Geoscience Reference
In-Depth Information
(a)
Fig. 11.9 Azimuthal variation of phase velocities of 200-s
Rayleigh waves (expanded up to I = m = 3) (after Tanimoto
and Anderson, 1984, 1985).
motion directions, and little is expected since
200-s Rayleigh waves are sampling the mantle
beneath the lithosphere. Pn velocity correlates
well with spreading direction.
Lateral heterogeneity from
body waves
(b)
Fig. 11.8 The intermediate wavelength geoid is controlled
by processes in the upper mantle (slabs, slow asthenosphere).
(a) The l = 6 component of a global spherical harmonic
expansion of Love-wave phase velocities. These are sensitive
to shear velocity in the upper several hundred kilometers of
the mantle. Note that most shields are fast (gray areas), and
oceanic and tectonic regions are slow (white areas). Hot
regions of the upper mantle, in general, cause geoid highs
because of thermal expansion and uplift of the surface and
internal boundaries. Tectonic and young oceanic areas are
generally elevated over the surrounding terrain. (b) Phase
velocity computed from the l = 6 geoid, assuming a linear
relationship between geoid height and phase velocity. Note
the agreement between these two measures of upper-mantle
properties (Tanimoto and Anderson, 1985).
Because of the irregular distribution of seismic
stations and seismic events, one cannot deter-
mine tomographic anomalies on a globally uni-
form basis from body waves. This limitation is
fundamental and cannot be cured by improve-
ments in ray theory, reference models or inver-
sion techniques. Many body-wave tomographic
maps and cross-sections simply reflect the ray
coverage and regions where the reference model
has been perturbed. If there is inadequate cov-
erage, the reference model is unchanged but
this does not mean it is right, or that contrasts
with adjacent regions are correct. In contrast to
surface-wave studies, however, the anomalies can
be fairly well localized geographically although
the depth extent is ambiguous. Average prop-
erties along rays or ray bundles can be deter-
mined accurately. Cross-sections oriented along
the great-circle paths between many sources and
receivers are the most useful. Tomographic cuts
in random directions across body-wave models
that do not include many sources and receivers
can be misleading. In principle, the most reliable
tomographic studies use both body waves and
surface waves, and correct for frequency effects.
azimuthal variation is low under North Amer-
ica and the central Atlantic, between Borneo
and Japan, and in East Antarctica. Maximum
velocities are oriented approximately northeast-
southwest under Australia, the eastern Indian
Ocean, and northern South America and east-
west under the central Indian Ocean; they vary
under the Pacific Ocean from north-south in
the southern central region to more northwest--
southeast in the northwest portion. The fast
direction is generally perpendicular to plate
boundaries. There is little correlation with plate
Search WWH ::




Custom Search