Plug-and-Play Stein variational gradient descent for Bayesian post-stack seismic inversion

Izzatullah, Muhammad; Alkhalifah, Tariq; Romero, J.; Corrales, M.; Luiken, Nick; Ravasi, Matteo

doi:10.3997/2214-4609.202310177

Cited by 3 publications

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In petrophysical inversion context, segmentation may be particularly beneficial as the link between facies and petrophysical properties is more direct than the link between facies and acoustic/elastic parameters as considered in Ravasi and Birnie (2022). In addition, our framework also permits the integration of new promising deep-based algorithms like the Plug-and-Play method with CNN- Based Denoisers (Romero et al, 2022) and its probabilistic extension (Corrales, Izzatullah, et al, 2022;Izzatullah et al, 2022).…”

Section: Discussionmentioning

confidence: 99%

Seis2Rock: A Data‐Driven Approach to Direct Petrophysical Inversion of Pre‐Stack Seismic Data

Corrales,

Hoteit,

Ravasi

2024

Earth and Space Science

Self Cite

View full text Add to dashboard Cite

The inversion of petrophysical parameters from seismic data represents a fundamental step in the process of characterizing the subsurface, with applications ranging from subsurface resource exploration to geothermal, carbon capture and storage, and hydrogen storage. We propose a novel, data‐driven approach, named Seis2Rock, that utilizes optimal basis functions learned from well‐log information to create a direct link between pre‐stack seismic data and so‐called band‐limited petrophysical reflectivities. Seis2Rock is composed of two stages: training and inference. During training, a set of optimal basis functions are identified by performing singular value decomposition on one or more synthetic amplitude variation with offset gathers created from measured or rock‐physics synthesized elastic well logs. In inference, seismic pre‐stack data are first projected into a set of band‐limited petrophysical reflectivities using the previously computed basis functions; this is followed by regularized post‐stack seismic inversion of the individual properties. In this work, we apply the Seis2Rock methodology to a synthetic data set based on the Smeaheia reservoir model and to the open Volve field data set. Our numerical results reveal the ability of the proposed method in recovering accurate porosity, shale content, and water saturation models. Finally, the proposed methodology is applied in the context of reservoir monitoring to invert time‐lapse, pre‐stack seismic data for water saturation changes.

show abstract

Section: Discussionmentioning

confidence: 99%

Seis2Rock: A Data‐Driven Approach to Direct Petrophysical Inversion of Pre‐Stack Seismic Data

Corrales,

Hoteit,

Ravasi

2024

Earth and Space Science

Self Cite

View full text Add to dashboard Cite

show abstract

Full-Waveform Inversion Using a Learned Regularization

Sun,

Yang,

Liang

et al. 2023

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

VIP - Variational Inversion Package with example implementations of Bayesian tomographic imaging

Zhang,

Curtis

2024

Seismica

View full text Add to dashboard Cite

Bayesian inference has become an important methodology to solve inverse problems and to quantify uncertainties in their solutions. Variational inference is a method that provides probabilistic, Bayesian solutions efficiently by using optimisation. In this study we present a Python Variational Inversion Package (VIP), to solve inverse problems using variational inference methods. The package includes automatic differential variational inference (ADVI), Stein variational gradient descent (SVGD) and stochastic SVGD (sSVGD), and provides implementations of 2D travel time tomography and 2D full waveform inversion including test examples and solutions. Users can solve their own problems by supplying an appropriate forward function and a gradient calculation code. In addition, the package provides a scalable implementation which can be deployed easily on a desktop machine or using modern high performance computational facilities. The examples demonstrate that VIP is an efficient, scalable, extensible and user-friendly package, and can be used to solve a wide range of low or high dimensional inverse problems in practice.

show abstract

Plug-and-Play Stein variational gradient descent for Bayesian post-stack seismic inversion

Cited by 3 publications

References 0 publications

Seis2Rock: A Data‐Driven Approach to Direct Petrophysical Inversion of Pre‐Stack Seismic Data

Seis2Rock: A Data‐Driven Approach to Direct Petrophysical Inversion of Pre‐Stack Seismic Data

Full-Waveform Inversion Using a Learned Regularization

VIP - Variational Inversion Package with example implementations of Bayesian tomographic imaging

Contact Info

Product

Resources

About