HR: 0830h
AN: S11E-0335    [PDF]
TI: Wave-equation Imaging of Teleseismic Body-wave Coda
AU: * Shragge, J C
EM: jeff@sep.stanford.edu
AF: Stanford University, Geophysics, Mitchell Bldg., Stanford, CA 94305 United States
AU: Artman, B W
EM: brad@sep.stanford.edu
AF: Stanford University, Geophysics, Mitchell Bldg., Stanford, CA 94305 United States
AB: Historically, characterization of the lithosphere with information in teleseismic body-wave coda has been realized with collections of 1-D receiver functions. However, the promise of larger, fully 3-D teleseismic data sets with finer spatial sampling (e.g. the US-ARRAY project) motivates the investigation of industry-oriented imaging algorithms in the context of crustal/upper mantle studies. In this milieu, we present a multi-dimensional structural imaging method based on wave-equation migration of multi-component teleseismic array data for v(x,y,z) media. Although the advantages of wave-equation methods are well known to explorationists, these methods are rarely used in teleseismic investigation. Possible reasons for their disuse are unconventional source characteristics, and the heretofore limited number, and irregular distribution, of receivers. However, given sufficient receiver density teleseismic body-wave coda may be readily imaged with a wave-equation processing strategy through the use of a modified shot-profile migration algorithm. Shot-profile migration requires separate depth extrapolation of source and receiver wavefields. The source wavefields are modeled using the slowness vector of the incident body-wave that dictates the time-slope of the impulsive line (in 2-D) or plane sources (in 3-D). The receiver wavefield is the scattered energy in the body-wave coda after a deconvolution with the estimated source-signature. The two wavefields are then independently extrapolated through separate velocity models according to the wave-equation, and an imaging condition is applied at each model location to generate the image. The presence of forward- and backscattered arrivals of P,SV and SH polarity within the coda requires various combinations of migration parameters to independently focus different scattering modes. Accordingly, seven different images can be produced through appropriate permutations of velocity models and source wavefield propagation direction. Importantly, the application of this imaging technique to 3-D data sets needs only a 3-D velocity model. We present multi-event, stacked, migrated images of all possible scattering modes for both synthetic and field data sets. The imaging strategy is tested on a 2-D, finite-difference modeled data representing a subduction/suture earth model. The method is then applied to the IRIS-PASSCAL CASC93 data acquired in central Oregon, USA, and generates interpretable images of the Cascadia subduction zone.
UR: http://sepwww.stanford.edu/sep/people/jeff/agu2003SHRAGGE.txt
DE: 7218 Lithosphere and upper mantle
DE: 7260 Theory and modeling
DE: 7294 Instruments and techniques
SC: Seismology [S]
MN: 2003 Fall Meeting