Environmental Engineering Reference
In-Depth Information
accounts for about 70% of the planet's mass (Jackson
1998; Helffrich and Wood 2001). The transfer of heat
through the mantle takes place mostly by convection,
heat-induced motion of solid yet (on geological time
scales) flowing rock.
But the dynamic link between the convecting mantle
and the overlying lithosphere remains conjectural. Plate
tectonics have commonly been seen as nothing but the
surface manifestation of mantle convection, but a more
accurate view may be to see the convection patterns
as the result of plate tectonics (D. L. Anderson 2002).
Oceanic plates have negative buoyancy because they are
cooler and hence heavier that the underlying mantle and
descend into it more rapidly than the heat conduction
warming the slab (Peacock 2003). As a result, the sub-
ducting lithosphere remains cooler and heavier than its
surroundings even after more than 100 Ma. The gravita-
tional pull of subducted slabs is seen as the prime mover
of plate tectonics, but the coupling between slabs and
plates is unclear, and the motion is best explained by
invoking both slab pull and slab suction forces (Conrad
and Lithgow-Bertelloni 2002). We are also not sure
about the way the mantle convection takes place. Is it a
whole-mantle process or a layered convection separated
into two giant cells by a discontinuity at a depth of 660
km (O'nions, Hamilton, and Evensen 1980; M. W. Kel-
logg Co. 1998; Hofmann 2003)?
Geochemists, basing their arguments on compositional
differences within the mantle, have favored the layered
model; seismologists, pointing to cold subducting slabs
that penetrate deep below the 600-km boundary, have
argued against any layered mixing. A new hypothesis has
tried to bridge this gap by proposing a slow-rising deep
mantle and a fast-rising upper mantle, which are sepa-
rated by a thin discontinuity at 410 km, where the mate-
rial undergoes dehydration-induced partial melting that
explains the difference in composition (Bercovici and
Karato 2003). And, going even deeper, we are uncertain
about the degree to which the lower mantle interacts
with the Earth's core (C. A. Lee 2004) and now believe
that the core-mantle boundary may be structurally as
complex as the Earth's crust (Garnero 2000).
Another controversial topic has been the genesis, per-
sistence, and stability of hotspots. J. T. Wilson (1963)
chose the term for those volcanic phenomena (ridges
or subduction zones) that are not associated with plate
boundaries and hence appear to represent a convection
mode independent of plate tectonics. By far the best
known example of a massive hotspot is the one that cre-
ated the Hawaii islands and the chain of seamounts in the
North Pacific. Other well-known island hotspots include
Iceland, Azores, Galapagos, Samoa, and Tahiti; the best
examples of continental hotspots are the Yellowstone
and Africa's Afar and Hoggar. Morgan (1971) attributed
hotspots to fixed, deeply anchored (near the mantle-core
boundary), and fairly narrow (50-100 km across) plumes
of hot rock whose relatively high speed (up to 50 cm/a)
and buoyancy carry them through the upper mantle and
the overlying (and much slower moving) plates. The
worldwide total of putative hotspots eventually grew to
more than 5,000, but only about 50 volcanic locales de-
serve to be in this category (fig. 2.10) (Courtillot et al.
2003).
The very existence of plumes feeding volcanic hotspots
is disputed. Nataf (2000) argued that direct evidence for
actual plumes is weak, and any plumes in the lower man-
tle would be difficult to detect by seismic means. Heat
flow measurements do not show highly elevated temper-
atures for eruptive plumes, and the plumes are not fixed
relative to one another: relative motion between the
