Developing Finite-Frequency Regional Pn Velocity Models
SCIENCE APPLICATIONS INTERNATIONAL CORP SAN DIEGO CA
Pagination or Media Count:
Head waves and Pn waves in particular are important in studying the predominantly layered velocity structure of the Earth and discriminating the seismic sources. Conventional studies of head waves have used high-frequency ray approximation in wave propagation. While recent studies have shown the diffractive nature and the three dimensional 3-D sensitivities of finite-frequency turning waves, analogs of head waves in a continuous velocity structure, the finite-frequency effects and sensitivity kernels of head waves are yet to he carefully examine We present the results of a numerical study on the finite-frequency behavior of head waves. A reference model with a low-velocity layer over a high-velocity half-space is used. Velocity anomalies of various sixes are placed on either side of the interface at different locations. A 3-D fourth- order staggered-grid finite-difference method is used to calculate synthetic waveforms, and travel time anomalies are measured by cross correlations of seismograms for models with and without velocity anomalies. The results show that finite-frequency head waves are sensitive to the 3-D velocity perturbation in a more complex way than predicted by ray theory. The peak travel time sensitivity is located near the two piercing points at which the head wave reaches and leaves the interface. The Sensitivity is much smaller elsewhere on the ray path. Fresnel zones can be observed from the pattern of positive and negative sensitivities, with the strongest negative sensitivity in the first Fresnel zone. Unlike turning waves, the head wave has a nonzero sensitivity right beneath the interface along the source-receiver path. But at some distance below the interface, the sensitivity has a local minimum. We are in the process of constructing 3-D full-wave sensitivity kernels from 3-D reference velocity models and compare the measured travel time anomalies with the predictions from the 3-D sensitivity kernels.
- Theoretical Mathematics