Scholarlink

<p>In seismic reliability analysis of engineering structures, stochastic ground-motion time histories are routinely required as input excitation. The stochastic finite-fault method has become a cornerstone tool for high-frequency ground-motion simulation in engineering seismology. However, the conventional EXSIM implementation can be further improved in its treatment of corner frequency, source duration, and phase representation. In this study, several representative source- and phase-related updates previously reported in the literature were integrated within the Ground Motion Simulation System (GMSS), a MATLAB-based stochastic finite-fault simulation framework. Compared with EXSIM, the GMSS incorporates a rupture-velocity-dependent corner frequency and source duration model that physically links normalized rupture velocity (<i>V</i><sub>rup</sub>/<i>β</i>₀) to subfault corner frequency. An improved scaling factor <i>H</i><sub><i>ij</i></sub> is introduced by weighting each subfault’s contribution according to its relative slip within the total rupture, thereby achieving greater theoretical consistency in the seismological model that governs subfault Fourier amplitude spectra. Furthermore, the phase spectrum is refined to honor propagation-induced frequency-dependent characteristics. The enhanced method is rigorously validated using the well-constrained finite-fault slip model of the 2013 <i>M</i><sub>W</sub> 6.7 Lushan thrust earthquake, together with regionalized attenuation and site parameters, against recorded strong-motion data at 18 near-field stations. Results demonstrate that GMSS accurately reproduces observed Fourier amplitude spectra (FAS), 5%-damped Pseudo-acceleration response spectra (PSA), and peak ground acceleration (PGA). The incorporation of non-stationary phase characteristics markedly improves the time-domain envelope and phase realism of synthetic waveforms, enhancing the physical fidelity and predictive accuracy of broadband ground-motion simulation. The GMSS framework therefore provides a practical tool for stochastic finite-fault simulation and seismic hazard assessment in regions lacking abundant strong-motion records. Nevertheless, both methods still show limited performance in the low-frequency range below 1&#xa0;Hz.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Integrated stochastic finite-fault simulation of near-fault ground motions: application to the 2013 MW 6.7 Lushan earthquake

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月11日

查看全文
  • Dongdong Yuan,
  • Xueliang Chen,
  • Chen Wang,
  • Chenyue Hu,
  • Lijun Qiu

摘要

In seismic reliability analysis of engineering structures, stochastic ground-motion time histories are routinely required as input excitation. The stochastic finite-fault method has become a cornerstone tool for high-frequency ground-motion simulation in engineering seismology. However, the conventional EXSIM implementation can be further improved in its treatment of corner frequency, source duration, and phase representation. In this study, several representative source- and phase-related updates previously reported in the literature were integrated within the Ground Motion Simulation System (GMSS), a MATLAB-based stochastic finite-fault simulation framework. Compared with EXSIM, the GMSS incorporates a rupture-velocity-dependent corner frequency and source duration model that physically links normalized rupture velocity (Vrup/β₀) to subfault corner frequency. An improved scaling factor Hij is introduced by weighting each subfault’s contribution according to its relative slip within the total rupture, thereby achieving greater theoretical consistency in the seismological model that governs subfault Fourier amplitude spectra. Furthermore, the phase spectrum is refined to honor propagation-induced frequency-dependent characteristics. The enhanced method is rigorously validated using the well-constrained finite-fault slip model of the 2013 MW 6.7 Lushan thrust earthquake, together with regionalized attenuation and site parameters, against recorded strong-motion data at 18 near-field stations. Results demonstrate that GMSS accurately reproduces observed Fourier amplitude spectra (FAS), 5%-damped Pseudo-acceleration response spectra (PSA), and peak ground acceleration (PGA). The incorporation of non-stationary phase characteristics markedly improves the time-domain envelope and phase realism of synthetic waveforms, enhancing the physical fidelity and predictive accuracy of broadband ground-motion simulation. The GMSS framework therefore provides a practical tool for stochastic finite-fault simulation and seismic hazard assessment in regions lacking abundant strong-motion records. Nevertheless, both methods still show limited performance in the low-frequency range below 1 Hz.