Large‐scale ground motion simulation of the 2016 Kumamoto earthquake incorporating soil nonlinearity and topographic effects

Abstract

A large‐scale physics‐based simulation is conducted to investigate ground motion distribution in the Mw 7.0 2016 Kumamoto earthquake, Japan. The model simulates the earthquake scenario from fault rupture to wave propagation, and localized site response with consideration of the combined effect of soil nonlinearity and topographic amplification. The simulation domain is 51 km × 43 km × 25 km, and the obtained ground motion time histories are compared satisfactorily with recordings from KiK‐net and K‐NET. Ground motion distribution considering nonlinear soil response and topographic amplification is presented. A 3D equivalent linear model is developed to mimic the soil nonlinearity, and it is demonstrated that neglecting soil nonlinearity could over‐predict peak ground acceleration (PGA) and underestimate peak ground velocity (PGV) near the fault. The topographic amplification factors (TAFs) of PGA and PGV are found between 0.5 and 2.0, with a correlation coefficient of 0.7 between them. Predictive equations are proposed to correlate TAFs of PGA and PGV with topographic features, which are represented by relative heights obtained at different length scales. Finally, major earthquake damages are summarized with reference to the obtained ground motion intensity map.

Multi‐input integrative neural network for soil seismic response modeling at KiK‐net downhole array sites

“…With the increase in computing power, state-of-the-art methods for two-or three-dimensional models have been developed to simulate soil seismic response in more complex conditions. [15][16][17] However, creating an accurate physics-based soil seismic response model requires extensive prior knowledge of soil site characteristics, including physical and mechanical parameters. Unfortunately, obtaining precise parameters proves challenging due to the inherent complexity of geotechnical materials.…”

“…Nonlinear soil seismic response models are able to incorporate various constitutive models 12–14 and simulate the behavior of diverse soil types in small time steps. With the increase in computing power, state‐of‐the‐art methods for two‐ or three‐dimensional models have been developed to simulate soil seismic response in more complex conditions 15–17 . However, creating an accurate physics‐based soil seismic response model requires extensive prior knowledge of soil site characteristics, including physical and mechanical parameters.…”

“…of training samples processed simultaneously during one iteration of the model's training phase, influencing memory requirements and computational efficiency. We set learning rate ∈ [0.1, 0.01, 0.001, 0.0001], kernel size ∈[11,13,15,17,19], and batch size ∈ [256, 512, 1024, 2048, 4096]. Consequently, a total of 100 models are trained and tested using the dataset described in Section 3.…”

Impact of Randomized Soil Properties and Rock Motion Intensities on Ground Motion

“…Shear wave velocity (V s ) profiles, layering, and material nonlinearity all influence the amplitude of ground shaking. Historical data have consistently demonstrated an evident amplification of ground motion when unfavorable local soil conditions are present [1][2][3][4][5][6][7][8][9]. For example, the M7.8 earthquake that struck Turkey in 2023 resulted in extensive damage to numerous buildings and claimed thousands of lives [6].…”

Engineering validation of BB-SPEEDset, a data set of near-source physics-based simulated accelerograms

“…Although numerical simulations have become, starting from the 60s of the past century, the core of most engineering activities, including earthquake engineering, approaches for PBS are still relatively young, owing, on one side, to the high computational burden to deal with large multi-scale domains (Chen et al, 2023; Gatti et al, 2018; Infantino et al, 2021a; Mazzieri et al, 2013) up to fault-to-structure frameworks (Kato and Wang, 2021; McCallen et al, 2021) and, on the other side, to the complexity in modeling, within a broad frequency range, the coupled effect of the seismic source, the propagation path (including models for P- and S-waves quality factors), and the shallow geological layers (possibly including nonlinear soil response).…”