Earth as a Planet: Surface and Interior 211
FIGURE 24 Shear velocity anomalies just above
and below the boundary between the upper mantle
and lower mantle in model S262D1 of Gu et al.,
The differences indicate a drastic change in the
pattern of the anomalies, most likely associated
with a serious impedance to flow.Mediterranean. Even though elements of these two struc-
tures are present in our model, they are not equally well
defined. Also, there are many other features of comparable
amplitude. This is also true with respect to models pub-
lished by scientists at Berkeley and at Scripps, who used
parameterization similar to that in Figure 22. Intensive ef-
forts are made to understand the differences between the
results of two different approaches to tomography.
The map at 2800 km depth shows the velocity anoma-
lies as the CMB is approached. The ring of high velocities
circumscribing the Pacific basin is already visible at 2000
km; it strengthens considerably over the next 500 km and
increases even further toward the CMB. In the wavenum-
ber domain of spherical harmonics, the spectrum of lateral
heterogeneities is very red, being dominated by degree,
2 and 3. This is the dominant signal in the lower mantle,
very clear in properly displayed data. The location of the
ring of fast velocities corresponds to the location of sub-
duction zones during the past 200 Ma. The large red (slow)
regions are sometimes called the African and the Pacific
“superplumes.” Their origin is unknown; they, most likely,
represent both thermal and chemical heterogeneity. There
is a good correlation between the location of the two super-
plumes and distribution of hotspots at the Earth’s surface,
indicating a degree of connection between the tectonics at
the surface and conditions near the core–mantle boundary.
Figure 25 gives two views of low-pass filtered anoma-
lies in the lower mantle in a model of by Ritsema et al. in
1999, plotted in Cartesian coordinates: the red is a 0.6%
isosurface and blue is+0.6%. We see the circum-Pacific
ring of fast anomalies and the two low velocity anomalies:
one very concentrated under the Pacific and a more diffuse
one under the Atlantic and Africa. Their radial continuity
throughout the lower mantle indicates that they cannot be
explained by processes at the core–mantle boundary alone.
The origin of this large-amplitude, very large wavelength
signal has not yet been explained by geodynamic model-
ing, although an assumption that the velocity and gravity
anomalies are correlated leads to a good prediction of the
geoid at the gravest harmonics.
It was believed since 1977, the time of publication of the
first large-scale GST study, that three-dimensional images
of lateral heterogeneity in the mantle will be an essential tool
in addressing some of the fundamental problems in earth
sciences. The results accumulated since then confirm that
statement even though much progress is still to be made.
Cooperation among the different fields of Earth sciences
(geodynamics, mineral physics, geochemistry, seismology,
geomagnetism) it the requisite condition to fulfill this goal.FIGURE 25 Low-pass filtered S-velocity model of Ritsema et al.
in a three-dimensional projection; the top 800 km of the
structure is removed.