Geology Reference
In-Depth Information
describing the relative importance of convection
view is that the 670-km phase transition favors
the accumulation of cold material at the base of
the transition zone. When the negative thermal
buoyancy of the accumulated material overcomes
the positive component of buoyancy, then the slab
rapidly penetrates into the lower mantle, deter-
mining the formation of a descending column
of cold material and simultaneously allowing
the upward injection of hot material across the
670-km discontinuity (e.g., Zhong and Gurnis
1994
). In this scenario, mantle convection is nor-
mally layered, but catastrophic events of mass
exchange (and associated whole-mantle convec-
tion) periodically occur to restore the equilibrium
between upper and lower mantle. As pointed out
by Christensen (
1995
), such a “hybrid” regime
requires values of
Ra
greater than 10
6
, whereas
for
Ra
< 10
6
layered convection would prevail.
An alternative viewpoint has been proposed by
Don Anderson and others since the early 1980s
(e.g., Anderson and Natland
2005
). In this model,
the 670-km discontinuity represents a
chemical
boundary and convection is essentially confined
to the upper mantle. However, the major point
of controversy is probably the interpretation of
mantle plumes. In the original formulation of
Morgan (
1971
), these are hot and narrow regions
of active upwelling that originate as instabilities
within hot thermal boundary regions at the base
of the mantle. This model was formulated to ex-
plain the age progression and linear arrangement
of volcanic island chains and aseismic ridges,
which would originate from the motion of tec-
tonic plates over
hot-spots
fixed at the base of
the lower mantle. In many modern models of
mantle circulation, plumes involve large mass
transport from the base of the mantle to the
Earth's surface. Together with deep slab penetra-
tion, they represent an essential feature of
whole
-
mantle convection
(van der Hilst et al.
1997
),
in which layered circulation is considered as a
local transient phenomenon. The alternative sce-
nario proposed by Don Anderson and colleagues
presents a lower mantle characterized by sluggish
convection (very low
Ra
), where any deep mantle
plume would be suppressed by the effects of
pressure on viscosity, thermal conductivity and
thermal expansion (Anderson
2002
). Conversely,
the asthenosphere is considered as a layer charac-
terized by local chemical heterogeneity, as well as
variations of fertility and melting point, that are
consequent to subduction of young plates, aseis-
mic ridges, and seamounts, and the delamination
of lower continental crust (Anderson and Natland
2005
). In this view, all upwellings are passive
and anomalous plume magmatism only reflects
higher fertility, not higher temperature (Anderson
2006
). Therefore, plumes are ultimately viewed
as thermo-chemical heterogeneities, which can
be transported passively by upper mantle currents
(e.g., Hawaii) or sampled by migrating ridges
(e.g., Iceland, Galapagos).
Recent advances in mantle tomography and
the study of seismic scattering in the lower mantle
have confirmed the presence of heterogeneities
within the entire mantle. For example, Shearer
and Earle (
2004
) have found that the upper
mantle has strong wave scattering, which
determines 3-4 % rms velocity heterogeneity
at 4-km scale length, whereas the lower mantle
heterogeneity would be only 0.5 % rms at 8-km
scale length. Most importantly, three independent
studies published in 2004 confirmed the existence
of two large nearly antipodal thermochemical
“superplumes” in the lower mantle, which are
now known, respectively, as the
Pa cifi c
and the
African superplumes
(Trampert et al.
2004
;Ishii
and Tromp
2004
; Ritsema et al.
2004
). These
deep structures represent large low-shear-velocity
provinces (LLSVP), which are characterized
by anomalously high density
and
temperature
(Fig.
1.15
), which implies that they constitute
chemically
distinct regions of the lower mantle.
The chemical diversity of these regions can
be described in terms of anomalously high per-
ovskite and iron content (Fig.
1.16
). The contrast-
ing effect of high density and high temperature
on the buoyancy of LLSVPs determines compli-
cate geodynamic behaviour of these regions. As
the thermal expansivity decreases with increas-
ing depths, in the lowermost 1,000 km of the
lower mantle the positive thermal buoyancy of the
LLSVPs is surmounted by the negative chemical
buoyancy, thereby the lower part of these regions
