Geoscience Reference
In-Depth Information
ocean, a fast region centered southeast of South
America. At 100 km depth the overall variations
are smaller than at shallower depth. Triple junc-
tions are slow. Below 200 km depth, the correla-
tion with surface tectonics starts to break down.
Shields are fast, in general, but ridges do not
show up systematically. The East African region,
centered on the Afar, is slow. The south-central
Pacific is faster than most shields. An interesting
feature is the belt of slow mantle at the Pacific
subduction zones, a manifestation of the volcan-
ism and marginal sea formation induced by the
sinking ocean slab. Below 340 km, the same belt
shows up as fast mantle; the effect of cold sub-
ducted material that was formerly part of the
surface thermal boundary layers. Many ridge seg-
ments are now fast. At larger depths the reso-
lution becomes poor, but these trends seem to
persist.
At intermediate depths, regions of uprising
(ridges) or downwelling (subduction zones) have
an SV
Z
6
P
4
2
SV
SH
s
,
y
2
4
6
V (km
/
s)
Fig. 20.5
Velocities as a function of angle and fluid
properties in granite containing aligned ellipsoidal cracks
(orientation shown at origin) with porosity
=
0.01 and aspect
ratio
=
0.05. The short dashed curves are for the isotropic
uncracked solid, the long dashes for liquid-filled cracks (
K
L
=
100 kbar) and the solid curves for gas-filled cracks (
K
L
=
0.1 kbar) (after Anderson
et al
., 1974).
SH anisotropy. Shallow depths (50 km)
show very large anisotropy variations (
>
10%).
From observed Pn anisotropy and measured
anisotropy of olivine, such values are not unrea-
sonable. At 100 km the amplitude of the vari-
ations is much smaller (
±
to 5% has been inferred from surface-wave
studies in order to fit Love waves and Rayleigh
waves. Azimuthal anisotropy can be averaged
out. We are then left with only polarization
anisotropy and can use a transversely isotropic
parameterization.
Shear-velocity maps and cross-sections are
shown in Figures 20.6 through 20.8. Shear veloc-
ity shows a strong correlation with surface tec-
tonics down to about 200 km. Deeper in the
mantle the correlation vanishes, and some long-
wavelength anomalies appear. At 50--100 km
depth heterogeneities are closely related to sur-
face tectonics. All major shields show up as
fast regions (Canada, Africa, Antarctica, West
Australia, South America). All major ridges show
up as slow regions (East Pacific, triple junctions
in the Indian Ocean and in the Atlantic, East
African rift). The effect of the fast mantle beneath
the shields is partially offset by the thick crust.
Old oceans also appear to be fast, but not as
fastasshields.Afewregionsseemtobeanoma-
lous, considering their tectonic setting: a slow
region around French Polynesia in average age
5%), but the pattern
is similar. The Mid-Atlantic Ridge has SV
>
SH,
whereas the other ridges show no clear-cut trend.
Under the Pacific there appear to be some par-
allel bands trending northwest-southeast with a
dominant SH
±
SV anomaly. This is the expected
anisotropy for horizontal flow of olivine-rich
aggregates. At 340 km, most ridges have SH
>
<
SV
(vertical
flow).
Antarctica
and
South
America
<
have a strong SV
SH anomaly (horizontal flow).
North America and Siberia are almost isotropic
at this depth. They exhibit, however, azimuthal
anisotropy. The central Pacific and the eastern
Indian Ocean have the characteristics of verti-
cal flow. These regions have faster than average
velocities at shallow depths and may represent
sinkers.
Age-dependent transverse isotropy
A wide band of anomalous bathymetry and mag-
matism
extends
across
the
Pacific
plate
from
