HR: 0830h
AN: S41C-0105    [PDF]
TI: Interpreting Focal Mechanisms in a Heterogeneous Stress Field
AU: * Smith, D E
EM: desmith@gps.caltech.edu
AF: California Institute of Technology, 252-21 Seismolab Caltech, Pasadena, CA 91125 United States
AU: Heaton, T H
EM: heaton_t@caltech.edu
AF: California Institute of Technology, 252-21 Seismolab Caltech, Pasadena, CA 91125 United States
AB: There is evidence that slip in earthquakes is spatially heterogeneous (with possibly a fractal distribution). This implies stress that is highly heterogeneous. In addition, there is evidence that the small-scale spatial heterogeneity is significantly larger than any uniform component of stress. Strain resulting in local stress changes of 100 MPa or more have been reported (McGill and Rubin, 1999), whereas GPS observations in Southern California indicate regional stress changes of 1 MPa per century, and energy estimates of large earthquakes predict average shear stress changes of less than 10 MPa (Kanamori and Heaton, 2000). Additional support for stress heterogeneity, has been shown in work by Rivera and Kanamori (2002) and Aagaard et al. (2001), Widely used techniques for determining the orientations and relative magnitudes of principal stresses (Gephart, 1990; Gephart and Forsyth, 1984; Michael, 1991; Michael, 1984, 1987) invert focal mechanisms from earthquakes. The objective is to find a spatially uniform component of the deviatoric stress field in hopes that it is indicative of the background stress. These techniques work only if the uniform component of stress is sufficiently large with respect to random spatial heterogeneity. We hypothesize that stress is very heterogeneous and the heterogeneity is large in comparison to uniform background stress. We also hypothesize that inversion results, which are intended to reveal uniform background stress, are actually characterizing secularly increasing tectonic stress. In other words, earthquakes are a biased sample of points brought closer to failure by tectonic stress. If these hypotheses are correct, they could have major implications for how stress inversions are interpreted for crustal stress studies and hazard analysis. Therefore, our plan is: 1) generate suites of synthetic focal mechanism from a variety of stress fields. Specifically, a random number generator with a Gaussian distribution will create an initial 3-D stress field. Various smoothing filters will be applied to create the different spatial heterogeneities, including some with fractal distributions. A uniform background component of stress will be applied. Then last, spatially uniform tectonic stress that linearly increases with time will be added to bring points to failure. A Hencky-Mises yield condition will determine the time to failure. 2) invert the focal mechanism suites using the Gephart and Forsyth (1984) technique and the Michael (1987) technique, 3) then compare the inversion results to the original focal-mechanism generating stress fields. This will give us a direct numerical test of the inversion methods. Our three main objectives are: 1) determine how the degree of spatial heterogeneity affects the inversion, 2) compare the effects of different types of heterogeneity, 3) determine to what degree inversions reflect spatially uniform, external tectonic stress.
UR: http://www.gps.caltech.edu/~desmith/research.html
DE: 3250 Fractals and multifractals
DE: 7223 Seismic hazard assessment and prediction
DE: 7260 Theory and modeling
DE: 8150 Plate boundary--general (3040)
DE: 8164 Stresses--crust and lithosphere
SC: Seismology [S]
MN: 2003 Fall Meeting