Environmental Engineering Reference
In-Depth Information
Natural Processes That Emit Mercury into the Air
into surface waters locally through photochemical pro-
cessing. They estimated that the net fl ux out of the marine
boundary layer is only about 1500 tons/yr. Using a three-
dimensional global model that was constrained to observed
mercury concentrations in the ocean, Strode et al. (2007)
calculated a signifi cantly larger net fl ux of 2800 tons/yr. A
large fraction of this, however, was considered to be reemis-
sion of previously deposited mercury.
For forest fi res, most of the mercury released probably
came from recent deposition to the soil and vegetation, and
so is pa r t ly reemission of prev ious a nt hropogenic a nd nat u ra l
emissions (see the section on “Mercury Emissions: Summary,
Uncertainty, and Validation,” below). Weiss-Penzias et al.
(2007) estimated global emissions of mercury from for-
est fi res of 670
A number of natural processes emit mercury to the
atmosphere. The origins of the mercury emitted from
these processes may be purely natural, or they may be a
mix of natural and anthropogenic (reemissions) (Gustin
et al., 2008). Volcanoes, geothermal vents, and naturally
enriched soils release mercury that originated in deep
reservoirs, so can be considered purely natural emissions.
Land, emissions, forest fi res, and ocean emissions are
mixed sources because a signifi cant fraction of their Hg
burden was previously deposited, which includes some
anthropogenic mercury (Selin et al., 2008).
Quantifying natural sources of mercury in the air is
diffi cult because of the large range of source types, natural
variability, concomitance with anthropogenic mercury, and
the global scale of the problem. Nonetheless, it is critical
to quantify these sources, as without an understanding
of natural sources it is diffi cult to understand the scale of
anthropogenic infl uence. Ideally, we would like to quantify
preindustrial mercury emissions as a reference point for
present-day emissions. Since we obviously cannot go back
in time and make direct observations, the estimation of
preindustrial emissions must be done indirectly through
models and sediment records.
Based on a small number of measurements, estimated
mercury releases from volcanoes are 100-800 tons/yr
(Varekamp and Buseck, 1986; Nriagu and Becker 2003;
Pyle and Mather 2003). The only known estimate of global
geothermal emissions (60 tons/year) is from Varekamp and
Buseck (1986) based on average Hg content in hot springs
and a global estimate of convective heat fl ux.
The emission of Hg from naturally enriched soils and
exposed mineral deposits has been studied on small
scales. A very limited number of global estimates have
been made by scaling up the local emissions, but these are
highly uncertain (Gustin et al., 2008). An early estimate
of 500 tons/yr for global mercury emissions from soils
was made by Lindqvist, et al. (1991). The fl ux of mercury
from soil is a complicated function of soil concentration,
light, moisture, temperature, and other factors (e.g. Gustin,
et al., 1999; Ericksen, et al., 2006; Xin, et al., 2007). In one
study, rainfall was found to increase the fl ux of mercury
from desert soils by approximately an order of magni-
tude (Lindberg, et al., 1999), thus showing the complex-
ity of quantifying soil fl ux. Based on recent observations
and an understanding of soil mercury concentrations in
geologically enriched regions across the United States,
Ericksen et al., (2006) estimated a release of 100 tons/yr
from soils in the contiguous United States. Several models
of the global mercury cycle have used a global land fl ux of
500 tons/yr (Seigneur, et al., 2004; Selin et al., 2007).
Oceanic emissions are believed to be the largest compo-
nent of all natural emissions. Mason and Sheu (2002) esti-
mated that oceanic emissions of mercury are 2600 tons/yr;
however, a signifi cant fraction of this is “recycled: back
330 tons/yr based on observations of the
Hg:CO ratio in 10 biomass burning plumes. Wiedinmeyer
and Friedli (2007) used a similar approach to estimate U.S.
emissions of mercury from wildfi res, and arrived at 44 tons/
yr with an uncertainty of 50%. In summary, natural emis-
sions of Hg are an important component of the global cycle,
but there are large uncertainties with regard to our estimates
of these emissions.
Anthropogenic Sources of Mercury in the Air
Emissions inventories of anthropogenic sources of mercury
have been compiled for most developed countries. These
are based on direct stack tests to determine the amount
of mercury being emitted for a particular facility. The
emission per unit of production is termed an “emission
factor.” The emission factor from one facility is often used
to estimate the emissions from another facility that has not
had a direct stack test. This, of course, assumes that both
factories operate similarly, with identical fuel and emission
controls. While the emissions for large facilities can be
relatively well known, emissions from smaller facilities are
often excluded from emission inventories, and this can
result in a signifi cant error in the total emissions. The most
important example is Chinese emissions from coal com-
bustion and metals smelting. Because of less centralized
consumption of coal (Wong et al., 2006) and a lack of data
regarding activity levels and emissions factors from smaller
operations and remote regions (Wu et al., 2006), there is
considerable uncertainty (
44%) about the total estimate
(Streets et al., 2005).
Based on emissions inventories, anthropogenic sources
are believed to emit approximately 2200-3400 tons/yr of
mercury to the atmosphere. While there are many different
source types, the largest sources, in order of importance,
are coal combustion, gold production, nonferrous metal
smelting, cement production, caustic soda manufacturing,
and waste incineration (Pacyna et al., 2006; Selin et al.,
2008). Coal combustion is the largest source globally, and
is responsible for about 1400 tons/yr, which is nearly two
thirds of the global anthropogenic total (Pacyna et al., 2006).
In most fi rst-world countries, coal combustion is carried
