Time–Frequency Representation and Spatial Correlation Model for Ground Motions with Different Fault Mechanisms

The different fault mechanisms of earthquakes can lead to distinct time–frequency representations of ground motions. Describing the time–frequency representation of ground motions involves utilizing the time–frequency power spectral density (TFPSD) and applying the time–frequency analysis method (i.e., S transform). However, the models developed in the literature mostly focus on strike-slip earthquake only, and, most importantly, the spatial correlation of the parameters for the developed model was never investigated. In the present study, probabilistic models of the TFPSD, correlated with three commonly used seismological parameters—moment magnitude, rupture distance, and shear-wave velocity of the top 30 m of soil—for earthquake ground motions specifically related to the reverse and normal fault mechanisms are developed. The developed models employ 10 parameters to effectively capture the complexity and variability of TFPSD characteristics for a given ground motion. Further, a spatial correlation model of the parameters is also proposed for different scenario events by investigating the spatial correlation between the residuals of the developed model. Finally, an illustrative example is given to show the use of the developed TFPSD model and spatial correlation model for the simulation of ground motions within a regional field.