1887
Volume 74, Issue 1
  • E-ISSN: 1365-2478

Abstract

ABSTRACT

We propose a novel physics‐guided deep learning framework for geophysical inversion that integrates Langevin Monte Carlo (LMC) sampling to quantify uncertainties in model parameters. A statistical sampling strategy is employed to enhance computational efficiency by reducing the number of required samples while preserving diversity and informativeness. The training data for the supervised learning networks are iteratively expanded with outputs from a stochastic sampler and their corresponding observed responses, ensuring representative coverage of the model space. The Jensen–Shannon divergence is adopted as the loss function for training the network model, in which the Gaussian assumption is applied to enable analytical computation. The developed workflow is evaluated on reservoir porosity inversion, where it successfully reconstructs porosity patterns in the subsurface, yielding results that closely match the reference model. Compared to traditional LMC algorithm applied to the entire data cube, the proposed approach attains substantial computational efficiency by leveraging an active learning strategy that identifies and utilizes a limited yet representative subset of the observations. The results demonstrate the effectiveness of the proposed method, highlighting its potential for application to a wide range of geophysical inverse problems.

© 2026 European Association of Geoscientists & Engineers. © 2026 European Association of Geoscientists & Engineers.
Loading

Article metrics loading...

/content/journals/10.1111/1365-2478.70130
2026-01-18
2026-02-09

  • From This Site

    /content/journals/10.1111/1365-2478.70130
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Journal -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

References

  1. Azevedo, L., and A.Soares. 2017. Geostatistical Methods for Reservoir Geophysics. Springer Nature.
     [Google Scholar]
  2. Alfarhan, M., M.Deriche, and A.Maalej. 2020. “Robust Concurrent Detection of Salt Domes and Faults in Seismic Surveys Using an Improved U‐Net Architecture.” IEEE Access10: 39424–39435.
     [Google Scholar]
  3. Bernard, E.2022. Introduction to Machine Learning. Wolfram.
     [Google Scholar]
  4. Bishop, C. M.1995. “Training With Noise Is Equivalent to Tikhonov Regularization.” Neural Computation7, no. 1: 108–116.
     [Google Scholar]
  5. Brooks, S., A.Gelman, G.Jones, and X.Meng. 2011. Handbook of Markov Chain Monte Carlo. CRC Press.
     [Google Scholar]
  6. Chen, F., Z.Zong, and X.Yin. 2024a. “Seismic Scattering Inversion for Multiple Parameters of Overburden‐Stressed Isotropic Media.” Geophysics89, no. 6: T319–T336.
     [Google Scholar]
  7. Chen, F., Z.Zong, X.Yin, et al. 2024b. “Anisotropy Parameters Estimation in Stress‐Induced Orthorhombic Reservoirs Based on Step‐Wise Bayesian Inversion of Azimuthal Seismic Data.” IEEE Transactions on Geoscience and Remote Sensing62: 5933512.
     [Google Scholar]
  8. Colombo, D., E.Turkoglu, W.Li, E.Sandoval‐Curiel, and D.Rovetta. 2021a. “Physics‐Driven Deep Learning Inversion With Application to Transient Electromagnetics.” Geophysics86, no. 3: E209–E224.
     [Google Scholar]
  9. Colombo, D., E.Turkoglu, W.Li, and D.Rovetta. 2021b. “Coupled Physics‐Deep Learning Inversion.” Computers and Geosciences157: 104917.
     [Google Scholar]
  10. Colombo, D., E.Turkoglu, E.Sandoval‐Curiel, D.Rovetta, and W.Li. 2023. “Machine Learning Inversion via Adaptive Learning and Statistical Sampling: Application to Airborne Micro‐TEM for Seismic Sand Corrections.” Geophysics88, no. 3: K51–K68.
     [Google Scholar]
  11. Colombo, D., E.Turkoglu, E.Sandoval‐Curiel, and T.Alyousuf. 2024. “Self‐Supervised, Active Learning Seismic Full‐Waveform Inversion.” Geophysics89, no. 2: U31–U52.
     [Google Scholar]
  12. Das, V., A.Pollack, U.Wollner, and T.Mukerji. 2019. “Convolutional Neural Network for Seismic Impedance Inversion.” Geophysics84, no. 6: R869–R880.
     [Google Scholar]
  13. Deutsch, C. V., and A. G.Journal. 1998. GSLIB: Geostatistical Software Library and User's Guide. Oxford University Press.
     [Google Scholar]
  14. Dvorkin, J., and A.Nur. 1996. “Elasticity of High‐Porosity Sandstones: Theory for Two North Sea Data Sets.” Geophysics61, no. 5: 1363–1370.
     [Google Scholar]
  15. Feng, R.2020. “Estimation of Reservoir Porosity Based on Seismic Inversion Results Using Deep Learning Methods.” Journal of Natural Gas Science and Engineering77: 103270.
     [Google Scholar]
  16. Feng, R.2021. “Improving Uncertainty Analysis in Well Log Classification by Machine Learning With a Scaling Algorithm.” Journal of Petroleum Science and Engineering196: 107995.
     [Google Scholar]
  17. Feng, R.2025. “Geophysical Inversion via Hierarchical Bayesian Deep Learning With Statistical Sampling.” Geophysical Prospecting73: e70113. https://doi.org/10.1111/1365‐2478.70113.
     [Google Scholar]
  18. Feng, R., N.Balling, D.Grana, J. S.Dramsch, and T. M.Hansen. 2021. “Bayesian Convolutional Neural Networks for Seismic Facies Classification.” IEEE Transactions on Geoscience and Remote Sensing59, no. 10: 8933–8940.
     [Google Scholar]
  19. Fichtner, A., A.Zunino, and L.Gebraad. 2019. “Hamiltonian Monte Carlo Solution of Tomographic Inverse Problems.” Geophysical Journal International216, no. 2: 1344–1363.
     [Google Scholar]
  20. Figueiredo, L. P., D.Grana, F. L.Bordignon, M.Santos, M.Roisenberg, and B. B.Rodrigues. 2018. “Joint Bayesian Inversion Based on Rock‐ Physics Prior Modeling for the Estimation of Spatially Correlated Reservoir Properties.” Geophysics83, no. 5: M49–M61.
     [Google Scholar]
  21. Glorot, X., A.Bordes, and Y.Bengio. 2011. “Deep Sparse Rectifier Neural Networks.” In Proceedings of the 14th International Conference on Artificial Intelligence and Statistics. PMLR.
  22. Grana, D., T.Fjeldstad, and H.Omre. 2017. “Bayesian Gaussian Mixture Linear Inversion for Geophysical Inverse Problems.” Mathematical Geosciences49: 493–515.
     [Google Scholar]
  23. Grana, D., L.de Figueiredo, and K.Mosegaard. 2022. “Markov Chain Monte Carlo for Petrophysical Inversion.” Geophysics87: M13–M24.
     [Google Scholar]
  24. Izzatullah, M., T.van Leeuwen, and D.Peter. 2021. “Bayesian Seismic Inversion: A Fast Sampling Langevin Dynamics Markov Chain Monte Carlo Method.” Geophysical Journal International227: 1523–1553.
     [Google Scholar]
  25. Jin, Y., X.Wu, J.Chen, and Y.Huang. 2019. “Using a Physics‐Driven Deep Neural Network to Solve Inverse Problems for LWD Azimuthal Resistivity Measurements.” In 60th Annual Logging Symposium. SPWLA.
  26. Li, W.2018. “Classifying Geological Structure Elements From Seismic Images Using Deep Learning.” In SEG Technical Program Expanded Abstracts 2018. Society of Exploration Geophysicists.
  27. Liu, M., D.Vashisth, D.Grana, and T.Mukerji. 2023. “Joint Inversion of Geophysical Data for Gelogic Carbon Sequestration Monitoring: A Differentiable Physics‐Informed Neural Network Model.” Journal of Geophysical Research: Solid Earth128: e2022JB025372.
     [Google Scholar]
  28. Mavko, G., T.Mukerji, and J.Dvorkin. 2010. The Rock Physics Handbook. Cambridge University Press.
     [Google Scholar]
  29. Moench, A. F.1984. “Double‐Porosity Models for a Fissured Groundwater Reservoir With Fracture Skin.” Water Resources Research20, no. 7: 831–846.
     [Google Scholar]
  30. Neal, R. M.2011. MCMC Using Hamiltonian Dynamics. Handbook of Markov Chain Monte Carlo. Chapman & Hall/CRC.
     [Google Scholar]
  31. Pham, N., X.Wu, and E.Zabihi Naeini. 2020. “Missing Well Log Prediction Using Convolutional Long Short‐Term Memory Network.” Geophysics85, no. 4: WA159–WA171.
     [Google Scholar]
  32. Phan, S., and M. K.Sen. 2010. “Porosity Estimation From Seismic Data at Dickman Field, Kansas From Carbon Sequestration.” In SEG Expanded Abstracts. Society of Exploration Geophysicists.
  33. Rall, L. B.1981. Automatic Differentiation: Techniques and Applications. Springer Berlin Heidelberg.
     [Google Scholar]
  34. Roberts, G. O., and R. L.Tweedie. 1996. “Exponential Convergence of Langevin Distributions and Their Discrete Approximations.” Bernoulli2, no. 4: 341–363.
     [Google Scholar]
  35. Sambridge, M., and K.Mosegaard. 2002. “Monte Carlo Methods in Geophysical Inverse Problems.” Reviews of Geophysics40, no. 3: 3_1–3_29. https://doi.org/10.1029/2000RG000089.
     [Google Scholar]
  36. Schütze, H., and C. D.Manning. 1999. Foundations of Statistical Natural Language Processing. MIT Press.
     [Google Scholar]
  37. Talkhestani, A. A.2015. “Prediction of Effective Porosity From Seismic Attributes Using Locally Linear Model Tree Algorithm.” Geophysical Prospecting63, no. 3: 680–693.
     [Google Scholar]
  38. Tarantola, A.2005. Inverse Problem Theory and Methods for Model Parameter Estimation. SIAM.
     [Google Scholar]
  39. Wu, Y., Y.Liu, Z.Zhou, D. C.Bolton, J.Liu, and P.Johnson. 2018. “Deep‐Detect: A Cascaded Region‐Based Densely Connected Network for Seismic Event Detection.” IEEE Transactions on Geoscience and Remote Sensing57: 62–75.
     [Google Scholar]
  40. Yan, Y., and H.Li. 2024. “Simultaneous Inversion of Four Physical Parameters of Hydrate Reservoir for High Accuracy Porosity Estimation.” Geophysical Prospecting72, no. 9: 3202–3216.
     [Google Scholar]
  41. Yu, H., Z.Wang, R.Rezaee, et al. 2018. “Porosity Estimation in Kerogen‐Bearing Shale Gas Reservoirs.” Journal of Natural Gas Science and Engineering52: 575–581.
     [Google Scholar]
  42. Zhou, D., X.Wen, X.He, and Z.He. 2022. “Effective Porosity Seismic Inversion for Porous Media Saturated With an Ideal Fluid Using Simulated Annealing.” Lithosphere12: 1738611.
     [Google Scholar]
  43. Zwartjes, P., and J.Yoo. 2021. “First Break Picking With Deep Learning—Evaluation of Network Architectures.” Geophysical Prospecting70, no. 2: 318–342.
     [Google Scholar]
/content/journals/10.1111/1365-2478.70130
Loading
Efficient Bayesian Active Learning with Langevin Dynamics for Reservoir Porosity Inversion
Geophysical Prospecting 74, e70130 (2026); https://doi.org/10.1111/1365-2478.70130
/content/journals/10.1111/1365-2478.70130
/content/journals/10.1111/1365-2478.70130
Loading

Data & Media loading...

  • Article Type: Research Article
Keyword(s): active learning; Langevin Monte Carlo (LMC) sampling; physics guided; uncertainty analysis

Most Cited This Month Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error