T H E Horse Canyon Earthquake of August 2, 1975-two-stage Stress-release Process in a Strike-slip Earthquake by Stephen Hartzell and James

A moderate strike-slip earthquake (ML = 4.8) occurred on the San Jacinto fault system about 60 km northwest of the Salton Sea on August 2, 1975. Analysis of main shock and aftershock data suggest that stress release during this earthquake took place in two stages. During one stage faulting occurred over a relatively small source area (source radius of 0 . 5 km), with a rapid dislocaton rate (rise time ~0.1 sec), possibly associated with an asperity on the fault. During the second stage of faulting, the rupture front grew, but at a much slower rate (rise time ~10 sec), to a final Source radius of ~1.0 km. The above model explains the larger moment estimate based on 20-sec surface waves compared to shorter period body-wave estimates, and also the apparent increase in source dimension with time. The model allows for large stress drops over small source dimensions, but when averaged over the final extent of the rupture plane, stress drops are much lower. The rupture of the asperity is characterized by a moment of 6.5 x 1022 dyne-cm and a stress drop of about 225 bars. The total moment is about 3.0 x 1023 dyne-cm with an averaged stress drop over the fault plane of approximately 90 bars and a dislocation of 25 cm. Observations similar to the ones reported on here have been noted for other earthquakes with a wide range of magnitudes, including: a few large earthquakes in Japan, the 1971 San Fernando earthquake and some of its aftershocks, the 1975 Oroville earthquake, and some swarm events in the Imperial Valley. These observations suggest that a two-stage rupture mechanism may be a fairly common o c c u r r e n c e in shallow faulting and may reflect possible large variations in stress over a length scale of kilometers within the crust. INTRODUCTION The Horse Canyon earthquake (ML = 4.8) occurred on the San Jacinto fault system on August 2, 1975 approximately 60 km northwest of the Salton Sea. This earthquake is noteworthy because it was fairly well instrumented, and affords an opportunity to study the source mechanism of an earthquake in a relatively simple geological setting, predominantly hard, granitic rock. Such a location should reduce complications in the records due to scattering. Also, the Horse Canyon earthquake produced no obvious strain precursor at the Pifion Flat Geophysical Observatory (Berger and Wyatt, 1978), on the three-component laser strain meter just 13 km northeast of the epicenter. The mechanism of this earthquake is therefore important from an earthquake prediction standpoint. The reported origin time is 0 hr 14 min 7.5 sec, with an epicenter of 33 ° 31.4'N, 116 ° 33.5'W and a depth of 12 km {Pasadena). The epicenter lies near the trifurcation of the San Jacinto fault, southeast of Anza, California (see Figure 1). The faultplane solution in Figure 1 is a lower hemisphere projection reported by Kanamori (1976). The s h a d e d a r e a s indicate compression. The strike and dip of the fault plane are 307.2 ° and 71.9 ° , respectively, and of the auxiliary plane, 42.2 ° and 75.0 ° , respectively. The solution is based on first motions from 28 Southern California stations and is fairly well constrained. Amplitudes from five stations in the Canadian network at distances greater than 20 ° , give an average body-wave magnitude of 4.3 (Basham, personal communication, 1976). Using 12-sec Rayleigh waves at eight 1161 117000 ' 3 4 0 0 0 ' 1162 STEPHEN HARTZELL AND JAMES N. BRUNE Canadian stations and the procedure of Marshall and Basham (1972}, an average surface-wave magnitude of 3.6 is obtained. Using the mb versus Ms plot of Marshall and Basham (1972), the Horse Canyon earthquake discriminates from explosions as expected. About a dozen aftershocks in the magnitude range from 1.0 to 2.0 were recorded at Palomar (PLM) 38 km from the epicenter, within the first 12 hr following the main shock. The epicentral region is characterized by a high level of microearthquake activity,