Group B, Poster #226, Ground Motions

Data-constrained fault zone structure improves 0-3 Hz deterministic ground motion predictions for the 2019 M7.1 Ridgecrest earthquake

Te-Yang Yeh, & Kim B. Olsen
Poster Image: 

Poster Presentation

2022 SCEC Annual Meeting, Poster #226, SCEC Contribution #12257 VIEW PDF
We have simulated 0-3 Hz deterministic wave propagation in the Southern California Earthquake Center (SCEC) Community Velocity Model (CVM) version CVM-S4.26-M01 for the 2019 Mw 7.1 Ridgecrest earthquake incorporating a data-constrained high resolution fault zone model (Zhou et al., 2022). The aim of the simulations were to investigate the effects of the near-fault low-velocity zone (LVZ) on near-source and far-field ground motions. Strong motion data recorded at 161 stations were used to estimate the optimal parameters for the near-surface geotechnical layer (GTL) and frequency-dependent anelastic attenuation in the model domain, extending from the Ridgecrest area into the Los Angeles basin.... The optimal model was used to quantify the individual effects of the near-fault LVZ and the GTL on the ground motions. Our results show that the near-fault LVZ significantly perturbs the predicted ground motions in the near-source region, and increases the peak ground velocities in the western Los Angeles Basin (200 km from the source). The fault zone structure improves modeling of the long-period features in the data and lengthens the coda wave trains, in better agreement with observations. On the other hand, a calibrated GTL is the model feature that most significantly improves the spectral energy toward the observed level. Additionally, the simulated basin amplification was validated against the observations, which were computed based on spectral accelerations (SA) at different periods as well as Effective Amplitude Spectra (EAS). The numerical simulation is able to replicate the observed complex spatial pattern of basin amplification at 1 s period, as well as the smoother patterns coherent with the basin structure that emerge at periods longer than 3 s. We found larger misfits between observed and simulation at 1 and 3 s periods, whereas the model performs very well for 6 s and 8 s periods. Using simple 1D transfer functions, we show that deepening of the basin structure is capable of reducing the misfit between observed and simulated amplification for spectral accelerations at periods of 3 s and longer.