Abstract

Strong motion records of the 1989 Loma Prieta earthquake are inverted to determine a model of the rupture history. Uncorrected horizontal and vertical accelerograms are integrated to particle velocity time histories for 38 stations within an epicentral range of 75 km. The time histories are bandpassed filtered with corners at 0.05 and 1.0 Hz. These bandpassed time histories are inverted using a nonlinear method to solve for the distribution of slip amplitudes and rupture times at specified locations on the fault plane. The fault plane is specified a priori: 38 km long and 17 km wide, extending from 3 to 19 km depth at a constant dip of 70°. Starting models have rupture times based on constant rupture velocities of 2.5, 2.8, and 3.0 km/sec and uniform slip with rise times of 0.5, 1.0, 1.5, and 3.0 sec.

The waveform inversion results show the strike-slip displacement is concentrated at the southern end of the rupture (rake = 156°) and the dip-slip displacement is concentrated at the northern end of the rupture (rake = 115°). The average total slip is partitioned almost equally between strike slip and dip slip (rake = 137°). The hypocentral area has an unusually small amount of slip with almost no slip in a region just to the north and up dip from the hypocenter. The rupture front is complex, propagating up dip to the south faster than it propagates to the north. The region of maximum strike slip to the southeast radiates simultaneously with the region of maximum dip slip to the northwest. The average rupture velocity is 3.0 km/sec, approximately 0.83 times the local shear wave speed. The calculated seismic moment is 3.5 ± 0.5 × 1026 dyne-cm.

First Page Preview

First page PDF preview
You do not currently have access to this article.