Large-Scale 3D Ground-Motion Synthesis with Physics-Inspired Latent Operator Flow Matching

TL;DR

This paper introduces GMFlow, a physics-inspired generative model that rapidly produces large-scale, realistic 3D ground-motion time histories for earthquake hazard analysis, significantly reducing computational costs.

Contribution

The paper presents GMFlow, a novel flow-based generative framework that efficiently synthesizes large-scale ground-motion data conditioned on physical parameters, enabling rapid hazard assessment.

Findings

Generates spatially coherent ground motion across 9 million points in seconds.

Achieves a 10,000-fold speedup over traditional physics-based simulations.

Validates effectiveness on simulated earthquake scenarios in the San Francisco Bay Area.

Abstract

Earthquake hazard analysis and design of spatially distributed infrastructure, such as power grids and energy pipeline networks, require scenario-specific ground-motion time histories with realistic frequency content and spatiotemporal coherence. However, producing the large ensembles needed for uncertainty quantification with physics-based simulations is computationally intensive and impractical for engineering workflows. To address this challenge, we introduce Ground-Motion Flow (GMFlow), a physics-inspired latent operator flow matching framework that generates realistic, large-scale regional ground-motion time-histories conditioned on physical parameters. Validated on simulated earthquake scenarios in the San Francisco Bay Area, GMFlow generates spatially coherent ground motion across more than 9 million grid points in seconds, achieving a 10,000-fold speedup over the simulation workflow, which opens a path toward rapid and uncertainty-aware hazard assessment for distributed infrastructure. More broadly, GMFlow advances mesh-agnostic functional generative modeling and could potentially be extended to the synthesis of large-scale spatiotemporal physical fields in diverse scientific domains.