Broadband ground motion simulation procedures typically utilize physics-based modeling of source and path effects at low frequencies coupled with semi-stochastic procedures at high frequencies. The high-frequency procedure considered here combines a deterministic Fourier amplitude spectrum that is a function of closed-form source, path, and site models with a random phase. Previous analysis of the simulation procedure in the ShakeOut exercise demonstrated faster distance attenuation and lower intra-event dispersion of high-frequency ground motions than in empirical ground motion equations. We increase crustal damping (Q ) to remove the distance attenuation bias and introduce random site-to-site variations to the Fourier amplitudes using a log-normal standard deviation ranging from 0.45 for Mw < 7 to zero for M w 8. We repeat the simulation of the ShakeOut event with the increased crustal damping and a revised source characterization, with increased slip heterogeneity reflecting more recent recommendations. The revised simulation procedure for ShakeOut produced ground motions without a distance attenuation bias and with near-source dispersion that is generally compatible with empirical models. However, far-field dispersion remains lower than empirical models.
Full List of PEER Reports: click here.
