Multi-Modal Surface Wave Inversion using Physics Guided Variational Autoencoders
Abstract
The growth in seismic acquisition techniques has led to large-scale surveys, making traditional manual dispersion analysis logistically and computationally prohibitive. While deep learning techniques can offer an automated alternative, standard convolutional neural networks (CNNs) often fail to account for the non-uniqueness of the inverse problem and can produce geologically unrealistic, smooth velocity models. To address these challenges, we propose a physics-guided multi-modal variational autoencoder (PG-VAE) for scalable near-surface characterization. Our framework introduces a multi-stream fusion strategy that maps independent latent features from both raw shot gathers and spectral-dispersion images by leveraging the physics of Rayleigh-wave dispersion. We implement a specialized decoder architecture with learned transposed convolutions and residual refinement to enforce a structural bias towards layered earth models. Physics guidance is incorporated via a differentiable surrogate network that serves as a forward-consistency regularizer, penalizing predictions that violate the physics of Rayleigh-wave dispersion. The network is trained on diverse synthetic data generated via elastic modeling, and then validated on unseen synthetic data, and subsequently evaluated on an ultra-high-density field dataset from the Devine test site in Texas. We also employ gradient-weighted class activation mapping (Grad-CAM) to examine the input features contributing to the network predictions. The proposed method could serve as a scalable tool for near-surface characterization and as an initial model-building tool for high-resolution elastic full-waveform inversion.
Related
- EnviFlux (v1.0): a simplified surface flux inversion tool based on four-dimensional variational data assimilation (4D-Var)
- Structure-aware Bayesian full waveform inversion via diffusion-regularized Stein variational gradient descent
- Bayesian Geophysical Inversion using Stein Mixture Variational Inference for Multimodal or Complex Solutions
- Linear geophysical inversion with seismic image-guided kriging
- DispSAC: Surface-wave dispersion inversion using soft actor-critic reinforcement learning
BibTeX
@misc{swaminadhan2026multimodal,
title = {{Multi-Modal Surface Wave Inversion using Physics Guided Variational Autoencoders}},
author = {Sujith Swaminadhan and Andrey Bakulin and Sergey Fomel and Yangkang Chen},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.02337},
eprint = {2610.02337},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.02337},
}
RIS
TY - UNPB
TI - Multi-Modal Surface Wave Inversion using Physics Guided Variational Autoencoders
AU - Sujith Swaminadhan
AU - Andrey Bakulin
AU - Sergey Fomel
AU - Yangkang Chen
PY - 2026
DA - 2026/10/01
PB - arXiv
DO - 10.48550/arxiv.2610.02337
UR - https://arxiv.org/abs/2610.02337
AB - The growth in seismic acquisition techniques has led to large-scale surveys, making traditional manual dispersion analysis logistically and computationally prohibitive. While deep learning techniques can offer an automated alternative, standard convolutional neural networks (CNNs) often fail to account for the non-uniqueness of the inverse problem and can produce geologically unrealistic, smooth velocity models. To address these challenges, we propose a physics-guided multi-modal variational autoencoder (PG-VAE) for scalable near-surface characterization. Our framework introduces a multi-stream fusion strategy that maps independent latent features from both raw shot gathers and spectral-dispersion images by leveraging the physics of Rayleigh-wave dispersion. We implement a specialized decoder architecture with learned transposed convolutions and residual refinement to enforce a structural bias towards layered earth models. Physics guidance is incorporated via a differentiable surrogate network that serves as a forward-consistency regularizer, penalizing predictions that violate the physics of Rayleigh-wave dispersion. The network is trained on diverse synthetic data generated via elastic modeling, and then validated on unseen synthetic data, and subsequently evaluated on an ultra-high-density field dataset from the Devine test site in Texas. We also employ gradient-weighted class activation mapping (Grad-CAM) to examine the input features contributing to the network predictions. The proposed method could serve as a scalable tool for near-surface characterization and as an initial model-building tool for high-resolution elastic full-waveform inversion.
ER -