Ambient Noise Full Waveform Inversion with Neural Operators

2025-03-19

AI Summary

Conventional full-waveform inversion (FWI) suffers from high computational cost, cycle-skipping issues in adjoint-state-based gradient computation, and limited utilization of multi-phase seismic information due to the Born approximation. Method: This work introduces neural operators—specifically Fourier Neural Operators (FNO) and DeepONet architectures—for the first time into real ambient-noise FWI, enabling fast, differentiable numerical solutions of the elastic wave equation. The approach integrates automatic differentiation, ambient-noise preprocessing, and multi-profile nodal array data, thereby circumventing iterative adjoint-state computations and linearization constraints. Contribution/Results: Applied to field data from the Los Angeles Basin, the method successfully reconstructs high-resolution shear-wave velocity structures. Inversion speed exceeds that of conventional FWI by over two orders of magnitude, while significantly enhancing joint multi-phase inversion capability. This establishes a novel, efficient, and fully differentiable modeling paradigm for large-scale seismic imaging.

Technology Category

• Search and Optimization: Mixed Discrete/Continuous Search
• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Learning & Optimization for CV

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

Numerical simulations of seismic wave propagation are crucial for investigating velocity structures and improving seismic hazard assessment. However, standard methods such as finite difference or finite element are computationally expensive. Recent studies have shown that a new class of machine learning models, called neural operators, can solve the elastodynamic wave equation orders of magnitude faster than conventional methods. Full waveform inversion is a prime beneficiary of the accelerated simulations. Neural operators, as end-to-end differentiable operators, combined with automatic differentiation, provide an alternative approach to the adjoint-state method. Since neural operators do not involve the Born approximation, when used for full waveform inversion they have the potential to include additional phases and alleviate cycle-skipping problems present in traditional adjoint-state formulations. In this study, we demonstrate the application of neural operators for full waveform inversion on a real seismic dataset, which consists of several nodal transects collected across the San Gabriel, Chino, and San Bernardino basins in the Los Angeles metropolitan area.

Problem

• Accelerates seismic wave simulation using neural operators.
• Improves full waveform inversion efficiency and accuracy.
• Addresses cycle-skipping issues in traditional adjoint-state methods.

Innovation

• Neural operators accelerate seismic wave simulations.
• Neural operators enable full waveform inversion efficiently.
• Neural operators alleviate cycle-skipping in seismic analysis.