Development of an Estimation Method for the Seismic Motion Reproducibility of a Three-dimensional Ground Structure Model by combining Surface-observed Seismic Motion and Three-dimensional Seismic Motion Analysis

TL;DR

This work addresses the challenge of obtaining reliable 3D ground-structure models for seismic motion estimation by proposing a surface-observation–driven model-selection framework. The approach uses a Green's-function based analysis to fit surface seismic motions via a parametric input wave $f_i(t)=\sum_j c_{ij} p(t-(j-1)Δt)$, computing surface responses with $G^k_{ij}(m,t)$ and reconstructing $U^k_i(m,t)$, then evaluating $ERR(m)$ and solving $A c = b$ with a stable pseudoinverse via singular value decomposition. Results from a numerical experiment show that increasing surface observation points improves model discrimination, enabling identification of credible models (e.g., model000018) that better reproduce both observed surface motions and the implied input waves, compared to poorer candidates (e.g., model000228); the method is demonstrated on GPU-accelerated 3D seismic analyses with large degrees of freedom, highlighting practical feasibility. The study advances the reliability of 3D ground-structure models for earthquake damage mitigation and suggests avenues to extend parameterization and leverage HPC/AI to refine candidate models further.

Abstract

The ground structure can substantially influence seismic ground motion underscoring the need to develop a ground structure model with sufficient reliability in terms of ground motion estimation for earthquake damage mitigation. While many methods for generating ground structure models have been proposed and used in practice, there remains room for enhancing their reliability. In this study, amid many candidate 3D ground structure models generated from geotechnical engineering knowledge, we propose a method for selecting a credible 3D ground structure model capable of reproducing observed earthquake ground motion, utilizing seismic ground motion data solely observed at the ground surface and employing 3D seismic ground motion analysis. Through a numerical experiment, we illustrate the efficacy of this approach. By conducting $10^2$-$10^3$ cases of fast 3D seismic wave propagation analyses using graphic processing units (GPUs), we demonstrate that a credible 3D ground structure model is selected according to the quantity of seismic motion information. We show the effectiveness of the proposed method by showing that the accuracy of seismic motions using ground structure models that were selected from the pool of candidate models is higher than that using ground structure models that were not selected from the pool of candidate models.

Figures (10)

• Figure 1: Target system (left) and its numerical analysis model (right).
• Figure 2: Reference ground model. The model comprises two layers, with a flat surface and a sedimentary layer with varying thickness. The thickness of the sedimentary layer is illustrated in Fig. \ref{['fig:dem']}a).
• Figure 3: Estimated $ERR$ using one observation point
• Figure 4: Estimated incident wave using one observation point for event #1. Estimation accuracy is low even if $ERR$ is low.
• Figure 5: Estimated $ERR$ using nine observation points