Consideration of the rupture model uncertainties in the probabilistic seismic hazard analysis

Abstract

The uncertainties associated with the fault rupture model, i.e. down dip rupture width, subsurface rupture length and fault’s dip angle are not considered in the conventional probabilistic seismic hazard analyses (PSHA). The dip angle as an epistemic uncertainty is commonly assumed as a fixed value for a specific fault. Additionally, down dip rupture width and subsurface rupture length are determined from empirical relations.The main hypothesis of this paper was that the results of PSHA may be significantly influenced by considering the rupture model parameters as random variables instead of fixed values. The Monte Carlo simulation, as a powerful tool for uncertainty propagation analysis, was used for this objective. The NGA-West 2 database, as well as the Wells and Coppersmith (1994) [4] study was applied to describe the rupture model parameters in a stochastic manner. The results confirmed that hazard values derived from the Monte Carlo simulation method are significantly different from those derived from the classical approach. Depending on the attenuation relation and the style of faulting, theses differences leads to over/underestimation of hazard, especially at shorter periods in a characteristic band of region.