講演情報
[S08-25]Unraveling the Complex Rupture Processes of the 2024 Noto Peninsula Earthquake through 3D Wavefield Simulations and MCMC Inversion
*榎本 倫太郎1、金子 善宏1、唐 啓賢2、福島 洋2、後藤 浩之3、宮﨑 真一1 (1. 京都大学大学院理学研究科、2. 東北大学災害科学国際研究所、3. 京都大学防災研究所)
Numerical modeling of large intraplate earthquakes is often challenged by limited near-field data. The 2024 Noto Peninsula earthquake was observed by numerous near-field strong-motion seismograms, high-rate GNSS, and satellite observations. Many published fault models reveal complexity in the rupture process, and show considerable differences in rupture velocity and total event duration. We conducted wavefield simulations of the Noto Peninsula earthquake, integrating 3D velocity and inelastic attenuation models, and regional topography and bathymetry structure. To reduce the number of unknown parameters, we used a previously published fault slip model that fixed the fault geometry and final slip distribution derived from SAR and GNSS data and adjusted only the rupture timing and rise time of individual fault segments.We validate our model using all available near-field strong motion data, resulting in a well-constrained source process model. Our model constrains rupture velocity to ~1.4 to ~2.1 km/s and total event duration to 48 s. The preferred source model reveals multiple slip episodes and complex rupture evolution, including a backward-propagating rupture toward the mainshock hypocenter likely triggered by abrupt changes in local fault geometry. Comparison with two published models that relied on fewer near-field stations shows that our model achieves significantly lower misfit. We also explored the effect of velocity structure on resulting waveforms and found that near-field waveforms did not change significantly, and far-field waveforms exhibited a large misfit. Since the velocity structure in not well constrained, these results indicate the necessity of constraining the fault rupture using the near-field data, and by extension, the necessity of dense observation near the earthquake fault.Finally, using a Markov chain Monte Carlo (MCMC) approach, we explored trade-offs among key rupture parameters. This method enhances objectivity and provides new insights into rupture propagation in large intraplate earthquakes.
