HR: 0830h
AN: S21E-0349 [PDF]
TI: Global finite-frequency surface-wave tomography
AU: * Zhou, Y
EM: yingz@princeton.edu
AF: Department of Geosciences,
Princeton University, Guyot Hall, Princeton, NJ 08544 United States
AU: Nolet, G
EM: nolet@princeton.edu
AF: Department of Geosciences,
Princeton University, Guyot Hall, Princeton, NJ 08544 United States
AU: Dahlen, F
EM: fad@princeton.edu
AF: Department of Geosciences,
Princeton University, Guyot Hall, Princeton, NJ 08544 United States
AU: Laske, G
EM: gabi@igpp.ucsd.edu
AF: IGPP, University of California,
San Diego, San Diego, CA 92093 United States
AB:
We investigate finite-frequency effects in global surface-wave
tomography. We compare two-dimensional phase-velocity maps
inverted using ray theory and sensitivity kernels based
on Born approximation. The same comparison is also made
for three-dimensional upper mantle shear-wave velocity structures.
The 3-D sensitivity kernels are computed using a fast kernel computation
approach (Zhou et al, 2003). We incorporate the forward-scattering approximation,
and take sidelobes fully into account when
calculating the 2-D sensitivity kernels for phase-velocity
perturbations. \\
The global phase-delay measurements
used in this experiment has a frequency
range from 3 mhz to 20 mhz (Laske \& Masters, 1996).
The same local parameterization (average grid spacing $\sim 4^{\circ}$) is applied in
both the ray theoretical inversion and the finite-frequency kernel inversion.
In general, inversions based on ray theory and finite-frequency sensitivity kernels
show similar large-scale geographic pattern, both in
the 2-D phase-velocity maps and the 3-D shear-wave structures.
However, the difference is significant in the amplitude of the anomalies
and in the small-scale structures.
In 2-D phase-velocity maps, the difference becomes more apparent at long periods,
due to the increasing width of the Fresnel zone.
Theoretically, finite-frequency kernels are expected to converge to ray theory
whenever the scale of the anomaly is much larger than the wavelength. However, due to
limited path coverage in
tomographic practice, inversions based on ray theory often introduce considerable
artifacts in both the amplitudes and locations of the anomalies even when the real
structure is large in scale.
DE: 7200 SEISMOLOGY
DE: 7218 Lithosphere and upper mantle
DE: 7255 Surface waves and free oscillations
DE: 7260 Theory and modeling
SC: Seismology [S]
MN: 2003 Fall Meeting