The final, formatted version of the article will be published soon.
ORIGINAL RESEARCH article
Front. Earth Sci.
Sec. Geoinformatics
Volume 12 - 2024 |
doi: 10.3389/feart.2024.1463723
This article is part of the Research Topic The State-of-Art Techniques of Seismic Imaging for the Deep and Ultra-deep Hydrocarbon Reservoirs - Volume III View all articles
Wavefield reconstruction for full waveform inversion on noisy data
Provisionally accepted- 1 University of Science and Technology of China, Hefei, China
- 2 State Key Laboratory of Shale Oil and Gas Enrichment Mechanisms and Effective Development, Beijing, China
- 3 Sinopec Key Laboratory of Seismic Elastic Wave Technology, Beijing, China
- 4 SINOPEC Petroleum Exploration and Production Research Institute, Beijing, Beijing, China
In practice, full waveform inversion (FWI) can be challenging to use noisy low-frequency data to construct a low-wavenumber initial model. According to the first type of Rayleigh-Sommerfeld integral, we establish a wavefield reconstruction method and use the multiple reconstructed wavefield (MRW) to make FWI robust to noisy data. The MRW is the result of combining different reconstructed wavefields, and the wavefield components within the MRW that have similar properties are enhanced by superposition. The reflected waves in the MRW, which are critical for updating the deep portions of the velocity model, are enhanced. The dot product between the residual noise-generated wavefield and the wavefield that does not contribute to the deep model updates is detrimental to the FWI. Due to the phenomenon of superposition, the enhanced reflected waves in the MRW significantly mitigate this detrimental dot product during FWI updates. We use the MRW to optimize the gradient of the FWI, yielding satisfactory results despite significant noise interference. By incorporating the MRW into the FWI, we effectively mitigate overfitting problems associated with noisy data and improve the robustness of the FWI.
Keywords: Full waveform inversion (FWI), Noisy data, modelling, Rayleigh-Sommerfeld integral, Kirchhoff integral
Received: 12 Jul 2024; Accepted: 02 Dec 2024.
Copyright: © 2024 Yu, Wei, Jia and Zhu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence:
Xiaofeng Jia, University of Science and Technology of China, Hefei, China
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.