Geoscience Reference
In-Depth Information
grain-growth kinetics. Much remains to be
investigated on the grain-size evolution and
rheological properties on the lower mantle
minerals and their assemblages.
temperature (the D'' layer is a thermal boundary
layer where temperature increases by
1000 K
(Lay
et al
., 2008)). Geodynamic evidence for a
weak D'' layer is presented by Cadek and Fleitout
(2006) and Nakada and Karato (2012).
∼
(b) Is the D'' layer weak?
Ammann
et al
. (2010)
calculated the point defect mobility in lower
mantle minerals including post-perovskite. They
found large anisotropy in point defect mobility in
post-perovskite, and assuming that the diffusion
coefficient along the fastest direction controls the
creep rate by dislocation creep they concluded
that post-perovskite is weaker than co-existing
minerals when deformed by diffusion-controlled
dislocation creep. However, their conclusion is
not valid because (1) no constraints are avail-
able for defect concentration and (2) the averaging
scheme of anisotropic diffusion used by Ammann
et al
. (2010) is incorrect (Karato, 2010a). Cur-
rently no experimental studies are available on
plastic deformation of post-perovskite that can
be applied to deformation in the D'' layer. There
are a number of papers reporting deformation of
post-perovskite as cited above, but none of these
results is applicable to the D'' layer because all
of these experiments were conducted at low tem-
peratures at unknown strain-rates (and stress in
most cases). Hunt
et al
. (2009) conducted defor-
mation experiments on post-perovskite analogue
(CaIrO
3
) and concluded that there is weakening
associated with transformation from perovskite
to post-perovskite phase and suggested that this
is evidence for a weak post-perovskite phase.
However, the validity of their conclusion is ques-
tionable because weakening during a phase trans-
formation is observed in many materials caused
by the internal stress (or strain) (e.g., Zamora &
Poirier, 1983) and such an observation does not
necessarily mean that a newly formed phase is
weaker than the pre-existing phase.
Even though the notion of weak post-perovskite
proposed by Ammann et al. (2010) is questionable,
the D'' layer is likely to be weak compared to the
regions above. This is simply because of high
4.7 Some Speculations on the Rheological
Properties of Other Planets
Physical conditions in the planets such as the
Moon, Venus, Mars and Mercury are largely the
same as those in Earth with somewhat lower
pressures than Earth. Also the major element
compositions are also similar (except for modest
difference in iron content). Therefore inferring
rheological properties in these planets does not
require any new experimental data.
Most critical is the knowledge of water distri-
bution in these planets (or satellites). Although
observations relevant to the rheological proper-
ties in these planets are scarce, some observations
such as tidal energy dissipation provide some
constraints. For example, tidal
Q
of the Moon
is
40-50 (Williams
et al
., 2001), and that of
Mars is
∼
80 (Lainey
et al
., 2007; Bills
et al
.,
2005). Because tidal dissipation occurs mostly
in the deep mantle (e.g., Peale & Cassen, 1978),
these observations imply that deep mantles of
these planets have relatively low viscosity (tidal
Q
of Earth due to solid Earth is
∼
280 (Ray
et al
.,
2001)).
Q
and viscosity scale as
Q/Q
o
=
∼
(
η/η
o
)
α
with
α
≈
0.3 (Karato, 2008) (see also Chapter 3,
this volume, above). If such a relation is used,
then the viscosity of the deep mantle of the Moon
is estimated to be
10
19
Pa s and that of Mars is
∼
10
20
Pa s. A plausible cause for these low vis-
cosities is a large amount of water (hydrogen)
in the deep mantles of these planets (or satel-
lites) (for water in the Moon see Hauri
et al
.,
2011). Electrical conductivity measurements by
remote sensing (electromagnetic induction) on
these planets will provide additional data to es-
timate the water contents (see Chapter 5, this
volume, below).
∼
