The 25 November 1988 Saguenay, Quebec, earthquake: Source parameters and the attenuation of strong ground motion

Abstract

The Saguenay earthquake of 25 November 1988 occurred close to the southern margin of the Saguenay Graben in southern Quebec. It was caused by almost purely dip-slip faulting centered at a depth of 26 km with a P axis oriented northeast-southwest. This faulting mechanism is similar to those of the larger historical earthquakes in eastern North America, but the focal depth is substantially greater than all but one of these events. The seismic moment estimated from regional PnI waves and teleseismic long-period body waves is 5 × 10^(24) dyne-cm., corresponding to a moment magnitude of 5.8. The source duration of the earthquake is estimated to be 1.8 sec, corresponding to a stress drop of 160 bars, which is not significantly higher than the average stress drop of 120 bars estimated from previous large earthquakes in eastern North America. In order to simultaneously match the recorded ground motion amplitudes of strong-motion acceleration, strong-motion velocity, and teleseismic short-period and long-period body waves, it is necessary to use a source function having a complex shape that implies the presence of asperities and larger local stress drops. The large set of strong-motion recordings of the Saguenay earthquake has been used to validate a procedure for estimating strong ground motion attenuation based on a simple wave propagation model. The most important feature of the recorded strong motions is that their peak amplitudes do not decay significantly with distance inside 120 km, but then decay abruptly beyond 120 km. Profiles of recorded accelerograms with absolute times indicate that at distances beyond 64 km the peak ground motions are due to strong postcritical reflections from velocity gradients in the lower crust. The principal shear-wave arrivals and the variation of their peak amplitudes with distance were reproduced in synthetic seismograms generated using a regional crustal structure model. The critical distances for the postcritical reflections were short because of the deep focal depth of the event, causing the elevation of ground motion amplitudes out to 120 km. Similar studies of earthquakes in other regions of eastern North America indicate that the strength of the postcritical reflections, and the distance ranges over which they are dominant, are controlled by the focal depth and crustal structure. Regional variations in crustal structure thus give rise to predictable regional variations in strong ground motion attenuation.