Structure-aware Bayesian full waveform inversion via diffusion-regularized Stein variational gradient descent
Abstract
Full waveform inversion (FWI) is a powerful tool for high-resolution subsurface model reconstruction, but it remains highly nonlinear and ill-posed, especially when the initial model is inaccurate or the acquired data are noisy and incomplete. Annealed Stein variational gradient descent (ASVGD) provides a particle-based Bayesian framework for FWI by approximating the posterior distribution with an ensemble of velocity models, thereby enabling uncertainty quantification beyond a single deterministic solution. However, conventional ASVGD-FWI commonly relies on simple handcrafted priors, such as Gaussian or smoothing assumptions, which are often insufficient to represent complex geological structures. To address this limitation, we introduce a learned diffusion model (DM) as an implicit regularization mechanism within the ASVGD-FWI framework. Instead of explicitly evaluating the diffusion prior density, the trained diffusion model acts as a learned geological projection operator that guides updated particles toward the distribution of plausible velocity models. In this way, the proposed method combines data consistency from FWI, posterior diversity from ASVGD, and structural realism from the diffusion prior. Numerical experiments demonstrate that the diffusion-regularized ASVGD-FWI produces more geologically consistent mean models, suppresses nonphysical artifacts, and yields uncertainty estimates that are more concentrated around structurally complex and poorly constrained regions. These results suggest that learned diffusion priors can effectively enhance Bayesian FWI by improving both inversion stability and the interpretability of posterior uncertainty.
Related
- Bayesian Geophysical Inversion using Stein Mixture Variational Inference for Multimodal or Complex Solutions
- Bayesian Joint Velocity and Impedance Inversion via Diffusion Models Conditioned on Common Image Gathers
- Multi-Modal Surface Wave Inversion using Physics Guided Variational Autoencoders
- Bayesian inversion of multilayer CO2 migration from seismic plume observations using a graph-based finite-rate invasion-percolation model
- Separating Weight and Likelihood Tail Effects in Bayesian Neural Groundwater Inversion
BibTeX
@misc{li2026structureaware,
title = {{Structure-aware Bayesian full waveform inversion via diffusion-regularized Stein variational gradient descent}},
author = {Chao Li and Mrinal K. Sen and Sergey Fomel and Yangkang Chen},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.19187},
eprint = {2610.19187},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.19187},
}
RIS
TY - UNPB
TI - Structure-aware Bayesian full waveform inversion via diffusion-regularized Stein variational gradient descent
AU - Chao Li
AU - Mrinal K. Sen
AU - Sergey Fomel
AU - Yangkang Chen
PY - 2026
DA - 2026/10/13
PB - arXiv
DO - 10.48550/arxiv.2610.19187
UR - https://arxiv.org/abs/2610.19187
AB - Full waveform inversion (FWI) is a powerful tool for high-resolution subsurface model reconstruction, but it remains highly nonlinear and ill-posed, especially when the initial model is inaccurate or the acquired data are noisy and incomplete. Annealed Stein variational gradient descent (ASVGD) provides a particle-based Bayesian framework for FWI by approximating the posterior distribution with an ensemble of velocity models, thereby enabling uncertainty quantification beyond a single deterministic solution. However, conventional ASVGD-FWI commonly relies on simple handcrafted priors, such as Gaussian or smoothing assumptions, which are often insufficient to represent complex geological structures. To address this limitation, we introduce a learned diffusion model (DM) as an implicit regularization mechanism within the ASVGD-FWI framework. Instead of explicitly evaluating the diffusion prior density, the trained diffusion model acts as a learned geological projection operator that guides updated particles toward the distribution of plausible velocity models. In this way, the proposed method combines data consistency from FWI, posterior diversity from ASVGD, and structural realism from the diffusion prior. Numerical experiments demonstrate that the diffusion-regularized ASVGD-FWI produces more geologically consistent mean models, suppresses nonphysical artifacts, and yields uncertainty estimates that are more concentrated around structurally complex and poorly constrained regions. These results suggest that learned diffusion priors can effectively enhance Bayesian FWI by improving both inversion stability and the interpretability of posterior uncertainty.
ER -