HR: 1330h
AN: S12A-0369    [PDF]
TI: Modeling teleseismic P and SH static offsets for Great strike-slip earthquakes
AU: * Ji, C
EM: jichen@gps.caltech.edu
AF: Seismological Laboratory, Dvision of Geological and Planetary Sciences, California Institute of Technology, 252-21,caltech, Pasadena, CA 91125 United States
AU: Tan, Y
EM: ytan@gps.caltech.edu
AF: Seismological Laboratory, Dvision of Geological and Planetary Sciences, California Institute of Technology, 252-21,caltech, Pasadena, CA 91125 United States
AU: Helmberger, D
EM: helm@gps.caltech.edu
AF: Seismological Laboratory, Dvision of Geological and Planetary Sciences, California Institute of Technology, 252-21,caltech, Pasadena, CA 91125 United States
AU: Tromp, J
EM: jtromp@gps.caltech.edu
AF: Seismological Laboratory, Dvision of Geological and Planetary Sciences, California Institute of Technology, 252-21,caltech, Pasadena, CA 91125 United States
AB: Most body-wave modeling of earthquakes involves P and SH waves observed between $30\deg$ to $90\deg$. Traditionally, only the "far field" contributions, whose waveforms are related to the moment-rate function ($\dot{M}(t)$), are considered. The "near field" contribution, whose waveform is determined by the moment release function ($M(t)$), has generally been neglected. According to the "far field" approach, the displacements should ultimately become zero. However, in broadband records of recent large strike-slip events, such as the 2001 Kunlun and 2002 Denali earthquakes, displacement offsets of up to two thirds of the peaks of the P and SH phases are maintained until they are disturbed by the later arriving PP or SS phases. Hence, "near field" contributions must be included when we study the source processes of such events. We investigate complete synthetic seismograms generated by three basic double-couple fault types: $90\deg$ strike-slip, $90\deg$ dip-slip, and $45\deg$ dip-slip using several numerical solutions. We find that for a shallow earthquake, a $90\deg$ strike-slip fault can excite large P and SH near field terms, while the other two fault types generate negligible amplitudes. Our theoretical analysis indicates that this observation is entirely controlled by the source excitation characteristics. More importantly, the amplitude ratio of the "far field" to the "near field" is roughly proportional to the reciprocal of the shear velocity at the hypocenter. Hence considering "near-field" contributions provides unique constraints not only on the entire strike-slip component of fault-slip, but also on the down-dip extension of the rupture. While the "near field" terms are important, it is computationally difficult to generate high-frequency Green's functions at teleseismic distances using normal-mode or spectral-element (SEM) methods. The reflectivity method also fails to produce the near field term of the P waves due to the inaccurate earth-flatten approximation. Therefore, we develop a ray-based algorithm using "near-field" source excitation coefficients and correcting for the geometrical spreading in the spherical earth with normal-mode synthetics. Broadband synthetics generated based upon this new method compare favorably with SEM simulation. The new approach is used to study the source process of the 2002 Denali earthquake along with local strong motion and GPS observations.
DE: 7200 SEISMOLOGY
DE: 7203 Body wave propagation
DE: 7212 Earthquake ground motions and engineering
DE: 7215 Earthquake parameters
DE: 7260 Theory and modeling
SC: Seismology [S]
MN: 2003 Fall Meeting