科研之友
微信
新浪微博
Facebook
分享链接
摘要
Based on the results obtained by analysing a large number of dynamic strike slip and dipping fault ruptures we construct a kinematic rupture model generator that incorporates some of the key source parameters extracted from the dynamic rupture models. In this kinematic rupture model, the slip rate function includes the final slip, the rise time, the local rupture velocity and the peak time (a measure of the impulsive part of the slip rate function). A four-dimensional correlation matrix is used to describe the spatial interdependency between the four source parameters-each of which is modelled as correlated random field. Each source parameter has a different marginal distribution determined from the analysis of the dynamic models. The marginal distributions are allowed to change as a function of distance from the hypocentre. The autocorrelation of each parameter is modelled by a power spectrum that follows a power law. The value of the spectral decay of the power law for the different source parameters is based on the results obtained from the dynamic rupture models. Finally, the values of rise time and peak time are adjusted such that the moment rate function fits a Brune spectrum for a specified corner frequency. We validate the rupture model generator using observed strong motion near field recording for the 1994 Northridge earthquake and the 1989 Loma Prieta earthquake. We perform comparisons with the rupture model generator of Liu et al. We find that only considering the response spectral bias of a best model is not sufficient to validate a rupture model generator. Thus we test the predictive power of the two rupture model generators if multiple ruptures are computed. Overall, the new rupture model generator yields a better prediction, compared to Liu et al., for the two validation events, especially in predicting observed PGV values and pseudospectral velocity at low frequencies (< 1 Hz).
- 出版日期2013-3
