Automation of surface wave analysis and inversion method by application of neural networks

Introduction. Surface wave analysis and inversion (SWI) is a valuable tool for constructing near-surface velocity models. It offers high noise immunity and does not require specialized acquisition systems. However, manual extraction of surface wave dispersion characteristics and the need for inverse operator fitting make SWI timeconsuming and impractical for large datasets.

Aim. This study aims to accelerate and automate near-surface velocity model construction using deep learning neural networks within the SWI framework.

Materials and methods. Deep machine learning methods, including a convolutional autoencoder model and a fully-connected neural network, were applied to achieve this goal. The developed algorithms were tested on synthetic seismic data generated using the matrix propagator method and validated using field data from ground seismic surveys.

Results. The proposed neural network architectures demonstrate high accuracy in automatically extracting and inverting surface wave dispersion curves. The mean absolute percentage error for extracted curves was 1 %, and 5 % for reconstructed velocity models on the test dataset. The developed algorithms were successfully applied to real seismic data from an oil and gas field in Western Siberia for automated near-surface model construction.

Conclusions. The set of developed algorithms, based on trained neural networks, off ers a new and effective implementation of the SWI method. It automates and significantly accelerates near-surface model construction through the processing of surface wave data, overcoming the limitations of manual methods and providing a powerful tool for seismic analysis.

1. Yablokov A.V., Serdyukov A.S., Loginov G.N., Baranov V.D. An artifi cial neural network approach for the inversion of surface wave dispersion curves // Geophysical Prospecting. 2021, vol. 69, no. 7, pp. 1405–1432.

2. Cheng F. et al. Watching the cryosphere thaw: Seismic monitoring of permafrost degradation using distributed acoustic sensing during a controlled heating experiment // Geophysical Research Letters. 2022, vol. 49, no. 10, pp. e2021GL097195.

3. Bai J., Yilmaz O. Model-based surface wave analysis and attenuation // 80th EAGE Conference and Exhibition 2018. — European Association of Geoscientists & Engineers. 2018, no. 1, pp. 1–5.

4. Askari R., Ferguson R.J., Isaac J. H., Hejazi, S. H. Estimation of S-wave static corrections using CMP cross-correlation of surface waves // Journal of Applied Geophysics. 2015, vol. 121, pp. 42–53.

5. Socco L.V., Boiero D. Improved Monte Carlo inversion of surface wave data // Geophysical Prospecting. 2008, vol. 56, no. 3, pp. 357–371.

6. Dal Moro G., Pipan M., Gabrielli P. Rayleigh wave dispersion curve inversion via genetic algorithms and marginal posterior probability density estimation // J. Appl. Geophys. 2007, vol. 61, no. 1, pp. 39–55.

7. Pei D., Louie J.N., Pullammanappallil S.K. Application of simulated annealing inversion on high-frequency fundamentalmode Rayleigh wave dispersion curves // Geophysics. 2007, vol. 72, no. 5, pp. R77–R85.

8. Cox B.R., Teague D.P. Layering ratios: a systematic approach to the inversion of surface wave data in the absence of a priori information // Geophysical Journal International. 2016, vol. 207, no. 1, pp. 422–438.

9. Song X., Tang L., Zhao S., Zhang X., Li L., Huang J., Cai W. Grey Wolf Optimizer for parameter estimation in surface waves // Soil Dynamics and Earthquake Engineering. 2015, vol. 75, pp. 147–157.

10. Aleardi M., Stucchi E. A hybrid residual neural network–Monte Carlo approach to invert surface wave dispersion data // Near Surface Geophysics. 2021, no. 4. pp. 397–414.

11. Yablokov A., Lugovtsova Y., Serdyukov A. Uncertainty quantifi cation of multimodal surface wave inversion using artifi cial neural networks // Geophysics. 2023, vol. 88, no. 2, pp. KS1–KS11.

12. Serdyukov A.S., Yablokov A.V., Duchkov A.A., Azarov A.A., Baranov V.D. et al. Slant f-k transform of multichannel seismic surface wave data // Geophysics. 2019, vol. 84, no. 1, pp. A19–A24.

13. Lai C.G., Foti S., Rix G.J. Propagation of data uncertainty in surface wave inversion // Journal of Environmental & Engineering Geophysics. 2005, vol. 10, no. 2, pp. 219–228.

14. Kamashev А.М., Duchkov А.А. Application of machine learning for adaptive subtraction of multiple refl ected waves // Geofi zicheskiye tekhnologii. 2023, no. 1, pp. 54–65.