Integrated seismic-hazard analysis of the Wasatch Front, Utah

We examined the combined effects of different sources that influence earthquake hazard of the populated Wasatch Front, Utah. We first evaluated the fault-stress interaction of the two largest historic earthquakes of the Intermountain Seismic Belt (ISB): the 1959 Hebgen Lake, Montana (M_s 7.5) and the 1983 Borah Peak, Idaho (M_s 7.3) earthquakes, which experienced multisegment, normal-faulting ruptures. Estimates of the static-stress change for these events revealed an increase in Coulomb failure stress in areas of extended aftershocks. These observations suggested that fault-stress analysis is applicable in evaluating the spatial distribution of aftershocks after large, normal-faulting earthquakes in the same extensional-stress regime of the eastern Basin-Range, including the Wasatch Front, Utah, which encompasses the 370-km-long Wasatch Fault and surrounding faults. On the basis of this result, we applied the modeling technique to the historically seismically quiescent Wasatch Fault and examined the relation between the pattern of stress change and the space-time distribution of paleoearthquakes. Ages and locations of the Wasatch Fault paleoseismic data imply 17 single-segment or 11 multisegment ruptures in the past 5.6 kyr. We prefer the multisegment model because almost all of the large, historical, normal-faulting earthquakes in the ISB were multisegment. Fitting the along-fault displacements by an analytic half-ellipse function provides a first-order distribution of the displacement over an entire rupture length and allows new estimates of fault slip rates and seismic moment. With these data, we also estimated the occurrence rate of M_w > 6.6 Wasatch paleoearthquakes and showed that the rate was about three times higher than that inferred by the historical seismicity. This result, along with new information from a "megatrench" on the Wasatch Fault near Salt Lake City, that revealed no scarp-forming earthquakes between ca. 9.0 and 15.5 ka, suggests that large Wasatch Fault events in the past 5.6 kyr may be clustered. The recurrence rate of large earthquakes estimated by the geodetic-measured strain, on the other hand, is three to four times higher than that estimated by the long-term fault-slip rate. This difference, together with the observation of a low regional strain rate from historical seismicity, suggests that cumulative, aseismic deformation may be significant on the Wasatch Fault. We also examined a major consideration for along-strike segmentation, namely, how stress "contagion" could affect the probability of failure of adjacent faults. Including paleoearthquake-derived fault-slip rates, global positioning system derived geodetic moment rates, and the effect of stress contagion in the earthquake-hazard estimation for a specific location in the Salt Lake Valley revealed an increase in the annual frequency of peak ground acceleration ≥ 0.25g by a factor of 1.4, 4.0, and 5.4, respectively, compared with that derived from historical seismicity only.