1-2 % yr −1 since 1990, which has been attributed to the phase-out of Hg from commercial products and declines in coal Hg emissions as a co-benefit of SO 2 and NO x emission controls . Observations used in Holmes et al. (2010) also included very high concentrations at four East Asian sites (3-7 ng m −3 STP; Nguyen et al., 2007;Sakata and Asakura, 2007;Feng et al., 2008;Wan et al., 2009), whereas the more recent observations over East Asia used here (n = 6) are all less than 3 ng m −3 STP (Sheu et al., 2010;Fu et al., 2012a, b;. Only one site (Mt. Changbai, China) has data available during both time periods, and observed TGM concentrations decreased there by a factor of 2 (Fu et al., 2012b). Thus, our atmospheric loadings may be more representative of recent (ca. 2010) conditions. We calculate several atmospheric lifetimes to understand the processes driving Hg deposition: the chemical lifetime of Hg 0 against oxidation, the chemical lifetime of Hg II against reduction, the lifetime of Hg II against deposition, and the lifetime of total gaseous mercury (TGM ≡ Hg 0 + Hg II (g)) against deposition. We find that the lifetime of Hg 0 against oxidation to Hg II in the troposphere is 2.7 months, with Br atom-initiated pathways contributing 97 % of total oxidation. This chemical lifetime is smaller than most prior model estimates, e.g., 6 months in Holmes et al. (2010), which used lower Br concentrations. The addition of NO 2 and HO 2 as second-stage oxidants is also important in increasing the rate of HgBr conversion to Hg II , more than compensating for faster thermal decomposition of HgBr relative to . Our results are consistent with Shah et al. (2016), who estimated a global annual tropospheric Hg 0 lifetime against oxidation to Hg II by Br atoms of 1.2 to 2.8 months, based on their simulation of measurements of Hg II over the southeastern US in summer during the Nitrogen, Oxidants, Mercury and Aerosol Distributions, Sources and Sinks (NO-MADSS) campaign (Gratz et al., 2015). The dominance of NO 2 and HO 2 as HgBr oxidants reflects their high atmospheric abundance and is consistent with results from a box modeling study over the Pacific (Wang et al., 2014). We find Cl atoms contribute less than 1 % of Hg 0 oxidation globally because the supply of Cl atoms (mean tropospheric mixing ratio of 2 × 10 −4 ppt) is limited by fast conversion to the stable reservoir HCl. Aqueous-phase pathways in clouds contribute 2 %. Our work suggests that the dominant Hg II species produced in the gas phase are BrHgONO and BrH-gOOH, though the actual speciation of Hg II in the atmosphere would likely be modified through cycling in aerosols and clouds (Hedgecock and Pirrone, 2001;Lin et al., 2006). We did not consider heterogeneous Hg 0 oxidation on particle surfaces (e.g., Vidic et al., 1998;Flora et al., 1998;Lee et al., 2004) due to inadequate information to formulate atmospheric rates. Further study of heterogeneous Hg 0 oxidation is needed (Ariya et al., 2015).
The vertical distribution of Hg 0 → Hg II oxidation rates is shown in the bottom panel of Fig. 1 in units of number density (relevant to the mass budget) and mixing ratio (relevant to transport). Only 1 % of Hg 0 oxidation occurs in the stratosphere, consistent with previous work . Most Hg oxidation by mass occurs in the extratropical free troposphere, consistent with the Br distribution (Schmidt et al., 2016), and is faster in the Northern Hemisphere because of higher NO 2 concentrations. Hg 0 oxidation in  was fastest in the Southern Ocean MBL due to high Br concentrations from sea salt aerosol debromination, but debromination in that region is unlikely since the sea salt aerosol remains alkaline (Murphy et al., 1998;Alexander et al., 2005;Schmidt et al., 2016). In general, the addition of second-stage oxidants HO 2 and NO 2 shifts Hg oxidation to lower latitudes relative to Holmes et al. (2010). This has implications for Hg deposition, which we discuss in Sect. 5. Hg 0 oxidation in terms of mixing ratio features a secondary maximum in the tropical upper troposphere, where the Br / BrO ratio is high (Parrella et al., 2012;Fernandez et al., 2014).
We find that the shorter chemical lifetime of Hg 0 in our simulation relative to Holmes et al. (2010) must be balanced by faster atmospheric reduction to reproduce observed TGM concentrations. This is implemented by adjusting the photoreduction coefficient α in Table 1. The resulting mean lifetime of Hg II against reduction in the troposphere is 13 days. Shah et al. (2016) similarly found that faster Hg 0 oxidation as needed to match NOMADSS observations required faster Hg II reduction, with a tropospheric Hg II lifetime against reduction of 19 days. The lifetime of tropospheric Hg II against deposition is relatively long, 26 days (see Fig. 2), because most Hg II production occurs in the free troposphere, where precipitation is infrequent (Fig. 1). Hg II reduction in the model is faster in the Northern Hemisphere than in the Southern Hemisphere because of higher OA concentrations. Hence Hg II concentrations are higher in the Southern Hemisphere than in the Northern Hemisphere (Fig. 1).
We find here that the Hg II lifetime against reduction is shorter than against deposition, emphasizing the importance of reduction in controlling the atmospheric Hg budget. By contrast, Holmes et al. (2010) found an Hg II tropospheric lifetime of 50 days against reduction and 36 days against deposition, which led them to conclude that no reduction would be needed if Hg 0 oxidation kinetics were decreased within their uncertainty. This is no longer possible in our simulation because of the much faster Hg 0 oxidation. We conclude that atmospheric Hg II reduction must take place, although the mechanism is not established. With Hg II reduction, the overall lifetime of tropospheric TGM against deposition in our simulation is 5.2 months, similar to the estimate of 6.1 months in Holmes et al. (2010). We discuss the consistency of this estimate with observations in the next section. Beyond this simulation of the mean, we successfully reproduce the observed standard deviation of TGM concentrations, which places an independent constraint on the atmospheric lifetime of Hg against deposition (Junge, 1974;Hamrud, 1983). This constraint had previously been expressed in terms of the interhemispheric gradient of TGM from ship cruises, leading to TGM lifetime estimates ranging from 4.4 months to 2 years (Slemr et al., 1981;Fitzgerald et al., 1983;Lindqvist and Rodhe, 1985;Lamborg et al., 2002). However, we find the interhemispheric gradient in surface concentrations is not a sensitive diagnostic of lifetime because surface atmospheric Hg in the Southern Hemisphere is controlled more by atmosphere-ocean exchange than by transport from the Northern Hemisphere. For example, changing the simulation of atmosphere-ocean exchange from the slab ocean to the MITgcm (Sect. 3.3) increases the TGM lifetime while also increasing the interhemispheric gradient due to changed ocean emissions. We conducted a sensitivity simulation doubling the tropospheric TGM lifetime in the model (to 10.4 months) by decreasing the rate of Hg 0 oxidation. This produced less than 20 % of a decline in the interhemispheric gradient, but the relative standard deviation of TGM concentrations across land stations decreased by 40 %. Thus the overall variability of TGM concentrations provides a more sensitive constraint on the TGM atmospheric lifetime, while the interhemispheric gradient can be misleading. The model's ability to reproduce the observed standard deviation across land sites supports our simulated TGM atmospheric lifetime of 5.2 months.

Seasonality
Figure 4 compares simulated and observed seasonal cycles of TGM concentrations at land stations at northern midlatitudes (30 to 60 • N) and in the Southern Hemisphere (see Fig. 3 for site locations). Also shown are results from a sensitivity simulation with the 2-D surface-slab ocean model (Soerensen et al., 2010). The observed seasonal cycle is successfully reproduced in the Northern Hemisphere. The February maximum and September minimum are driven in the model in part by Hg 0 oxidation and in part by ocean evasion. Seasonality in radical oxidant concentrations outweighs temperature effects, leading to fastest oxidation in summer and slowest in winter. Modeled oceanic Hg 0 evasion from the North Atlantic Ocean peaks in winter and early spring, because higher wind speeds increase mixing and entrainment of reducible Hg II from subsurface waters and enhance the rate of air-sea exchange, and Hg removal via particle settling during this time is low (Sorensen et al., 2010(Sorensen et al., , 2013. Similar to Song et al. (2015), we find that the 2-D slab ocean overestimates the observed seasonality at northern midlatitude sites (Fig. 4). Atlantic and Pacific Ocean seawater Hg 0 measurements indicate this overestimate results from oversimplified ocean physics and the fixed boundary condition for subsurface ocean concentrations in the slab ocean model (Soerensen et al., 2014). Previous work shows one-way coupling to the MITgcm ocean simulation improves simulated surface ocean Hg 0 concentrations compared to observations − Southern Hemisphere Figure 4. Mean seasonal variation and spatial standard deviation of total gaseous mercury (TGM) concentrations for northern midlatitude sites (see Fig. 3) and Southern Hemisphere sites (Amsterdam Island and Cape Point). Observations are compared to our standard GEOS-Chem simulation coupled to the MITgcm 3-D ocean model and to a sensitivity simulation coupled to the 2-D surfaceslab ocean model in GEOS-Chem. (Soerensen et al., 2014). Here, the two-way coupling to the MITgcm also improves the seasonality in atmospheric Hg.
In the Southern Hemisphere, Amsterdam Island in the Indian Ocean and Cape Point on the South African coast have similar monthly average concentrations and show no significant seasonal variation (Angot et al., 2014;Slemr et al., 2015). By contrast, the model at those sites has strong seasonality, as previously shown for the standard model version v9-02 . We find a modest improvement in moving from the slab ocean simulation to the MITgcm, but the seasonal bias is still large. Long-range transport of atmospheric Hg from Antarctica could influence observed seasonality at these sites (Angot et al., 2016). Capturing the seasonality of atmospheric Hg in the Arctic using GEOS-Chem required parameterization of unique sea ice, oceanic, and riverine dynamics (Fisher et al., , 2013Y. Zhang et al., 2015); a similar analysis for the Southern Ocean region, with its distinct productivity dynamics impacting Hg cycling (e.g., Cossa et al., 2011;Gionfriddo et al., 2016), has not yet been performed.

Vertical distribution and the stratosphere
Hg 0 concentrations in the model drop rapidly above the tropopause (Fig. 1), consistent with observations (Talbot et al., 2007;Slemr et al., 2009). Modeled Hg 0 oxidation in the lower stratosphere is almost exclusively from Br atoms. There is no significant Hg II reduction there because OA concentrations and relative humidity are very low. Extensive TGM observations in the lowermost stratosphere at northern extratropical latitudes are available from the Civil Aircraft for the Regular Investigation of the atmosphere Based on an Instrument Container (CARIBIC) program, which collects measurements aboard commercial aircraft (Slemr et al., 2009. The early stratospheric data were biased low, and we focus on corrected data available for April 2014-January 2015 . TGM in the stratosphere is expected to be mainly Hg 0 because Hg II is incorporated into aerosol (Murphy et al., 2006;Lyman and Jaffe, 2012). Figure 5 compares the stratospheric TGM concentrations measured in CARIBIC to model values sampled along the flight tracks. The CARIBIC data also include ozone (O 3 ) and CO concentrations (Brenninkmeijer et al., 2007). Here we use O 3 concentration as a chemical coordinate for depth into the stratosphere and exclude tropospheric data as diagnosed by [O 3 ] / [CO] < 1.25 mol mol −1 . We correlate the logarithm of TGM concentrations, as a measure of first-order loss, to the O 3 concentrations. The observations reach higher ozone than the model, indicating that they sample air with greater stratospheric influence. The model underestimates the log(TGM)-ozone slope by a factor of 2. To test whether this underestimate is due to slow Hg 0 oxidation, we performed a sensitivity simulation doubling the stratospheric Hg 0 oxidation rate (green line in Fig. 5). The model slope increased by only 32 %. This suggests that the log(TGM)-ozone relationship in the lowermost stratosphere is set in part by mixing rather than solely by chemistry (Xiao et al., 2007). The model underestimate of the log(TGM)-ozone slope could reflect excessive dynamical mixing in the lower stratosphere, a well-known problem in stratospheric transport models (Schoeberl et al., 2003;Tan et al., 2004), or errors in the timescales of air transit across the tropopause, which vary on the order of months to years (Orbe et al., 2014;Ploeger and Birner, 2016).
5 Implications for global Hg deposition Figure 6 compares simulated and observed annual Hg wet deposition at sites in North America, Europe, and China for 2007-2013. The model captures the spatial variability across sites in North America relatively well (r = 0.57). Some of this variability is driven by precipitation amount, which leads to higher modeled deposition along the northwest coast and over the North Atlantic. The maximum along the coast of the Gulf of Mexico is due to deep convection scavenging uppertropospheric air enriched in Hg II (Guentzel et al., 2001;Holmes et al., 2016). Previous GEOS-Chem simulations with Br-initiated oxidation of Hg 0 failed to capture this maximum because Hg II production favored higher latitudes Amos et al., 2012). The inclusion of NO 2 and HO 2 as second-stage oxidants in our simulation shifts Hg II production to lower latitudes and matches the general location of the Gulf of Mexico maximum, although the magnitude is still underestimated. Increasing horizontal resolution in the model could improve the definition of the maximum .
Over Europe, deposition is fairly uniform and low in the model and observations. This reflects decreased anthropogenic Hg II emissions in the region from emissions controls on coal-fired power plants (Klimont et al., 2013;Muntean et al., 2014;Zhang et al., 2016). There is no area of frequent deep convection, unlike the Gulf of Mexico for the US.
Observations of Hg wet deposition over China have recently become available (Fu et al., , 2016. Values at urban sites are high, likely reflecting local Hg II emissions, and are correlated with high concentrations of particulate Hg II (Fu et al., 2016). Values at rural sites are much lower and comparable to wet deposition observed in North America and Europe despite higher TGM concentrations in China (see Fig. 3). We simulate these low values successfully in our model, whereas a simulation with the standard GEOS-Chem v9-02 overestimates them by a factor of 4. This reflects in part our dependence of Hg II reduction on OA concentrations, which are particularly high in China (Heald et al., 2011). Decreasing wet deposition in this region in the model has implications for the long-range transport of atmospheric Hg from East Asia (Weiss-Penzias et al., 2007). Figure 7 shows the global distribution of wet and dry Hg II deposition in the model. This deposition is the main Hg source to the open ocean (Sunderland and Mason, 2007;Soerensen et al., 2010). We find that 80 % of global Hg II deposition is to the oceans, compared to 71 % in Holmes et al. (2010), and that 60 % of this deposition is wet. Even though surface Hg 0 concentrations are higher over land, most  Fu et al. (2015Fu et al. ( , 2016 over China (9 sites). In the China panel, circles represent rural sites, and diamonds represent urban sites as identified in Fu et al. (2015Fu et al. ( , 2016. For the MDN and EMEP networks, which collect weekly or monthly integrated samples, we only include sites with at least 75 % of annual data for at least 1 year between 2007 and 2013; for China we only include sites with at least 9 months of data over the 2007-2013 period. MDN data are formally quality-controlled, while for EMEP data we rely on a subset of sites that have been quality-controlled (O. Travnikov, personal communication, 2016  Hg 0 oxidation takes place in the free troposphere and is globally distributed (Fig. 1). Concentrations of Br atoms are higher over oceans, reflecting their marine origin (Schmidt et al., 2016). The dependence of Hg II reduction in our mech-anism on the formation of Hg-organic complexes further shifts Hg II deposition away from continents where OA concentrations are highest (Heald et al., 2011). The change in Br distribution and dominance of HO 2 and NO 2 as second-stage oxidants for HgBr shifts Hg II deposition to lower latitudes relative to the simulation by Holmes et al. (2010), where most deposition occurred over high-latitude oceans. We now find 49 % of global total Hg II deposition is to tropical oceans (30 • S-30 • N). The global ocean is a net sink for total atmospheric Hg of 1700 Mg yr −1 (Fig. 2), with approximately half taken up by tropical oceans. A large Hg II deposition flux to the tropical oceans is suggested by recent cruise observations of high oceanic Hg 0 concentrations across the Intertropical Convergence Zone (ITCZ), which GEOS-Chem previously underestimated (Soerensen et al., 2014).

Conclusions
The atmospheric redox chemistry of mercury (Hg 0 / Hg II ) determines the global patterns of Hg deposition to surface ecosystems. Here we developed an updated chemical mechanism for use in atmospheric models and evaluated it in the GEOS-Chem global model to gain new insights into the global Hg budget and the patterns of Hg deposition. As part of this work we also developed a new coupling between GEOS-Chem and a 3-D ocean general circulation model (MITgcm), resulting in a fully resolved simulation of Hg transport and chemistry in the atmosphere-ocean system.
The updated atmospheric Hg 0 / Hg II redox mechanism includes gas-phase Br as the main Hg 0 oxidant in the troposphere and stratosphere, and second-stage oxidation of HgBr by a number of radical oxidants including NO 2 and HO 2 . Br concentrations in GEOS-Chem are from the recent simulation of Schmidt et al. (2016), and they are higher than previous models and more consistent with recent aircraft and satellite observations of BrO. Atmospheric reduction of Hg II is hypothesized to take place by photolysis of aqueous-phase Hg II -organic complexes. This is parameterized in GEOS-Chem as a function of the local concentration of organic aerosol.
The global mass of atmospheric Hg simulated by GEOS-Chem is 4400 Mg, including 3900 Mg in the troposphere. The tropospheric lifetime of Hg 0 against oxidation is 2.7 months. Observations of the atmospheric variability of total gaseous mercury (TGM ≡ Hg 0 + Hg II (g)) suggest an atmospheric lifetime against deposition of about 6 months. Thus, Hg II must be reduced in the atmosphere, which is consistent with recent observations of atmospheric Hg isotope fractionation. Matching the observed mean surface TGM concentrations in GEOS-Chem implies a tropospheric Hg II lifetime of 13 days against reduction and 26 days against deposition. This results in an overall tropospheric lifetime for TGM of 5.2 months against deposition and enables a successful simulation of the observed relative standard deviation of TGM concentrations across terrestrial sites. The interhemispheric difference in TGM concentrations had previously been interpreted to suggest a longer TGM lifetime against deposition, but we show this is misleading because TGM in the Southern Hemisphere is mostly controlled by oceanic emissions rather than transport from the Northern Hemisphere.
The observed seasonality of TGM concentrations at northern midlatitudes (maximum in February, minimum in September) is reproduced by the model, where it is attributed to photochemical oxidation of Hg 0 and oceanic evasion, both with similar seasonal effects on TGM concentrations. Coupling GEOS-Chem to the MITgcm 3-D ocean model improves simulation of the seasonal amplitude by lowering oceanic evasion over the North Atlantic due to elimination of the static subsurface ocean boundary condition in the GEOS-Chem slab ocean and more realistic ocean physics. Observations at Southern Hemisphere sites show little seasonality, whereas the model features a spring maximum. More work is needed to understand this discrepancy.
Observations show a rapid depletion of TGM above the tropopause, and we examined whether this could provide a test for Hg 0 chemistry. For that purpose we used extensive lowermost-stratospheric observations from the CARIBIC program aboard commercial aircraft and characterized TGM first-order loss as the slope of the log(TGM)-ozone relationship. We find that the model underestimates the observed log(TGM)-ozone slope by a factor of 2 but that this slope is only moderately sensitive to the rate of Hg 0 oxidation and appears to be driven in part by mixing of air parcels with different stratospheric ages. This mixing is likely too fast in GEOS-Chem, which may explain the weaker TGM-ozone slope.
Hg wet deposition fluxes in the model are consistent with spatial patterns observed in North America and Europe. Inclusion of NO 2 and HO 2 as second-stage HgBr oxidants in the model shifts Hg II production to lower latitudes as compared to previous versions of GEOS-Chem and enables the model to capture the location of the observed maximum in wet deposition along the Gulf Coast of the US. However, the magnitude of this Gulf Coast maximum is still underestimated. In rural areas of China, wet deposition is observed to be low despite very high TGM concentrations. This is reproduced by the model, where it is due to fast Hg II reduction, driven in part by very high OA concentrations.
We find that 80 % of global Hg II deposition takes place over the oceans, reflecting in part the marine origin of Br as well as the relatively low marine OA concentrations and hence slow Hg II reduction. More Hg is deposited to the tropical oceans (49 % of total Hg II deposition) than in previous versions of GEOS-Chem, where debromination of sea salt aerosol drove fast Hg 0 oxidation and deposition to the Southern Ocean. Observations of the latitudinal gradient of Hg II wet deposition over the ocean would provide a sensitive test of Hg chemistry and improve understanding of Hg inputs to different ocean regions.
Code and data availability. Model code used in this study will be made publicly available at http://www.geos-chem.org in an upcoming version through the standard GEOS-Chem procedures. Observational data and code to create model-observation comparison figures are available for download at https://github.com/GC-Hg/ HgBenchmark/. Aircraft data from CARIBIC are available through the Global Mercury Observing System (GMOS) at http://sdi.iia.cnr. it/geoint/publicpage/GMOS/gmos_aircraft.zul.
Competing interests. The authors declare that they have no conflict of interest.