Geoscience Reference
In-Depth Information
perovskite, and the influence of hydrogen on
the diffusion in (Mg,Fe)O must be determined at
higher pressures.
conductivity profile can be interpreted by ''wet''
(
0.01wt %water) model with geophysically rea-
sonable temperature profiles.
Similarly some preliminary studies have been
published on the distribution of electrical conduc-
tivity in Mars (e.g., Verhoeven
et al
., 2005). The
Martian mantle corresponds to the upper mantle
and the transition zone of Earth, and therefore a
nearly complete data set on electrical conductiv-
ity is available. For Mars, the influence of high
iron content needs to be included. Results on or-
thopyroxene summarized above suggest that high
iron content enhances hydrogen-related conduc-
tivity as well as iron-related conductivity, but
the exact amount of enhanced conductivity is
unknown. More important is the influence of
hydrogen. However, the influence of water on
electrical conductivity was not included in the
previous studies. Water content in Martian man-
tle is controversial (e.g., Greenwood
et al
., 2008;
Elkins-Tanton, 2008; Guest & Smrekar, 2007;
Dreibus & Wanke, 1985; Breuer & Spohn, 2006).
Once the data set on the electrical conductivity
in Mars becomes available, a comparison to the
experimental data will provide strong constraints
on the hydrogen distribution in Mars, which will
help us understand the evolution of this planet.
∼
(d) Applications to the Moon and Mars
Because
of the absence of the surface oceans, the elec-
trical conductivity distribution in the Moon can
be well-constrained from the analysis of electro-
magnetic induction (Sonett
et al
., 1971; Sonett,
1982; Hood
et al
., 1982a,b). In these early studies,
the distribution of electrical conductivity was
interpreted based on the experimental data on
electrical conductivity of hydrogen-free (dry) sam-
ples (Figure 5.20). However, the strong influence
of hydrogen on electrical conductivity has re-
cently been demonstrated as has been reviewed
in this chapter, and the evidence for a substantial
amount of water in the lunar mantle is reported
(e.g., Saal
et al
., 2008; Greenwood
et al
., 2011;
Hauri
et al
., 2011). Consequently, it is impor-
tant to reinterpret the lunar conductivity profile
in terms of temperature and water distribution.
Because the pressure conditions in the Moon cor-
respond to the shallow upper mantle in Earth
(
<
5GPa), the results shown in Figures 5.15 and
5.17 can be directly applied to the Moon. One dif-
ference is that the oxygen fugacity in the Moon is
likely much lower than that in Earth (e.g., Righter
& Drake, 1996), which reduces conductivity by
iron but enhances proton conductivity (lunar iron
content is slightly larger than Earth but other-
wise the major element chemistry is similar to
Earth (except for a small difference in Fe content)
see e.g., Ringwood, 1979; Khan
et al
., 2007). The
temperatures in the lunar interior inferred from
the seismological and gravity observations (Khan
et al
., 2007; see also Lambeck & Pullan, 1980)
are considerably lower than those inferred from
electrical conductivity by Hood
et al
., 1982a,b
8
using the dry olivine and orthopyroxene data
(Figure 5.20). It can be shown that the lunar
5.6 Summary and Perspectives
In the geophysical literature, high electrical con-
ductivity (and low seismic wave velocities) was
often attributed to the presence of some fluids
including melts and aqueous fluids (Shankland
et al
., 1981; Jones, 1992). During the last sev-
eral years, there have been extensive studies on
the electrical conductivity of minerals with the
focus on the influence of hydrogen, and these
results show that a majority of observed high
conductivity can be explained by the solid-state
mechanisms involving hydrogen. After review-
ing key experimental details, we conclude that
if results from properly executed experimental
studies are used, the proposal by Karato (1990)
of hydrogen-enhanced conductivity is now well
supported for the lower crust, upper mantle and
transition zone minerals. In most of the Earth's
8
Hood et al. (1982b) assumed a high Al
2
O
3
content in
orthopyroxene that resulted in relatively low temper-
atures. This assumption is incorrect because most of
Al
2
O
3
is in either plagioclase or garnet. If the correct
Al
2
O
3
content were used one would obtainmuch higher
temperatures.
