Increasingly, fault slip rates are being used to constrain earthquake recurrence relationships for site-specific probabilistic seismic hazard (ground motion) assessments. This paper shows the sensitivity of seismic hazard assessments to variations in recurrence models and parameters that incorporate fault slip rates. Two models are considered to describe the partitioning of the slip rate or seismic moment rate into various magnitude earthquakes: an exponential magnitude distribution and a characteristic earthquake distribution. Assuming an exponential distribution, the activity rate, N(m0), is constrained by the upper bound magnitude, mu, the b-value for the region and the fault slip rate, S. For a given S, variations in mu and b-value have significant effects on recurrence and computed hazard, depending on whether the assumption is made that the seismicity rate is constant or the moment rate is constant.
There is increasing evidence that the characteristic earthquake model is more appropriate for individual faults than the exponential magnitude distribution. Based on seismicity data from areas having repeated large earthquakes, a generalized recurrence density function is developed, and the resulting recurrence relationship requires only mu, b-value, and S. A comparison of the recurrence relationships from this model with the historical seismicity and paleoseismicity data on the Wasatch and San Andreas faults shows a good match. The computed hazard based on the characteristic earthquake model differs from that obtained for the exponential model as a function of the fault-to-site distance and the acceleration level. One check on the recurrence models using slip rate is to compare over large regions activity rates based on seismicity data and slip rate data. Such a comparison in the western Transverse Ranges shows a reasonable match for both the exponential and characteristic earthquake models in the moderate-to-large magnitude range.