Deep-Learning Inversion of Seismic Data
Инверсия сейсмических данных с использованием глубокого обучения
2019-12-12
SCID: 54.1/xmkebzat
Discuss with AI
SeisInvNetdeep neural networksseismic explorationseismic inversionvelocity model reconstruction
Figures from the paper
Abstract (AI)
We propose a new method to tackle the mapping challenge from time-series data to spatial image in the field of seismic exploration, i.e., reconstructing the velocity model directly from seismic data by deep neural networks (DNNs). The conventional way of addressing this ill-posed inversion problem is through iterative algorithms, which suffer from poor nonlinear mapping and strong nonuniqueness. Other attempts may either import human intervention errors or underuse seismic data. The challenge for DNNs mainly lies in the weak spatial correspondence, the uncertain reflection-reception relationship between seismic data and velocity model, as well as the time-varying property of seismic data. To tackle these challenges, we propose end-to-end seismic inversion networks (SeisInvNets) with novel components to make the best use of all seismic data. Specifically, we start with every seismic trace and enhance it with its neighborhood information, its observation setup, and the global context of its corresponding seismic profile. From the enhanced seismic traces, the spatially aligned feature maps can be learned and further concatenated to reconstruct a velocity model. In general, we let every seismic trace contribute to the reconstruction of the whole velocity model by finding spatial correspondence. The proposed SeisInvNet consistently produces improvements over the baselines and achieves promising performance on our synthesized and proposed SeisInv data set according to various evaluation metrics. The inversion results are more consistent with the target from the aspects of velocity values, subsurface structures, and geological interfaces. Moreover, the mechanism and the generalization of the proposed method are discussed and verified. Nevertheless, the generalization of deep-learning-based inversion methods on real data is still challenging and considering physics may be one potential solution.
Key Findings
1
SeisInvNet consistently outperforms baseline methods on synthesized data and the proposed SeisInv dataset across multiple evaluation metrics.
2
SeisInvNets enhance each seismic trace with neighborhood information, observation setup, and global profile context to address weak spatial correspondence.
3
The method learns spatially aligned feature maps and enables every seismic trace to contribute to reconstruction of the entire velocity model.
4
The paper introduces end-to-end SeisInvNets for directly reconstructing spatial velocity models from seismic time-series data using deep neural networks.
5
The reconstructed models better match targets in velocity values, subsurface structures, and geological interfaces, while generalization to real data remains challenging.
Research Object
Seismic data and the corresponding subsurface velocity model
Research Subject
Deep-learning-based reconstruction of the subsurface velocity model from seismic data, including spatial correspondence and inversion accuracy
Publication Details
Publication Date
2019-12-12
Journal
Publisher
ISSN
Cited by
426
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest