Natural and Anthropogenic Mercury Sources
21
to be responsible for the large difference between the two global estimates; thus,
their contribution constitutes a large uncertainty factor and should be
reevaluated. Studies at Solfatara volcano, Italy, estimated the Hg emission at
0.9-4.5 g day-I (0.3-1.6 kg year-I); (Ferrara et al. 1994), which is comparable to
estimates for the Kilauea volcano, Hawaii (1.5 kg year-I), but much smaller than
estimates for more active volcanoes, e.g., the Colima volcano, Mexico
(440 kg year-I), or the Etna volcano, Italy (2.7 t year-I) (Varekamp and Buseck
1986). Data for Hg emissions from other geothermal sources are even rarer than
for volcanoes, but it has been measured that one volcanic geyser on Iceland alone
emits 8 kg year-I Hg to the atmosphere (Edner et al. 1991). Also, atmospheric Hg
concentrations over a geothermal area were elevated by about a factor of 10
compared to background regions (Gustin et al. 1996), but there is a need for
measurements of the contribution of geothermal sources to the global Hg
emissions. Although few data are published, a recent study reported mercury
emission rates in the order of 10-1000 ng m2
h- ' in geothermal areas of the
western US (Gustin and Lindberg 1997). Global Hg emission from geothermal
sources has been estimated at 60 t year-I (Varekamp and Buseck 1986).
4.3
Evasion of Mercury from the Earth's Subsurface Crust
It is reasonable to assume that anthropogenic influences on the Earth's
geochemistry does not go deeper than the immediate surface layer (probably
much less than 0.1 km). However, mercury is permanently released from deeper
regions of the Earth's crust and permeates as mercury vapor to the surface
through faults and fractures in bedrock. This phenomenon is being explored as a
potential indicator of earthquakes, since it has been observed that Hg
concentrations in soil air increase dramatically directly before earthquakes
occur. Crustal Hg emissions probably display extreme spatial and temporal
variation. This has been demonstrated by measurements of mercury concentrations in a subterranean vault in Precambrian Shield bedrock (Klusman and
Webster 1981) where pronounced diurnal and seasonal cycles were observed. Air
and soil temperature, barometric pressure, and relative humidity were identified
as the most important controlling factors. It has been argued that crustal
mercury emission on the global scale is mainly driven by crustal heat flow rather
than by bedrock mercury content (Varekamp and Buseck 1986). From this
approach, continental crustal mercury evasion from geologic sources alone has
been calculated by one author as 3000 to 6000 t year-I (Rasmussen 1994). This
does not include volcanic or geothermal zones, which have to be treated
separately due to their much higher heat flow. Overall, this would give a generally
larger natural geological mercury flux from the continents to the atmosphere
than generally assumed in other mass balances (e.g., see Sect. 4.1). A review of
earlier estimates of natural mercury sources made between 1970 to 1982
summarizes published values in the range 2500 to 30 000 t year-I (Lindqvist et al.
1984). A more recent work has ended up with an estimate of 3000 t year-I
(Nriagu and Pacyna 1988).
21
to be responsible for the large difference between the two global estimates; thus,
their contribution constitutes a large uncertainty factor and should be
reevaluated. Studies at Solfatara volcano, Italy, estimated the Hg emission at
0.9-4.5 g day-I (0.3-1.6 kg year-I); (Ferrara et al. 1994), which is comparable to
estimates for the Kilauea volcano, Hawaii (1.5 kg year-I), but much smaller than
estimates for more active volcanoes, e.g., the Colima volcano, Mexico
(440 kg year-I), or the Etna volcano, Italy (2.7 t year-I) (Varekamp and Buseck
1986). Data for Hg emissions from other geothermal sources are even rarer than
for volcanoes, but it has been measured that one volcanic geyser on Iceland alone
emits 8 kg year-I Hg to the atmosphere (Edner et al. 1991). Also, atmospheric Hg
concentrations over a geothermal area were elevated by about a factor of 10
compared to background regions (Gustin et al. 1996), but there is a need for
measurements of the contribution of geothermal sources to the global Hg
emissions. Although few data are published, a recent study reported mercury
emission rates in the order of 10-1000 ng m2
h- ' in geothermal areas of the
western US (Gustin and Lindberg 1997). Global Hg emission from geothermal
sources has been estimated at 60 t year-I (Varekamp and Buseck 1986).
4.3
Evasion of Mercury from the Earth's Subsurface Crust
It is reasonable to assume that anthropogenic influences on the Earth's
geochemistry does not go deeper than the immediate surface layer (probably
much less than 0.1 km). However, mercury is permanently released from deeper
regions of the Earth's crust and permeates as mercury vapor to the surface
through faults and fractures in bedrock. This phenomenon is being explored as a
potential indicator of earthquakes, since it has been observed that Hg
concentrations in soil air increase dramatically directly before earthquakes
occur. Crustal Hg emissions probably display extreme spatial and temporal
variation. This has been demonstrated by measurements of mercury concentrations in a subterranean vault in Precambrian Shield bedrock (Klusman and
Webster 1981) where pronounced diurnal and seasonal cycles were observed. Air
and soil temperature, barometric pressure, and relative humidity were identified
as the most important controlling factors. It has been argued that crustal
mercury emission on the global scale is mainly driven by crustal heat flow rather
than by bedrock mercury content (Varekamp and Buseck 1986). From this
approach, continental crustal mercury evasion from geologic sources alone has
been calculated by one author as 3000 to 6000 t year-I (Rasmussen 1994). This
does not include volcanic or geothermal zones, which have to be treated
separately due to their much higher heat flow. Overall, this would give a generally
larger natural geological mercury flux from the continents to the atmosphere
than generally assumed in other mass balances (e.g., see Sect. 4.1). A review of
earlier estimates of natural mercury sources made between 1970 to 1982
summarizes published values in the range 2500 to 30 000 t year-I (Lindqvist et al.
1984). A more recent work has ended up with an estimate of 3000 t year-I
(Nriagu and Pacyna 1988).