New Scaling Relationships of Earthquake Source Parameters for Stochastic Tsunami Simulation
Abstract
New scaling relationships of key earthquake source parameters are developed by uniformly and systematically analyzing 226 finite-fault rupture models from the SRCMOD database (http://equake-rc.info/srcmod/). The source parameters include the fault width, fault length, fault area, mean slip, maximum slip, Box-Cox power, correlation lengths along dip and strike directions, and Hurst number. The scaling relationships are developed by distinguishing tsunamigenic models from non-tsunamigenic models; typically, the former occurs in ocean and has gentler dip angles than the latter. The new models are based on extensive data, including recent mega-thrust events, and thus are more reliable. Moreover, they can be implemented as multivariate probabilistic models that take into account uncertainty and dependency of the multiple source parameters. The comparison between new and existing models indicates that the new relationships are similar to the existing ones for earthquakes with magnitudes up to about 8.0, whereas the relationships for the fault width and related parameters differ significantly for larger mega-thrust events. An application of the developed scaling relationships in tsunami hazard analysis is demonstrated by synthesizing stochastic earthquake source models in the Tohoku region of Japan. The examples are aimed at providing practical guidance as to how the developed scaling relationships can be implemented in stochastic tsunami simulation. The numerical results indicate that the effects of magnitude scaling of the source parameters and their uncertainties have major influence on the tsunami hazard assessment.
