Dynamic Rupture Modeling of the Transition from Thrust to Strike-Slip Motion in the 2002 Denali Fault Earthquake, Alaska
We use three-dimensional dynamic (spontaneous) rupture models to investigate the nearly simultaneous ruptures of the Susitna Glacier thrust fault and the Denali strike-slip fault. With the 1957 M_w 8.3 Gobi-Altay, Mongolia, earthquake as the only other well-documented case of significant, nearly simultaneous rupture of both thrust and strike-slip faults, this feature of the 2002 Denali fault earthquake provides a unique opportunity to investigate the mechanisms responsible for development of these large, complex events. We find that the geometry of the faults and the orientation of the regional stress field caused slip on the Susitna Glacier fault to load the Denali fault. Several different stress orientations with oblique right-lateral motion on the Susitna Glacier fault replicate the triggering of rupture on the Denali fault about 10 sec after the rupture nucleates on the Susitna Glacier fault. However, generating slip directions compatible with measured surface offsets and kinematic source inversions requires perturbing the stress orientation from that determined with focal mechanisms of regional events. Adjusting the vertical component of the principal stress tensor for the regional stress field so that it is more consistent with a mixture of strike-slip and reverse faulting significantly improves the fit of the slip-rake angles to the data. Rotating the maximum horizontal compressive stress direction westward appears to improve the fit even further.
Additional Information© 2004 Seismological Society of America. Manuscript received 3 February 2004. We thank Jeanne Hardebeck for guidance in interpreting stress orientations from focal mechanism inversions and analyzing our preferred stress orientations with the focal mechanism data. We also benefited from helpful discussions with Chen Ji and Douglas Dreger regarding their kinematic source inversions for this earthquake and thank Peter Hauessler for providing the locations of the surface rupture. We appreciated helpful reviews from Ruth Harris, David Oglesby, Paul Spudich, and an anonymous reviewer. The simulations were run on the Hewlett-Packard V-Class computer, located at the California Institute of Technology, with access provided by the Center for Advanced Computing Research. Some of the plots were created using the Generic Mapping Tools (GMT) (Wessel and Smith, 1998).
Published - S190.full.pdf