abstract
Strong-motion recordings at Port Hueneme and Pasadena, at hypocentral distances of 24 and 80 km, respectively, have been fit with a number of commonly used dislocation models. The most satisfactory model is a fault in which rupture spreads circularly from the focal point (Savage, 1966). The derived model shows bilateral rupture with a lower bound on the moment of 1.5 × 1024 dyne-cm. With an assumed rupture velocity of 3.1 km/sec the stress drops are in the range of 50 to 200 bars. The moment of the initial rupture is well determined, but much uncertainty can exist in the other fault parameters. For example, if the rupture velocity is 1.55 km/sec, an equally good fit to the Port Hueneme record can be achieved, but a stress drop close to 4000 bars is required. The area of the initial rupture is smaller than the area outlined by the aftershock distribution. There is evidence in the Port Hueneme record that the rupture process involved a multiple earthquake, which would help explain the inconsistency between the initial slip area and the area shown by aftershocks. Studies of one-dimensional wave propagation through the sediments beneath the Port Hueneme station suggest that the surface motions have been amplified by a factor of about 2.8 compared with an equivalent bedrock recording, although the wave forms have not been distorted significantly. The observed motions were compensated for the effect of the sediments before the model fitting.
A comparison of different dislocation models shows that rupture propagation effects are important in fitting the observed motions and that if the finite extent of the rupture surface is accounted for, the ω−2 high-frequency decay of the displacement spectrum often attributed to the Haskell dislocation model should be closer to ω−3.
