The February 9, 1971 San Fernando earthquake: A study of source finiteness in teleseismic body waves

Teleseismic P, SV, and SH waves recorded by the WWSS and Canadian networks from the 1971 San Fernando, California earthquake (ML = 6.6) are modeled in the time domain to determine detailed features of the source as a prelude to studying the near and local field strong-motion observations. Synthetic seismograms are computed from the model of a propagating finite dislocation line source embedded in layered elastic media. The effects of source geometry and directivity are shown to be important features of the long-period observations. The most dramatic feature of the model is the requirement that the fault, which initially ruptured at a depth of 13 km as determined from pP-P times, continuously propagated toward the free surface, first on a plane dipping 53°NE, then broke over to a 29°NE dipping fault segment. This effect is clearly shown in the azimuthal variation of both long period P- and SH-wave forms. Although attenuation and interference with radiation from the remainder of the fault are possible complications, comparison of long- and short-period P and short-period pP and P waves suggest that rupture was initially bilateral, or, possibly, strongly unilateral downward, propagating to about 15 km depth. The average rupture velocity of 1.8 km/sec is well constrained from the shape of the long-period wave forms. Total seismic moment is 0.86 × 10^(26) dyne-cm. Implications for near-field modeling are drawn from these results.