DOI: https://doi.org/10.26089/NumMet.v26r322

A structural two-level neural network approach to joint inversion of gravitational and magnetic fields

Authors

  • Yanfei Wang
  • Dmitry V. Churbanov
  • Raul L. Argun
  • Alexander V. Gorbachev
  • Alexander S. Leonov
  • Dmitry V. Lukyanenko

Keywords:

inverse problem
joint processing of gravitational and magnetic fields
neural networks
structural residual

Abstract

The paper proposes a new two-level neural network approach to joint inversion of gravitational and magnetic fields in the problem of mineral exploration. The lower level uses two neural networks with equivalent architecture. The first network calculates the distribution of gravitational field sources in a given domain based on measurements of this field on a certain remote surface. The second one finds the distribution of magnetic field sources in the same area based on magnetic field measurements. At the upper level, a structural residual between the two obtained source distributions is calculated, which determines the “degree of difference” of their geometries. The resulting neural network is trained to minimize this structural discrepancy. As a result, this two-level system implements a computational algorithm that finds geometrically close sources of fields of different nature. Test simulations demonstrate the high quality of joint inversion of model gravitational and magnetic fields within the proposed aproach.


Downloads

Published

2025-09-15

Issue

Vol. 26 (2025): Issue 3.

Section

Methods and algorithms of computational mathematics and their applications

Authors

Yanfei Wang

Institute of Geology and Geophysics, Chinese Academy of Sciences
• Professor

Dmitry V. Churbanov

Moscow Institute of Physics and Technology,
Department of Higher Mathematics
• Senior Researcher

Raul L. Argun

Lomonosov Moscow State University,
Faculty of Physics,
Department of Mathematics
Leninskie Gory, 1, building 2, 119991, Moscow
• Specialist

Alexander V. Gorbachev

Lomonosov Moscow State University,
Faculty of Physics,
Department of Mathematics
Leninskie Gory, 1, building 2, 119991, Moscow
• Phd Student

Alexander S. Leonov

National Research Nuclear University MEPhI,
Department of Higher Mathematics
• Professor

Dmitry V. Lukyanenko

Lomonosov Moscow State University,
Faculty of Physics,
Department of Mathematics
• Professor

References

  1. Vl. V. Voevodin, A. S. Antonov, D. A. Nikitenko, et al., “Supercomputer Lomonosov-2: Large Scale, Deep Monitoring and Fine Analytics for the User Community,” Supercomput. Front. Innov. 6 (2), 4-11 (2019).
    doi 10.14529/jsfi190201
  2. P. S. Novikov, “On the Uniqueness of the Solution to the Inverse Potential Problem,” Dokl. Akad. Nauk SSSR 18 (3), 165-168 (1938).
  3. L. N. Sretensky, “On the Uniqueness of Determining the Shape of an Attracting Body from the Values of Its External Potential,” Dokl. Akad. Nauk SSSR 99 (1), 21-22 (1954).
  4. A. N. Tikhonov and V. Y. Arsenin, Methods for Solving Ill-Posed Problems(Nauka, Moscow, 1979) [in Russian].
  5. M. S. Zhdanov, Analogs of the Cauchy Integral in the Theory of Geophysical Fields(Nauka, Moscow, 1984) [in Russian].
  6. K. Simirdanis, F. Simon, D. Oikonomou, and N. Papadoupoulos, “A Short Literature on Joint Inversion Methods in Geophysics,” (Research Report, IMS FORTH, Inrap (DST), 2019).
  7. A. N. Tikhonov, A. S. Leonov, and A. G. Yagola, Nonlinear Ill-Posed Problems(KURS, Moscow, 2017) [in Russian].
  8. A. S. Leonov, Solution of Ill-Posed Inverse Problems: Outline of Theory, Practical Algorithms, and MATLAB Demonstrations(Librokom, Moscow, 2010) [in Russian].
  9. Y. Wang, D. Lukyanenko, and A. Yagola, “Magnetic Parameters Inversion Method with Full Tensor Gradient Data,” Inverse Probl. Imaging 13 (4), 745-754 (2019).
    doi 10.3934/ipi.2019034
  10. E. N. Akimova, V. E. Misilov, and A. I. Tretyakov, “Methods and Algorithms of Computational Mathematics and Their Applications,” Numerical Methods and Programming (Vychislitel’nye Metody i Programmirovanie) 24 (4), 368-385 (2023).
    doi 10.26089/NumMet.v24r426
  11. R. Huang, Y. Zhang, S. Vatankhah, et al., “Inversion of Large-Scale Gravity Data with Application of VNet,” Geophys. J. Int. 231 (1), 306-318 (2022).
    doi 10.1093/gji/ggac190
  12. S. Yu and J. Ma, “Deep Learning for Geophysics: Current and Future Trends,” Rev. Geophys. 59 (3), Atticle ID. e2021RG000742 (2021).
    doi 10.1029/2021RG000742
  13. V. N. Strakhov, “On Equivalence in the Inverse Problem of Gravimetry with Variable Mass Density,” Dokl. Akad. Nauk SSSR 236 (2), 329-331 (1977).
  14. A. B. Bakushinsky and A. V. Goncharsky, Iterative Methods for Solving Ill-Posed Problems(Nauka, Moscow, 1989) [in Russian].
  15. V. K. Ivanov, V. V. Vasin, and V. P. Tanana, “Theory of Linear Ill-Posed Problems and Its Applications,” (Nauka, Moscow, 1978) [in Russian].
  16. R. Huang, S. Liu, R. Qi, and Y. Zhang, “Deep Learning 3D Sparse Inversion of Gravity Data,” J. Geophys. Res.: Solid Earth 126 (11), Article Number e2021JB022476 (2021).
    doi 10.1029/2021JB022476
  17. S.-B. Qiao, H.-P. Li, R. Qi, et al., “3D Gravity Inversion Using Cycle-Consistent Generative Adversarial Network,” Appl. Geophys. (2024).
    doi 10.1007/s11770-024-1096-5
  18. M. Jessell, J. Guo, Y. Li, et al., “Into the Noddyverse: a Massive Data Store of 3D Geological Models for Machine Learning and Inversion Applications,” Earth System Science Data 14 (1), 381-392 (2022).
    doi 10.5194/essd-14-381-2022
  19. E. Haber and D. Oldenburg, “Joint Inversion: a Structural Approach,” Inverse Probl. 13 (1) (1997).
    doi 10.1088/0266-5611/13/1/006
  20. L. A. Gallardo and M. A. Meju, “Characterization of Heterogeneous Near-Surface Materials by Joint 2D Inversion of dc Resistivity and Seismic Data,” Geophys. Res. Lett. 30 (13) (2003).
    doi 10.1029/2003GL017370
  21. M. Bosch and J. McCaughey, “Joint Inversion of Gravity and Magnetic Data under Lithologic Constraints,” The Leading Edge 20 (8), 877-881 (2001).
    doi 10.1190/1.1487299
  22. L. R. Dice, “Measures of the Amount of Ecologic Association between Species,” Ecology 26 (3), 297-302 (1945).
  23. Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner, “Gradient-Based Learning Applied to Document Recognition,” Proceedings of the IEEE 86 (11), 2278-2324 (1998).
    doi 10.1109/5.726791
  24. A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet Classification with Deep Convolutional Neural Networks,” Advances in Neural Information Processing Systems 25 (NIPS 2012).
  25. Y. Lin and Y. Wu, “InversionNet: A Real-Time and Accurate Full Waveform Inversion with Convolutional Neural Network,” J. Acoust. Soc. Am. 144 (3), 1683-1683 (2018).
    doi 10.1121/1.5067485
  26. K. Noh, D. Yoon, and J. Byun, “Imaging Subsurface Resistivity Structure from Airborne Electromagnetic Induction Data Using Deep Neural Network,” Explor. Geophys. 51 (2), 214-220 (2019).
    doi 10.1080/08123985.2019.1668240
  27. M. I. Shimelevich, E. A. Rodionov, I. E. Obornev, and E. A. Obornev, “Application of Convolutional Neural Networks in Inverse Problems of Geoelectrics,” Izvestiya, Physics of the Solid Earth 60 (6), 1215-1227 (2024).
    doi 10.1134/S1069351324701039
  28. S. He, H. Cai, S. Liu, et al., “Recovering 3D Basement Relief Using Gravity Data through Convolutional Neural Networks,” J. Geophys. Res.: Solid Earth 126 (10) (2021).
    doi 10.1029/2021JB022611
  29. F. Zhao, Y. Xu, N. Zheng, et al., “ARPENN: an Improved Deep Convolutional Neural Network for Bathymetry Inversion with Integrated Physical Constraints,” Geophys. J. Int. 241 (2), 891-900 (2025).
    doi 10.1093/gji/ggaf081
  30. O. Ronneberger, P. Fischer, and T. Brox, “U-Net: Convolutional Networks for Biomedical Image Segmentation,” Med. Image Comput. Comput. Assist. Interv. –- MICCAI 2015. 234-241 (2015).
  31. Y.-F. Wang, Y.-J. Zhang, L.-H. Fu, and H.-W. Li, “Three-Dimensional Gravity Inversion Based on 3D U-Net++,” Appl. Geophys. 18 (4), 451-460 (2021).
    doi 10.1007/s11770-021-0909-z
  32. H. Huang, L. Lin, R. Tong, et al., “UNet 3+: A Full-Scale Connected UNet for Medical Image Segmentation,” IEEE Int. Conf. on Acoustics, Speech and Signal Processing (ICASSP). 1055-1059 (2020).
  33. A. G. Yagola, Y. Wan, I. E. Stepanova, and V. N. Titarenko, Inverse Problems and Methods for Their Solution: Applications in Geophysics(Binom, Laboratoriya Znaniy, Moscow, 2017) [in Russian].
  34. G. Cybenko, “Approximation by Superpositions of a Sigmoidal Function,” Math. Control, Signals, and Systems (MCSS) 2 (4), 303-314 (1989).
    doi 10.1007/BF02551274
  35. A. Zafar, M. Aamir, N. Mohd Nawi, et al., “A Comparison of Pooling Methods for Convolutional Neural Networks,” Appl. Sci. 12 (17) (2022).
    doi 10.3390/app12178643
  36. A. Zhang, Z. Lipton, M. Li, and A. J. Smola, “Dive into Deep Learning,” CoRR. (2021). [arXiv: 2106.11342]. Cited September 4, 2025.
  37. I. Loshchilov and F. Hutter, “Decoupled Weight Decay Regularization,” Int. Conf. on Learning Representations (2019).
    https://openreview.net/forum?id=Bkg6RiCqY7 . Cited September 4, 2025.
  38. Z. Bai, Y. Wang, C. Wang, et al., “Joint Gravity and Magnetic Inversion Using CNNs’ Deep Learning,” Remote Sensing 16 (7) (2024).
    doi 10.3390/rs16071115
  39. R. F. Bulter, Paleomagnetism: Magnetic Domains to Geologic Terranes (Blackwell Scientific, Boston, 1992).

License

Copyright (c) 2025 Янфей Ван, Д. В. Чурбанов, Р. Л. Аргун, А. В. Горбачев, А. С. Леонов, Д. В. Лукьяненко

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.