FAULT SEGMENTATION AND EARTHQUAKE HAZARDS: HISTORIC AND PRE-HISTORIC RUPTURE ACCOMPANYING LARGE REVERSE FAULT EARTHQUAKES

Extensive surface rupture on reverse faults accompanying large earthquakes is rare globally. During the past century only a few large seismic ruptures have been well documented along a continental reverse fault. Among the few examples are the ruptures that caused the 1896 M~7.2 Rikuu (Japan), the 1952 M~7.5 Arvin Tehachapi (U.S.), the 1978 M~7.5, Tabas (Iran), the 1932 M ~ 7.6 Changma (China), and the 1999 Mw 7.6 Chi-Chi (Taiwan) earthquakes. The 1999 Chi-Chi earthquake has sparked questions about the potential size of an earthquake generated along reverse faults at a time when the use of fault-segment length in earthquake size prediction is under debate. Although large ruptures commonly terminate at geometrical complexities such as extensional stepovers or jogs, geologic and paleoseismic data suggest that, in general, segment boundaries have not proven to be a major impediment to rupture. For example, during the 1999 Chi-Chi earthquake, rupture nucleated at depth near the southern terminus of the Chelungpu fault, and propagated upward and northward across four segments. Average surface slip of about 5 m occurred along the southern fault segments. The rupture is an appropriate analog for large events on the Sierra Madre fault, located along the northern margin of the Los Angeles metropolitan area and Seattle fault in metropolitan Seattle. Like the Chelungpu, the Seattle and Sierra Madre faults are about 60-70 km long, with clear structural terminations at both ends. Paleoseismic investigations along the Seattle and Sierra Madre faults suggest that past ruptures of the thrust fault involved about 5 m of slip. Damage from large magnitude earthquakes along these faults would be substantially different than that produced in the 1994 Mw=6.7 Northridge blind earthquake thrust earthquake.