
Wavefield extrapolation of body waves for 3‐D imaging of earthquake sources
Author(s) -
Chang WenFong,
McMechan George A.
Publication year - 1991
Publication title -
geophysical journal international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.302
H-Index - 168
eISSN - 1365-246X
pISSN - 0956-540X
DOI - 10.1111/j.1365-246x.1991.tb04603.x
Subject(s) - extrapolation , geology , a priori and a posteriori , position (finance) , geodesy , energy (signal processing) , seismology , physics , mathematics , mathematical analysis , philosophy , epistemology , finance , economics , quantum mechanics
Summary Reverse‐time imaging of earthquake source parameters is extended from two to three spatial dimensions and from two‐ to three‐component recordings. Provided that the recording aperture is sufficiently large, and the data are not spatially aliased, source radiation patterns, in both time and space, can be reconstructed by elastic reverse‐time propagation of body waves; such reconstructions are necessarily partial as only the energy that was recorded is available for reconstruction. For a point source, the origin time and the (3‐D) spatial location of an event can be reconstructed by extracting the time and position, of the best focused energy from the backward propagating wavefield. For a spatially and temporally extended source, biased estimates of fault position and rupture/slip time history can be estimated if the recording aperture is sufficient; for smaller recording apertures, the time history can still be estimated if the fault location and geometry are known a priori. The latter is viable for a reasonable number (<200) of three‐component recordings, and so is potentially applicable to real data. All these processes assume that a smoothed representation of the 3‐D velocity distribution in the volume containing the source and receivers is available.