Geoscience Reference
In-Depth Information
in addition to the compositional heterogeneity,
much larger S velocity perturbation at least more
than
3
V
Φ
2
−
5%
to
−
6%
(
∼−
2%
by
the
composi-
V
P
tional effect,
−
3% to
−
4% by the thermal effect
1
(Wentzcovitch
et al
.,
2004,
2006)
should
be
0
observed. Moreover, since
1000 K tempera-
ture increase decreases the density by
+
1%,
which removes the gravitational stability of
MORB in the entire lower mantle. Therefore,
the hot and dense basaltic pile model (Nak-
agawa & Tackley, 2010) seems impossible to
explain LLSVP.
Sharp side boundaries observed between am-
bient mantle and LLSVP are often referred to
associated with possible chemical heterogeneities
at the base of mantle (e.g., Ritsema
et al
., 1998;
Ni
et al
., 2002, 2005; Helmberger
et al
., 2005;
Takeuchi
et al
., 2008), because a lateral temper-
ature variation with
∼−
−
1
−
2
V
S
−
3
0
50
100
150
P (GPa)
Fig. 7.12
Velocity contrasts between the pyrolite and
MORB aggregates (
V
MORB
−
V
Pyrolite
) as a function of
pressure. Uncertainties produced by applying
3%
fluctuations in the phase fractions are demonstrated by
shaded areas.
±
∼
1000 K, which is required
to produce the
3% low velocity anomaly,
is unlikely in a short horizontal length scale of a
few 100 km. However, the seismological and min-
eral physical joint modeling (Kawai & Tsuchiya,
2009) demonstrated that due to the large Clapey-
ron slope of the PPv transition and the existence of
thermal boundary layer in D
, even a 200
−
2
∼−
perturbation found in
V
P
(
ln
V
P
)of1%to2%
(Zhao, 2004) is hard to be reconciled by the
compositional contrast between these petrology
models alone. However, if considering
1000
K temperature anomaly, which decreases the
Pvelocityby
+
1% (Wentzcovitch
et al
., 2004;
2006), the thermochemical pile could explain the
observed
V
P
perturbation.
In contrast, the
ln
V
S
of
∼
300 K
thermal anomaly is enough to reproduce this ve-
locity anomaly between colder and hotter regions
with and without the PPv transition, respectively.
A few 100 K thermal fluctuation rather exists
highly likely even in a short horizontal scale. The
sharp side boundaries are therefore accounted for
mostly by combining lateral temperature varia-
tions
∼
1.5% calculated
in the deepest region (Figure 7.12) is fairly
comparable to the observed S velocity pertur-
bations
∼−
2% to 3% (Grand
et al
., 1997;
M egnin & Romanowicz, 2000; Antolik
et al
.,
2003; Takeuchi, 2007), again implying that the
compositional variation alone is not responsible
for the lateral S velocity heterogeneity. Man-
tle dynamics simulations (e.g., Nakagawa &
Tackley, 2005; Bull
et al
., 2009; Lassak
et al
.,
2010), however, report that the dense MORB com-
ponent accumulating at the CMB isolated from
the global mantle circulation causes a substantial
increase
of
±
the PPv phase change.
Since the approximated values were used for
the mineral phase fractions in the present calcu-
lations of the aggregate elasticity, we examined
uncertainties of the modeled properties by ap-
plying
+
±
3% fluctuations in the mineral phase
proportions and found that those yield insignif-
icant effects (Figure 7.12). The present results
are much based on more extensive analyses than
those of the previous study by Ohta
et al
. (2008),
although the conclusions obtained are basically
similar to those based on the static calculation
results (Tsuchiya, 2011). Summing up the above
discussion on
V
P
1000 K) in these
chemically distinct regions at the base of the
mantle (Nakagawa & Tackley, 2010), and then
dense and hot large thermochemical piles would
develop in the present mantle imaged as LLSVP.
The present calculations, however, indicate that
if such large temperature variation really exists
in
temperature
(
∼+
and
V
S
, we can conclude that
