← Back to Gallery

A Data-driven Ground Motion Synthesis Framework, With Physics-informed Broadband Time-series and Nonlinear Site Response

The proliferation of seismic record repositories through the expansion of seismic networks and high-performance computer simulations of earthquake scenarios, have made available ground-motion databases rich enough to make possible data-driven models of ground-motion synthesis that account for complex site effects. Data-driven methods offer a novel approach to describing these processes by directly learning the governing laws from sufficiently rich training data, while avoiding the use of simplified assumptions that limit the realism of models developed with traditional statistical tools. In this study, we demonstrate this new paradigm of learning the underlying physics in a data-driven framework –and quantifying as such, the sources of ground motion epistemic uncertainty, by coupling: (i) a Generative Adversarial Neural Operator (GANO), which synthesizes broadband three-component acceleration time histories at reference outcrop conditions, with (ii) a Fourier Neural Operator (FNO), which modifies the outcrop ground motions to account for the full nonlinear response of the near-surface soil layers. GANO was trained on a synthetic ground motion dataset for strike-slip and reverse events developed using the Southern California Earthquake Center (SCEC) Broadband Platform (BBP), while FNO was trained on non-linear one-dimensional wave propagation through smooth Bay Area velocity profiles using the site-response software, PySeismoSoil. A key advantage of the neural operator architectures in GANO and FNO compared to traditional neural networks, is their ability to learn the mapping between continuous function spaces as opposed to finite-dimensional sets, rendering the training and application of the model resolutions invariant (i.e., training can include input signals of different sampling frequencies without loss of information or generation of artifacts, while prediction can be performed on sampling frequencies independent of training). Verification analyses through residual and goodness of fit evaluations demonstrate that GANO can learn the magnitude and distance scaling, while FNO can correctly estimate the non-linear amplification for ground motions and profiles not included in the training dataset for 0.1 to 30hz frequency range. By appropriately conditioning data-driven algorithms, or combinations thereof like in this case, our work demonstrates the potential of using these methods to learn increasingly complex physics and their uncertainty over the entire frequency range of engineering interest, and to generate on demand time-histories appropriate for engineering design with high degree of realism. 

Session: From Earthquake Recordings to Empirical Ground-Motion Modelling - II

Type: Oral

Room: Tubughnenq’ 4

Date: 5/2/2024

Presentation Time: 10:45 AM (local time)

Presenting Author: Feiruo Xia

Student Presenter: Yes

Additional Authors