Geoscience Reference
In-Depth Information
coverage in the Pacific Ocean (Deuss, 2007; Cao
et al
., 2011). Some studies have combined SS
precursors and Pds receiver functions (Li
et al
.,
2003). The main problem with SS precursors is
their large and complicated Fresnel zone, making
it difficult to find any narrowmantle plumes. Cao
et al
. (2011) applied migration using the Radon
transform, which improves the resolution, and
studied a region around Hawaii in the Pacific.
Surprisingly, they found evidence for ponding of
hot plume material under the 660 km discontinu-
ity west of Hawaii, and not located directly under
the main island hotspot.
Older studies have sought for evidence of a
thin transition zone, which would be predicted
from olivine phase transitions for a hot geotherm
(Li
et al
., 2000, 2003). However, the influence of
garnet becomes important for the characteristics
of the 660 km discontinuity at higher temper-
atures (Figure 10.6). The garnet transition from
majorite to perovskite becomes dominant. The
majorite to perovskite transition has a positive
Clapeyron slope, leading to a deeper 660 km
discontinuity, instead of a shallower 660 km
discontinuity in hot regions. This implies that
moderately hot mantle plumes may indeed
have a thin transition zone, but extremely hot
mantle plumes would have an average thickness
transition zone because of a deeper 410 and a
deeper 660 km discontinuity. Using this new
definition, Deuss (2007) measured the transition
zone thickness in all proposed mantle plume
locations of the Courtillot
et al
. (2003) catalog.
It was found that many of these locations could
indeed have a hot mantle plumes underneath;
especially many of the larger plumes as suggested
by mantle tomography, were also seen with the
SS precursors. Tauzin
et al
. (2008) applied the
same idea using Pds receiver functions on a
global scale, and do not find evidence for a thin
transition zone under hotspot locations. They
also explain this because of the influence of the
majorite to perovskite transition at 660 km depth.
able to reach the core mantle boundary and hot
plumes rising form the core mantle boundary all
the way up to the surface (Davies & Richards,
1992). Alternatively, the mantle convects in two
layers, with subducting cold slabs not being able
to enter the lower mantle (McKenzie & Richter,
1981). The general idea is that in one layer con-
vection, the 660 km discontinuity will be due to
a phase transition (Hofmann, 1997). To test for
this idea, seismic studies have sought to correlate
the 660 km discontinuity topography with local
temperature variations. On the other hand, in
two layer convection, there is potentially a differ-
ence in composition between the upper and lower
mantle, and the 660 km discontinuity would be
a compositional boundary with slabs ponding
around 660 km depth.
Shearer and Masters (1992) found a correlation
between regions where the 660 km discontinuity
is depressed and subduction zone locations,
which they interpreted as evidence for slabs
being deflected horizontally above the 660 km
discontinuity instead of slabs subducting into the
lower mantle. However, the post-spinel transi-
tion also predicts a deeper 660 km discontinuity
in cold subduction zone regions, so it is actually
very difficult to decide if the boundary deflection
is due to a compositional boundary or a phase
transition. More recent seismic evidence now
suggests that there is at least some transport
of material across the 660 km discontinuity.
Mantle tomography shows slabs subducting
from the upper into the lower mantle, especially
in the region around Indonesia and for the
Farallon slab under North America (e.g. Ritsema
et al
., 2011). This is the same region where the
transition zone is thicker, which would agree
with an interpretation in terms of olivine phase
transitions. In other regions, slabs are seen to
stay above the 660 km discontinuity, for example
under the Japan arc (Ritsema
et al
., 2011).
Detailed discontinuity studies will aid in further
constraining how much transport happens across
through the 660 km discontinuity. For example,
Schmerr and Garnero (2007) found that both the
'410' and the '660' were deepened beneath South
America, consistent with the down-dip direction
10.6.3 Mantle convection
The major question is whether the mantle is
convecting as one layer, with cold slabs being
