Interfrequency Correlation of Characterized Source Modeling for Broadband Ground Motion Validation

Abstract:

Interfrequency correlation of simulated broadband ground motion is mostly affected by source modeling and Green’s function modeling. Several investigations are performed and modification are proposed for simulation methods on the SCEC Broadband Platform using the observed interfrequency correlation as a target. Here we investigate the performance of characterized source model that consists of asperities and the background slip area. This binarized source is originally based on the Das-Kostrov asperity model, and the correlation is high inside the same area, and very low for outside of the area. However, interfrequency correlation for simulated ground motions are similar to that of the observations. The low-frequency part of the Green’s functions is calculated using 1-D velocity models, and the high-frequency part, uses stochastic Green’s functions with random phase. Since the high-frequency content of the simulated ground motion is similar to that of small earthquakes, the simulated and observed interfrequency correlations for target earthquakes is expected to be similar. However, due to 1D wave propagation approximation the simulated and observed interfrequency correlations are different. To investigate this discrepancy, we tested the performance of characterized source modeling in ground motion simulations for the magnitude range of Mw 4.0 to Mw 8.0. We focused on the possible dependency of interfrequency correlation with magnitude. We found that the Fourier spectral content of the Kostrov-like slip-velocity functions, largely affects the low-frequency interfrequency correlation obtained with the characterized source model of the strong ground motion prediction recipe. The effects of the finite source with rupture progression on interfrequency correlation and their spatial distribution as a function of distance and azimuth from a source to stations need to be investigated. Acknowledgement: This project was partly supported by the JSPS Bilateral Open Partnership Joint Research Projects.