The 1989 Loma Prieta earthquake occurred along the stretch of the San Andreas fault zone within the southern Santa Cruz Mountains that last failed as a major earthquake in 1906. The southeastern end of the 1989 rupture marks the transition from stable, aseismic slip on the central creeping section of the San Andreas fault to unstable failure on the locked 1906 segment. We investigate this transition and the rupture characteristics of the 1989 earthquake using a three-dimensional P wave velocity model of the southern Santa Cruz Mountains section of the fault zone. The velocity model was determined by joint progressive inversion of 5422 P wave arrival times from 173 earthquakes, which included aftershocks of the 1989 earthquake and premainshock ''background'' events, recorded by the U.S. Geological Survey central California network. This velocity model was used to relocate 424 background earthquakes, the Loma Prieta mainshock, and 292 aftershocks. The velocity model is calibrated in terms of lithology using laboratory data for California Coast Ranges rock types and available seismic refraction profiles and is shown to be in good agreement with the surface geology. The model images a large anomalous high-velocity body at midcrustal depths within the rupture zone of the 1989 earthquake that the available evidence suggests might have gabbroic or other mafic composition. On the basis of the relationship of the lithological features interpreted from the velocity model to the seismicity and surface creep we propose a model in which the high-velocity body is primarily responsible for the transition from stable to unstable fault slip at Pajaro Gap.

The active plane of the San Andreas fault cuts throughout the body. The fault system attempts to circumvent this barrier by transferring slip to secondary faults, including splay faults that have propagated along the frictionally favorable contact between the high-velocity rock mass and Franciscan country rocks. However, the neat arrest of stable sliding causes stress to concentrate within the body, and the high-strength, unstable contact within it evolves from a barrier to the asperity that failed in the 1989 earthquake. The general features of the 1989 rupture predicted by this asperity model agree with several rupture histories computed for the earthquake.The model implies that as proposed by other workers, the Loma Prieta earthquake did not involve a repeat of the 1906 slip, which has an important bearing on earthquake recurrence estimates for the Santa Cruz Mountains segment of the fault. ¿ American Geophysical Union 1993