0. Russell Bullock, Jr.*
Atmospheric Sciences Modeling Division
Air Resources Laboratory
National Oceanic and Atmospheric Administration
Mail Drop 80
Research Triangle Park, North Carolina 27711
ton assignment to the National Exposure Research Laboratory, U.S. Environmental Protection Agency
The atmospheric pathway of the global mercury cycle is known to be the primary source
of mercury contamination to most threatened aquatic ecosystems. Current efforts toward
numerical modeling of atmospheric mercury are hindered by an incomplete understanding of
emissions, atmospheric transformations, and deposition processes. While much effort has been
made to quantify the total mass flux of mercury to the atmosphere from various natural and
anthropogenic sources, discrimination of the chemical and physical forms of these emissions is
just beginning in response to early modeling exercises showing this discrimination to be critical
for accurate modeling estimates of the sources responsible for observed mercury deposition. A
similar discrimination of ambient concentrations of mercury throughout the atmosphere is
needed in order to develop a clear understanding of atmospheric transformation processes, both
chemical and physical, which govern the length scale of atmospheric mercury transport and
patterns of its deposition in both wet and dry processes. In this paper, current atmospheric
mercury modeling techniques and the information obtained from them are described. A strategy
for future field research and numerical model development is proposed which is designed to
confidently identify the sources of atmospheric mercury responsible for observed contamination
of aquatic ecosystems.
During the 1960's, detrimental health effects from ingestion of mercury-laden seafood
were highly publicized after direct dumping of methyl-mercury compounds into the waters of
Minamata Bay in Japan led to unhealthy mercury concentrations in the local aquatic wildlife and
a subsequent public health disaster. Since that time, the direct disposal of mercury-laden wastes
into streams and water bodies has been nearly eliminated in most developed countries.
Nonetheless, mercury concentrations in fish tissue have remained alarmingly high in many areas,
including fresh-water species in water bodies where mercury disposal to water has never been
known to occur. Atmospheric deposition of mercury is now believed to be the primary source
of mercury in fish in many locations around the world. Many of the areas where fish
1 - Bullock

consumption advisories for mercury have been issued in the United States and Canada are far
removed from any known sources of atmospheric mercury, indicating that mercury is being
transported long distances through the air pathway.
The conceptual model of the global mercury cycle by Mason et al. (1994) (Figure 1)
deals with the various fluxes of mercury, but very little description of the chemical or physical
forms of that mercury is provided. Current theory suggests that a large part of the current
atmospheric mercury burden is in the form of a worldwide background concentration composed
almost entirely of elemental mercury (Hg) in gaseous form. Little, if any, elemental mercury
is believed to be in particulate form due to its high vapor pressure at normal atmospheric
temperatures. This global reservoir of Hg is believed to have increased in size by about a factor
of three over natural levels due to the relatively recent effects of human industrialization
compared to natural geological sources that have existed for eons (Expert Panel on Atmospheric
Mercury Processes, 1994; Fitzgerald, 1995). The atmospheric burden of oxidized mercury is
believed to be almost entirely in a divalent oxidation state (Hg2+) which can exist in both gaseous
(Hgg+) and particulate (Hg2+) forms depending on the mercury compound in question. The
chemical and physical properties of most Hg2+ compounds indicate that they should deposit from
the atmosphere much more rapidly than Hg through both wet and dry processes, thus explaining
the predominance of Hg in the global atmospheric reservoir.
During the late 1980's, rainfall samples taken in Sweden showed much larger mercury
concentrations than the water solubility of Hg and the local air concentrations of total mercury
would suggest (Iverfeldt, 1991), even in remote areas where anthropogenic sources of more
soluble Hg2+ were not believed to be important. This indicated that some type of oxidation
reaction was responsible for the production of dissolved and/or suspended mercury compounds
in cloud water before deposition to the surface in precipitation. Until recently, a confident
chemical discrimination of gaseous mercury concentrations in ambient air could not be obtained.
New air sampling technology is now being developed which promises to provide a much better
understanding of the reduction-oxidation balance of total gaseous mercury and the atmospheric
processes which might govern it.
Source-Based Modeling
Numerical modeling of atmospheric mercury transport and deposition in Europe by
Petersen et al. (1995) included no treatment of gas-phase mercury chemistry, since these
reactions are generally believed to be very slow, but did include an aqueous chemistry
mechanism which considers a reduction-oxidation balance for mercury in cloud water and
precipitation based on the effects of ozone, sulfite ions and carbon soot on the small amount of
Hg that is dissolved into cloud water from its limited solubility. In this mechanism, the
following aqueous chemical processes are modeled.
2 - Bullock

-	oxidation of dissolved Hg by ozone yielding Hg2"1
-	catalytic reduction of this Hg2+ by sulfite ions
-	adsorption of Hg2+ onto carbon soot particles suspended in the aqueous medium
Petersen et al. (1995) shows that these three simultaneous reactions can be considered in the
formulation of a precipitation scavenging ratio for elemental mercury gas as follows:
tffe0]	K A	I	[c 1
Prec,P = _i . _L_  \o 1  1 + K 
[iV] k- H- 9 1 3
lair	2 "S
k, is the second order rate constant for the aqueous oxidation of Hg by 03,
k2 is the first order rate constant for the aqueous reduction of Hg2+ by sulfite ions,
Hllg is the dimensionless Henry's Law coefficient for Hg,
I o3]aq is the aqueous concentration of ozone,
K3 is a model specific adsorption equilibrium constant (5.0 x 10"6 m4g"'),
[cSoot\aq is the total carbon soot aqueous concentration (g m"3), and
r is the assumed mean radius of soot particles (5.0 x 10~7 m).
Petersen et al. (1995) used this formulation for the precipitation scavenging ratio of Hg
along with prescribed scavenging ratios for Hg^+ and Hg+ in the Lagrangian modeling
framework of the long period model for sulfur used under the European Monitoring and
Evaluation Programme (EMEP) (Eliassen and Saltbones, 1983) to simulate mercury wet
deposition over central and northern Europe. This same technique has also been applied in the
Lagrangian modeling framework of the Regional Lagrangian Model of Air Pollution (RELM AP)
(Eder et al., 1986) to estimate wet deposition patterns for the various forms of mercury across
the continental United States (Bullock et al., 1997; U.S. EPA, 1997b). Each of these Lagrangian
modeling exercises simulated discrete air parcels containing Hg, Hg2+ and Hg2+ traveling with
the wind and depositing mercury in its various forms to the earth's surface without regard to
possible interactions with other parcels from the same or different mercury emission sources.
Gaseous and particulate Hg2+ emissions also contained in the simulated pollutant parcels were
treated separately from Hg in these modeling exercises, thus ignoring possible interactions of
emitted Hg2+ on the reduction-oxidation balance governing Hg wet deposition. In reality, any
Hg2+ dissolved into the water droplet directly from the air could affect the reduction-oxidation
balance between the total concentration of Hg and Hg2+ in the droplet.
These two Lagrangian modeling studies also addressed dry deposition of Hg, Hgj;+ and
Hg2+. However, our present scientific understanding of the dry deposition of mercury in its
various forms is very limited. In each case, dry deposition of Hg was assumed to be very slow
or negligible. Hg2+ was assumed to dry deposit much like common aerosol materials found in
continental background regions and Hg2+ was assumed to dry deposit in the same manner as has
been observed for gaseous nitric acid. The high vapor pressure, low chemical reactivity and low
water solubility of Hg supports the assumption that Hg would not be rapidly absorbed by bare
3 - Bullock

terrain, vegetation or water surfaces. In fact, it is probably emitted from mercuriferous soils and
some types of vegetation growing on them (Hanson et al., 1995). Observed Hg*+ concentrations
represent a very small fraction of the typical particulate mass loading of ambient air. Thus, its
presence should not significantly affect the size distribution or dry deposition rate of the total
particulate loading in air. If Hgp+ is evenly distributed among the ambient particle sizes, the use
of typical dry deposition velocities for general aerosols in the estimation of Hg*+ dry deposition
rates seems justified. However, there is little or no information on the size distribution of Hg+
particles relative to ambient particulate matter. Thus the assumption that Hgj;+ dry deposits much
like common aerosol materials is somewhat speculative and seems deserving of further study.
The use of nitric acid as a surrogate for Hgj;+ seems to be based on the assumption that HgCl2 is
the main component of Hgg+ in the atmosphere. The vapor pressure of HgCl2 has been
determined to be 8.99x103 Pa at 20C (Schroeder and Munthe, 1998), much larger than other
common inorganic mercury compounds. This implies a saturation air concentration for HgCl2
of approximately 1><10"3 g m"3 at a pressure of 1 atmosphere, far above any concentrations of
Hgg+ that would likely exist in ambient air. Although it is known that certain mercury
compounds can adsorb to carbon-rich aerosol matter, its high vapor pressure suggests that HgCl2
should exist primarily as a gas at normal atmospheric conditions, while other inorganic mercury
compounds are much more likely to exist in the solid phase.
Complex chemical interactions between numerous oxidizing substances are known to be
important in the generation of photochemical smog and acidic precipitation. The same degree
of complexity is believed to exist in the chemical processes that affect the oxidation state and
subsequent deposition of atmospheric mercury (Pleijel and Munthe, 1995). As previously
mentioned, a confident chemical discrimination of mercury in air samples has not been possible
until very recently (Lindberg and Stratton, 1998; Stevens et al., 1998). Application of this new
sampling technology in future field research programs will certainly refine our knowledge of
mercury transformation and deposition in the real atmosphere. Proof of the existence of complex
chemical reactions of mercury in air or in cloud water would make the use of discrete-parcel
modeling in Lagrangian frameworks obsolete and would require the use of more sophisticated
Eulerian modeling frameworks. Eulerian models, such as those now being developed for
integrated state of science modeling of acidic precipitation, photochemical smog, and
atmospheric particulate matter (Appleton, 1996), use fixed three-dimensional modeling grids
with pollutant mass transport represented by transfers between stationary grid volume elements.
By using these stationary grid volumes, chemical reactions involving pollutants from all sources
can be simulated.
Recently, a few organizations have begun to develop Eulerian models for atmospheric
mercury in order to simulate complex chemical interactions which may occur between the
various forms of mercury and numerous other chemicals which may facilitate and regulate the
forms of mercury which are eventually produced and deposited to the earth's surface (Pai et al.,
1997; Petersen et al., 1998). However, these models are still limited by a lack of independently
determined process parameters for: 1) the adsorption/desorption of mercury species on
atmospheric aerosols and cloud-water suspended particulate matter, 2) formation of Hgg+ and
Hgp+ species in air, and 3) sources of atmospheric methyl- and dimethyl-mercury (Schroeder and
4 - Bullock

Munthe, 1998). The U. S. EPA, Office of Research and Development is currently developing
a research strategy that includes both atmospheric field research for process definition and
development of an Eulerian modeling framework for process simulation. This research strategy
is also likely to include efforts to better approximate the chemical and physical forms of mercury
emissions from a variety of commercial and industrial source types.
Receptor Modeling
In addition to "source-based modeling" where the entire atmospheric lifetime of a
pollutant is simulated beginning from its emission source, "receptor modeling" has been
performed on a limited basis using measured air and rainfall concentration profiles for mercury
and various other elements, along with predetermined emission profiles for particular source
types to estimate the fraction of the mercury measured at monitoring sites that is attributable to
the particular sources for which emission profiles are available. The chemical mass balance
(CMB) receptor modeling technique uses statistical methods to correlate the relationship of
mercury with other elements in air and rainfall samples to similar relationships in air emissions
from suspected attributable sources. These receptor modeling studies of atmospheric mercury
are the subject of ongoing doctoral thesis research and have not yet been published. However,
comprehensive descriptions of receptor modeling techniques for source apportionment of
atmospheric aerosols can be found in Cooper and Watson (1980), Stevens and Pace (1984), and
Henry et al. (1984).
CMB modeling depends on the assumption that mercury behaves in a similar fashion to
the other elements in the atmosphere between emission from the source and reception at the
monitoring site. This assumption is becoming difficult to defend for all forms of mercury as we
learn more about the reactive behaviors of Hgg+ in comparison to the other elements typically
used for CMB modeling that are almost entirely particulate and relatively inert. Nonetheless,
CMB modeling should be appropriate for the study of particulate forms of mercury over
transport times brief enough to preclude significant adsorption of Hgg+ to aerosol particles or
desorption of Hgp+ back into the gas phase.
Mercury Air Emissions Inventory
The air emissions inventory is certainly one of the most critical elements of any air
quality modeling study. Even novice computer programmers know the rule "garbage in -
garbage out." All air quality simulation models, regardless of their complexity, are still just
computer programs that will not, or at least should not, produce good output data from bad input
data. The basic inputs to air quality simulation models are the time- and space-variant
descriptions of two things; emissions and meteorology. Of course, a description of the initial
concentration fields for all pertinent atmospheric materials is also required, but these initial
concentrations become less important over the period of simulation in many cases. Also, air
5 - Bullock

quality models rarely encompass the entire atmosphere and artificial boundaries for the model
domain must be established. Ideally, these boundaries are placed far enough from the area of
interest to be inconsequential and boundary fluxes can be considered just another form of
emission into, or removal from, the model domain. An accurate simulation of atmospheric
mercury processes using an inaccurate emissions inventory can produce just as much "garbage
out" as an inaccurate simulation using accurate emissions data.
There is reason to believe that current air emission inventories for mercury are still
significantly incomplete and inaccurate. For example, the latest published U.S. EPA estimate
for the air emission flux of total mercury from the U.S. chlor-alkali industry is 6.48x106 g a"1
(U.S. EPA, 1997a). However, a Report to EPA from The Chlorine Institute dated May 8,1998,
indicates mercury usage from 1990 to 1995 by this industrial sector to be 160 tons per year
(1.45x 108 g a"1), yet these facilities are not known to dispose of significant quantities of mercury
to any other media. As another example, the U.S. EPA was recently unable to calculate mercury
flux estimate for tailpipe emissions from mobile sources (diesel- and gasoline-powered, on road,
light-duty vehicles) due to uncertainties about analytical detection limits in test procedures
surveyed (U.S. EPA, 1997a). Preliminary results from an air sampling program conducted in
the Baltimore Harbor Tunnel show transient peaks in the mercury content of tunnel air,
suggesting that mobile sources might be an important source of mercury to the atmosphere
(Keeler, 1998). These are just two examples of how some anthropogenic source types may be
under-represented or simply ignored altogether in the development of anthropogenic mercury
emission inventories with only limited testing of actual suspected sources. Source testing is used
to estimate emission factors for each source category in terms of the mass of mercury emitted
per unit of fuel consumed or product produced. These emission factors are then applied to all
similar sources to arrive at county, state and national totals and, when possible, emission
estimates for individual point sources. This type of emission inventory has been compiled by
only a small fraction of nations worldwide. Obviously, we have a long way to go before global-
scale simulation models of atmospheric mercury can be used with much confidence.
Mercury is naturally a trace component of the atmosphere and is emitted from a number
of natural processes and sources such as volcanoes, geothermal vents, forest fires, and diffusion
from water bodies, soils and vegetation. Mercury is also emitted from land surfaces and water
bodies that have been contaminated by human activity, which complicates the definition of
"natural sources". Emissions of mercury that are not specifically from current human activity
nor from truly natural sources are sometimes referred to as "recycled anthropogenic emissions".
Natural and recycled anthropogenic emissions are thought to be mostly in the form of Hg, but
some Hgg+ and Hgj;+ could also be emitted from high temperature processes like volcanoes and
forest fires. Traces of dimethyl mercury can diffuse into the air above water bodies where
microbial action in sediments has methylated mercury in the water column, but this dimethyl
mercury is thought to quickly convert to Hgj;+ and Hgp+ forms from oxidation by hydroxyl
radicals in air (Niki et al., 1983). Various estimates of the current total global flux of natural
emissions of mercury have been made in recent years (Nriagu and Pacyna, 1988; Nriagu, 1989;
Mason et al., 1994), but it is uncertain to what extent recycled anthropogenic emissions are
included in those estimates. To date, there has not been developed any comprehensive inventory
6 - Bullock

of natural and recycled anthropogenic emissions with the spatial and temporal definitions
required by air quality simulation models.
Even in the absence of a fully descriptive mercury emissions inventory, we have gleaned
scientific knowledge from air quality simulation models that we believe are reasonably accurate
with regard to the atmospheric processes that govern mercury deposition. During the 1980's,
anthropogenic mercury emission inventories generally considered only the total mass of mercury
emitted. The vast majority of mercury in air was known to be in gaseous form and it was
generally assumed that this gas was mostly in the elemental state. Hgg+ could not be measured
in ambient air at that time and was suspected to be rapidly removed or converted to Hg in the
atmosphere. A literature survey at the beginning of the 1990's(Schroederetal., 1991) confirms
the general sense at that time that "Hg should play a pivotal role in the atmospheric chemistry
of this heavy metal." Modeling research has now shown that, while Hg is likely to be pivotal
to the global-scale distribution and eventual deposition of atmospheric mercury, Hgg+ and Hgp+
may be pivotal to the local-to-continental scales of mercury deposition through both wet and dry
atmospheric processes. By using rough estimates of the emissions of Hg, Hg^+ and Hgj;+ from
a variety of industrial sources, the importance of an accurate chemical and physical
discrimination of total-mercury emissions has now been demonstrated (Petersen et al., 1995;
U.S. EPA, 1997b) and quantified (Bullock, 1998). The need for "emissions speciation" for
atmospheric mercury modeling is now generally accepted by the research community and most
of the more recent anthropogenic source inventories recognize its importance (Schroeder and
Munthe, 1998).
Chemical and Physical Speciation of Mercury in Air
As described above, various experiments with atmospheric model simulations using
rather simple descriptions of mercury reactions have shown that the atmospheric transport
distances and deposition patterns of Hg, Hgg+ and Hgj;4 differ quite significantly and that air
emission inventories for mercury must resolve these separate forms. These modeling results also
indicate that atmospheric transformations of mercury between the physical phases and oxidation
states can greatly influence its atmospheric pathway. Thus, for numerical model simulations
to be used with any confidence for source attribution of observed atmospheric mercury
deposition, we must be certain that our definitions of the chemical and physical transformation
processes in air and in cloud water are complete and accurate. These modeled transformation
processes have, thus far, been described only from laboratory experimentation on individual
chemical or physical reactions. In order to assure that our models have treated all important
reactions of mercury and that these treatments are accurate, we must compare modeled
concentrations of the various species of mercury in ambient air to measured values. Until
recently, these measured values could not be obtained.
A method for chemical speciation of mercury in ambient air using a water mist chamber
has been developed (Lindberg and Stratton, 1995). This method was used from 1992 to 1995
in Tennessee and Indiana to characterize the concentrations and behavior of "reactive gaseous
mercury" (RGM). RGM is assumed to' be Hgj;+ compounds such as HgCl2, Hg(OH)2, or
7 - Bullock

compounds of other halides which are much more water soluble than Hg, and thus are dissolved
into the water mist used by this sampling method (Lindberg and Stratton, 1998). As expected,
the concentrations of RGM that were measured were only about 3% of the total gaseous mercury
concentration. However, due to the water solubility and chemical reactivity of RGM, this 3%
fraction of gaseous mercury could be responsible for most of the mercury depositing to the
surface in both wet and dry processes. RGM concentration showed high correlations with
temperature, solar radiation, 03, S02, and total gaseous mercury (TGM), much like other
regional air pollutants. These observations strongly suggest that the complete set of processes
that affect mercury transformation in the atmosphere are much more numerous and complex than
the early process formulations of Petersen et al. (1995).
A more chemically selective mercury sampling method has been developed which uses
an annular denuder coated with KC1 which is believed to selectively extract HgCl2 from the air
sampling stream (Stevens et al., 1998). This method is undergoing rigorous testing and it has
been integrated with a currently available continuous TGM monitoring device to provide a
comprehensive ambient air monitoring system that could be used to help characterize mercury
emission speciations and atmospheric transformations downwind. Once these new chemically
discriminate air sampling devices have been put to use in comprehensive field research programs
in tropical, temperate and arctic locations, it is likely that some serious modifications to current
mercury modeling process descriptions will be in order. The confidence with which we can use
simulation models for source attribution should also be greatly increased.
Dry Deposition and Gaseous Exchange
While we can reasonably expect Hgj;+ to dry deposit much like the other particulate
materials in ambient air, there is serious scientific uncertainty about the dry deposition of Hg
and Hgg+. Previous efforts to characterize the dry deposition of TGM have been confounded by
the very different behaviors of its elemental and oxidized components. Also, while particulate
dry deposition can generally be thought of as a one-way downward flux, gaseous exchange of
mercury between the atmosphere and the surface below can be a two-way process (Hanson et
al., 1995). Atmospheric mercury models cannot accurately simulate the transport and wet
deposition of mercury if dry deposition processes are not accurately simulated. Recently
developed mercury speciation methods for ambient air sampling should help greatly to advance
the scientific understanding of dry deposition of mercury in its individual forms over different
types of underlying land forms.
Gaseous exchange of mercury between water bodies and the atmosphere has long been
considered an important issue since mercuiy methylation and bio-accumulation is known to
occur primarily in aquatic ecosystems. Mercury flux across the air/water interface has previously
been modeled using simple Henry's law equilibrium assumptions beginning with the work of
Liss and Slater (1974). However, most observations of dissolved gaseous mercury in water
columns have found super-saturation with regard to Henry's law equilibrium values for Hg.
Various chemical and biological oxidation/reduction and methylation/demethylation processes
may be taking place within lakes and oceans and within their bottom sediments that cannot be
8 - Bullock

represented in purely atmospheric models. A complete water/sediment/biology model for all
water bodies in question may eventually be needed for atmospheric distribution and deposition
of mercury to be fully understood.
The exchange of gaseous mercury with soils is another cross-media issue that is gaining
attention due to some recent experimental measurement programs and the surprising data they
have yielded so far. Highly organic soils have generally been viewed as sinks for atmospheric
mercury, tightly binding mercury to organic materials. However, significant re-emission to air
has been documented in temperate forest soils (Lindberg et al., 1992) and from boreal forest soils
(Schroederetal., 1989; Xiao etal., 1991). Gaseous mercury flux measurement experiments over
mercury-rich geologic deposits in Nevada (Gustin et al., 1996) have recently been repeated and
compared to simultaneous flux measurement by numerous other methods. An unusual rain event
at this normally arid experimental site produced extremely large measured mercury fluxes to the
air that have yet to be explained. Obviously, these types of flux processes need to be explained
and defined in numerical atmospheric models.
Vegetation can also be either a sink or a source of gaseous mercury to the atmosphere.
Hanson et al. (1995) describes evidence for a "compensation point" where higher air
concentrations of Hg result in its deposition to vegetation and lower air concentration result in
emission of Hg from vegetation. More recently, Lindberg et al. (1998) has suggested that
forests are generally air emission sources of Hg. The balance of evidence suggests that mercury
is being deposited to terrestrial ecosystems mostly in oxidized forms and that some portion of
that mercury is converted to Hg and re-emitted to the atmosphere. Obviously this
atmospheric/terrestrial mercury balance needs to be fully understood before it can be
incorporated into atmospheric mercury simulation models. Speciated ambient air sampling to
obtain individual vertical flux estimates for Hg and Hgj;+ will obviously be a critical part of
developing this understanding.
Multi-media Modeling
As the discussion above indicates, the atmospheric pathway of mercury is only one part
of the complete global cycling of mercury. In order to model the atmospheric transport and
deposition of mercury, all fluxes of mercury to and from the surface below must be defined.
This surface may be an industrial plant where the flux of mercury is known to be upward and the
magnitude of that flux of Hg, Hgj;+ and Hgj;+ may be obtained from an emission inventory. On
the other hand, the surface below may be a deciduous forest, in which case a rough estimate of
the direction of the flux may be possible for each mercury species, but the magnitude of these
fluxes may depend on a number of biological and ecological factors that are not considered in
atmospheric models. A simple treatment of terrestrial ecology could be added to the atmospheric
model, but simple models of complex processes often give seemingly accurate results for the
wrong reasons. Improvements to the modeling of this situation obviously depend on the
development of significantly complete and accurate models of both the air and terrestrial media.
The same is true for the modeling of mercury fluxes across the air/water interface over a lake or
ocean. The concept of multi-media modeling seems straightforward until contemporary
9 - Bullock

modeling methods for air quality, water quality, aquatic ecology and terrestrial ecology are
compared. In most cases, they do not even transpose the real world into a numerical framework
in the same fashion, and current efforts to fully integrate media-specific models into a
comprehensive multi-media modeling framework are only at an early design phase. In the mean
time, atmosphere-based modeling'estimates ofthe cross-media exchange of mercury will remain
somewhat speculative.
Atmospheric transport and deposition is known to be a critical link in the chain of events
leading to human exposures to mercury that have alarmed medical and environmental experts.
Numerical modeling of this atmospheric transport and deposition has thus far been performed
with limited information about emissions, atmospheric transformations and deposition processes.
Nonetheless, this early-stage experimental modeling has identified the types of data and process
information that must be obtained for confident modeling assessments of the sources of mercury
responsible for observed ecological contamination. Scientific research has always been a one-
step-at-a-time endeavor, and simulation modeling of atmospheric mercury transport and
deposition has taken a first step. New developments in atmospheric mercury measurement
technology promise to make the next step in numerical modeling a most productive one.
The current scientific understanding of atmospheric mercury suggests that the processes
of nature that govern its deposition around the globe are too complex to be described using
Lagrangian modeling methods which simulate non-interacting parcels of mercury emissions over
the very long transport distances known to be typical of elemental mercury gas. There is solid
evidence that interactions of mercury emissions do indeed occur, both among the various forms
of mercury and with other atmospheric constituents. Thus, more complex and comprehensive
Eulerian-type modeling frameworks for atmospheric mercury are being developed. However,
they will remain largely vacant frameworks without additional field research to provide accurate
definitions of the processes they are capable of simulating. Looking yet another step further in
the future, plans have already been proposed to expand simulation modeling domains to
encompass the air, water and terrestrial media in one comprehensive multi-media modeling
framework. If such a modeling framework were developed today for the global mercury cycle,
it would stand even more vacant of process information than the Eulerian air-media frameworks
which are closer at hand. We do not yet understand many of the governing factors for cross-
media fluxes and bio-accumulation of mercury. Nonetheless, we do know that the global
mercury cycle is a multi-media process. The more we can make our numerical realizations
comparable to nature, the closer we are to the ultimate goals of simulation modeling.
10 - Bullock

Local & Regional
Deposition Hg
Appleton, E. L. (1996) Air quality modeling's brave new world. Environmental Science and Technology
Bullock, O. R., Jr., W. G. Benjey and M. H. Keating (1997) The modeling of regional-scale atmospheric mercury
transport and deposition using RELMAP. In: Joel E. Baker, editor. Atmospheric deposition of contaminants
to the Great Lakes and Coastal Waters. SETAC Press, Pensacola, pp. 323-347.
Bullock, O. R., Jr. (1998) Lagrangian modeling of mercury air emission, transport and deposition: An analysis
of model sensitivity to emissions uncertainty. Science of the Total Environment 213:1-12.
Cooper, J. A. and J. G. Watson (1980). Receptor oriented methods of air particulate source apportionment.
Journal of the Air Pollution Control Association 30:1116-1125.
Eder, B. K., D. H. Coventry, T. L. Clark, and C. E. Bollinger (1986) RELMAP: A regional Lagrangian model
of air pollution - users guide. Project Report, EPA/600/8-86/013, U.S. Environmental Protection Agency,
Research Triangle Park, NC.
Eliassen, A. and J. Saltbones (1983) Modelling oflong-range transport ofsulphur over Europe: a two-year model
run and some model experiments. Atmospheric Environment 17:1457-1473.
Expert Panel on Mercury Atmospheric Processes (1994) Mercury Atmospheric Processes: A Synthesis Report.
EPRI Report No. TR-104214.
Fitzgerald, W. F. (1995) Is mercury increasing in the atmosphere?: The need for an atmospheric mercury network
(AMNET). Water, Air and Soil Pollution 80:245-254.
Gustin, M. S., G. E. Taylor, T. L. Leonard and R. E. Keislar (1996) Atmospheric mercury concentrations
associated with geologically and anthropogenically enriched sites in central western Nevada. Environmental
Science and Technology 30:2572-2579.
Hanson, P.J., S.E. Lindberg, T.A. Tabberer, J.G. Owens and K.-H. Kim (1995) Foliar exchange of mercury vapor:
Evidence for a compensation point. Water, Air and Soil Pollution 80:373-382.
Henry, R. C., C. W. Lewis, P. K. Hopke and H. J. Williamson (1984) Review of receptor model fundamentals.
Atmospheric Environment 18:1507-1515.
Iverfeldt, A., (1991) Occurrence and turnover of atmospheric mercury over the nordic countries. Water, Air and
Soil Pollution 56:251-265.
Keeler, G.J. (1998) An oral presentation of recent mercury monitoring activities at the Science Experts Workshop
on Mercury in North America, Las Vegas, Nevada, U.S.A., October 6-8, 1998.
Lindberg, S. E., T. P. Meyers, G. E. Taylor, R. R. Turner, and W. H. Schroeder (1992) Atmosphere-Surface
Exchange of Mercury to a Forest: Results of Modelling and Gradient Approaches. Journal of Geophysical
Research 97(D2):2519-2528.
Lindberg, S. E., and W. J. Stratton (1995) Use of a refluxing mist chamber for measurement of gas-phase
mercury(II) species in the atmosphere. Water, Air and Soil Pollution 80: 1269-1278.
12 - Bullock

Lindberg, S. E., and W. J. Stratton (1998) Atmospheric mercury speciation: concentrations and behavior of
reactive gaseous mercury in ambient air. Environmental Science and Technology 32:49-57.
Lindberg, S. E., P. J. Hanson, T. P. Meyers and K.-H. Kim (1998) Air/surface exchange of mercury vapor over
forests - the need for a reassessment of continental biogenic emissions. Atmospheric Environment 32:895-
Liss, P. S. and P. G. Slater (1974) Flux of gases across the air-sea interface. Nature 247: 181-184.
Mason, R.P., W.F. Fitzgerald and F.M.M. Morel. (1994) The Biogeochemical Cycling of Elemental Mercury:
Anthropogenic Influences. Geochimica et Cosmochimicia Acta. 58(15):3191-3198.
Niki, H., P. D. Maker, C. M. Savage and L. P. Breitenbach (1983) A long-path fourier transform study of the
kinetics and mechanism for the HO-radical initiated oxidation of dimethyl mercury. Journal of Physics and
Chemistry 87: 4978-4981.
Nriagu, J. O. (1989) A global assessment of natural sources of atmospheric trace metals. Nature 338:47-49.
Nriagu, J. O. and J. M. Pacyna (1988) Quantitative assessment of worldwide contamination of air, water and soils
by trace metals. Nature 333:134-139.
Pai P., P. Karamchandani and C. Seigneur (1997) Simulation of the regional atmospheric transport and fate of
mercury using a comprehensive Eulerian model. Atmospheric Environment 31:2717-2732.
Petersen, G., A. Iverfeldt and J. Munthe (1995) Atmospheric mercury species over Central and Northern Europe.
Model calculations and comparison with observations from the Nordic Air and Precipitation Network for 1987
and 1988. Atmospheric Environment 29:47-68.
Petersen, G., J. Munthe, K. Pleijel, R. Bloxam and A. Vinod Kumar (1998) A comprehensive Eulerian modeling
framework for airborne mercury species: Development and testing of the tropospheric chemistry module
(TCM). Atmospheric Environment 32:829-843..
Pleijel, K. and J. Munthe (1995) Modeling the atmospheric chemistry of mercury: The importance of a detailed
description of the chemistry of cloudwater. Water, Air and Soil Pollution 80:317-324.
Schroeder, W. H., J. Munthe and O. Lindqvist (1989) Cycling of mercury between water, air and soil
compartments of the environment. Water, Air and Soil Pollution 48:337-347.
Schroeder, W. H., G. Yarwood and H. Niki (1991) Transformation processes involving mercury species in the
atmosphere - results from a literature survey. Water, Air and Soil Pollution 56:653-666.
Schroeder, W. H. and J. Munthe (1998) Atmospheric mercury - an overview. Atmospheric Environment 32:809-
Stevens, R. K., F. H. Schaedlich, D. R. Schneeberger, E. Prestbo, S. Lindberg, G. Keeler (1998) Automated
instrument designed to measure Hg and HgCl2 in near real time: design and operational characteristics.
Thirteenth Annual International Symposium on Measurement of Toxic and Related Air Pollutants, Cary, North
Carolina, September 1-3, 1998.
Stevens, R. K. and T. G. Pace (1984) Overview of the mathematical and empirical receptor models workshop
(Quail Roost II). Atmospheric Environment 18:1499-1506.
13 - Bullock

U.S. EPA (1997a) Mercury Study Report to Congress. Volume II: An Inventory of Anthropogenic Mercury
Emissions in the United States. Report number EPA-452/R-97-004
U.S. EPA (1997b) Mercury Study Report to Congress. Volume III: Fate and Transport of Mercury in the
Environment. Report number EPA-452/R-97-005
Xiao, Z. F., J. Munthe, W. H. Schroeder and O. Lindqvist (1991) Vertical fluxes of volatile mercury over forest
soil and lake surfaces in Sweden. Tellus 43B:267-279.
14 - Bullock

2 .
Current methods and research strategies for modeling
atmospheric mercury
0. Russell Bullock, Jr.
Same as Block 12
U.S. Environmental Protection Agency
Office of Research and Development
National Exposure Research Laboratory
Research Triangle Park, NC 27711
Preprint, FY-99
The atmospheric pathway of the global mercury cycle is known to be the primary source of
mercury contamination to most threatened aquatic ecosystems. Current efforts toward
numerical modeling of atmospheric mercury are hindered by an incomplete understanding of
emissions, atmospheric transformations, and deposition processes. While much effort has been
made to quantify the total mass flux of mercury to the atmosphere from various natural and
anthropogenic sources, discrimination of the chemical and physical forms of these emissions
is just beginning in response to early modeling exercises showing this discrimination to be
critical for accurate modeling estimates of the sources responsible for observed mercury
deposition. A similar discrimination of ambient concentrations of mercury throughout the
atmosphere is needed in order to develop a clear understanding of atmospheric transformation
processes, both chemical and physical, which govern the length scale of atmospheric mercury
transport and patterns of its deposition in both wet and dry processes. In this paper,
current atmospheric mercury modeling techniques and the information obtained from them are
described. A strategy for future field research and numerical model development is proposed
which is designed to confidently identify the sources of atmospheric mercury responsible for
observed contamination of aquatic ecosystems.