EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Structure‐Aware Bayesian Full Waveform Inversion via Diffusion‐Regularized Stein Variational Gradient Descent

Chao Li, Mrinal K. Sen, Sergey Fomel, Yangkang Chen

Journal of Geophysical Research: Machine Learning and Computation · 2026-12 · 3 (6)

DOI: 10.1029/2026jh001583 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Geophysical Research: Machine Learning and Computation, online erschienen am 19.11.2026. Verfasst von Chao Li und 3 weiteren Autor:innen. Die Arbeit nennt 57 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

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 DM 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.

Zitieren

APA

Li, C., Sen, M. K., Fomel, S., & Chen, Y. (2026). Structure‐Aware Bayesian Full Waveform Inversion via Diffusion‐Regularized Stein Variational Gradient Descent. Journal of Geophysical Research: Machine Learning and Computation, 3(6). https://doi.org/10.1029/2026jh001583

Vancouver

Li C, Sen MK, Fomel S, Chen Y. Structure‐Aware Bayesian Full Waveform Inversion via Diffusion‐Regularized Stein Variational Gradient Descent. Journal of Geophysical Research: Machine Learning and Computation. 2026;3(6). doi:10.1029/2026jh001583

BibTeX
@article{Li2026,
  author = {Li, Chao and Sen, Mrinal K. and Fomel, Sergey and Chen, Yangkang},
  title = {Structure‐Aware Bayesian Full Waveform Inversion via Diffusion‐Regularized Stein Variational Gradient Descent},
  journal = {Journal of Geophysical Research: Machine Learning and Computation},
  year = {2026},
  volume = {3},
  number = {6},
  doi = {10.1029/2026jh001583}
}
Kostenlos starten →