Uncovering the microseismic signals from noisy data for high-fidelity 3D source-location imaging using deep learning

Geophysics ◽  
2021 ◽  
pp. 1-129
Author(s):  
Omar M. Saad ◽  
Min Bai ◽  
Yangkang Chen
Keyword(s):  
Deep Learning ◽  
Signal To Noise Ratio ◽  
Singular Spectrum Analysis ◽  
Microseismic Monitoring ◽  
Source Location ◽  
Reverse Time ◽  
Reverse Time Migration ◽  
Coherent Noise ◽  
Time Migration ◽  
Microseismic Data

Localizing the microseismic event plays a key role in microseismic monitoring. However, microseismic data usually suffer from a low signal-to-noise ratio (S/N), which could affect the resolution of the microseismic source location. We have developed an unsupervised deep learning approach based on variational autoencoder (VAE) and squeeze-and-excitation (SE) networks for enhancing microseismic signals, as well as suppressing noise. First, the microseismic data are divided into several overlapped patches. Second, the VAE encodes the data, extracting the significant features related to the useful signals. Finally, the extracted latent features are decoded to uncover the useful signals and discard the others. The SE network is used to guide the VAE to preserve the useful information related to the clean signal by scaling the extracted features from the encoder part and concatenating them with the features of the decoder part. Our algorithm is evaluated using several synthetic and field examples. As a result, a robust denoising performance is shown despite the existence of a high level of random and coherent noise, for example, with an S/N as low as −32.45 dB. Then, the denoised signal can be used as input data to image the source location using a reverse time migration method, leading to better the location accuracy. Our algorithm performs the best when compared to benchmark methods such as f- x deconvolution and the damped multichannel singular spectrum analysis methods.

Plane-wave least-squares reverse time migration with a preconditioned stochastic conjugate gradient method

Geophysics ◽  
2018 ◽  
Vol 83 (1) ◽  
pp. S33-S46 ◽  
Author(s):  
Chuang Li ◽  
Jianping Huang ◽  
Zhenchun Li ◽  
Rongrong Wang
Keyword(s):  
Plane Wave ◽  
Least Squares ◽  
Conjugate Gradient ◽  
Sampling Method ◽  
Singular Spectrum Analysis ◽  
Reverse Time ◽  
Reverse Time Migration ◽  
Numerical Tests ◽  
Time Migration

This study derives a preconditioned stochastic conjugate gradient (CG) method that combines stochastic optimization with singular spectrum analysis (SSA) denoising to improve the efficiency and image quality of plane-wave least-squares reverse time migration (PLSRTM). This method reduces the computational costs of PLSRTM by applying a controlled group-sampling method to a sufficiently large number of plane-wave sections and accelerates the convergence using a hybrid of stochastic descent (SD) iteration and CG iteration. However, the group sampling also produces aliasing artifacts in the migration results. We use SSA denoising as a preconditioner to remove the artifacts. Moreover, we implement the preconditioning on the take-off angle-domain common-image gathers (CIGs) for better results. We conduct numerical tests using the Marmousi model and Sigsbee2A salt model and compare the results of this method with those of the SD method and the CG method. The results demonstrate that our method efficiently eliminates the artifacts and produces high-quality images and CIGs.

Mini-batch Least-Squares Reverse Time Migration In A Deep Learning Framework

Geophysics ◽  
2020 ◽  
pp. 1-61
Author(s):  
Janaki Vamaraju ◽  
Jeremy Vila ◽  
Mauricio Araya-Polo ◽  
Debanjan Datta ◽  
Mohamed Sidahmed ◽  
...  
Keyword(s):  
Deep Learning ◽  
Least Squares ◽  
Spatial Resolution ◽  
Data Science ◽  
Computational Cost ◽  
Computation Time ◽  
Reverse Time ◽  
Reverse Time Migration ◽  
Learning Framework ◽  
Time Migration

Migration techniques are an integral part of seismic imaging workflows. Least-squares reverse time migration (LSRTM) overcomes some of the shortcomings of conventional migration algorithms by compensating for illumination and removing sampling artifacts to increase spatial resolution. However, the computational cost associated with iterative LSRTM is high and convergence can be slow in complex media. We implement pre-stack LSRTM in a deep learning framework and adopt strategies from the data science domain to accelerate convergence. The proposed hybrid framework leverages the existing physics-based models and machine learning optimizers to achieve better and cheaper solutions. Using a time-domain formulation, we show that mini-batch gradients can reduce the computation cost by using a subset of total shots for each iteration. Mini-batch approach does not only reduce source cross-talk but also is less memory intensive. Combining mini-batch gradients with deep learning optimizers and loss functions can improve the efficiency of LSRTM. Deep learning optimizers such as the adaptive moment estimation are generally well suited for noisy and sparse data. We compare different optimizers and demonstrate their efficacy in mitigating migration artifacts. To accelerate the inversion, we adopt the regularised Huber loss function in conjunction. We apply these techniques to 2D Marmousi and 3D SEG/EAGE salt models and show improvements over conventional LSRTM baselines. The proposed approach achieves higher spatial resolution in less computation time measured by various qualitative and quantitative evaluation metrics.

scholarly journals Deep learning for enhancing multi-source reverse time migration

2022 ◽  
Author(s):  
Yaxing Li ◽  
Xiaofeng Jia ◽  
Xinming Wu ◽  
Zhicheng Geng
Keyword(s):  
Deep Learning ◽  
High Resolution ◽  
Plane Wave ◽  
Computational Efficiency ◽  
Reverse Time ◽  
Reverse Time Migration ◽  
Numerical Examples ◽  
Time Migration ◽  
High Resolution Images ◽  
Wave Migration

<p>Reverse time migration (RTM) is a technique used to obtain high-resolution images of underground reflectors; however, this method is computationally intensive when dealing with large amounts of seismic data. Multi-source RTM can significantly reduce the computational cost by processing multiple shots simultaneously. However, multi-source-based methods frequently result in crosstalk artifacts in the migrated images, causing serious interference in the imaging signals. Plane-wave migration, as a mainstream multi-source method, can yield migrated images with plane waves in different angles by implementing phase encoding of the source and receiver wavefields; however, this method frequently requires a trade-off between computational efficiency and imaging quality. We propose a method based on deep learning for removing crosstalk artifacts and enhancing the image quality of plane-wave migration images. We designed a convolutional neural network that accepts an input of seven plane-wave images at different angles and outputs a clear and enhanced image. We built 505 1024×256 velocity models, and employed each of them using plane-wave migration to produce raw images at 0°, ±20°, ±40°, and ±60° as input of the network. Labels are high-resolution images computed from the corresponding reflectivity models by convolving with a Ricker wavelet. Random sub-images with a size of 512×128 were used for training the network. Numerical examples demonstrated the effectiveness of the trained network in crosstalk removal and imaging enhancement. The proposed method is superior to both the conventional RTM and plane-wave RTM (PWRTM) in imaging resolution. Moreover, the proposed method requires only seven migrations, significantly improving the computational efficiency. In the numerical examples, the processing time required by our method was approximately 1.6% and 10% of that required by RTM and PWRTM, respectively.</p>

scholarly journals An efficient local imaging strategy based on illumination analysis with deep learning

2021 ◽  
Vol 9 ◽  
Author(s):  
Chao Rong ◽  
Xiaofeng Jia
Keyword(s):  
Deep Learning ◽  
Imaging Method ◽  
Target Area ◽  
Single Shot ◽  
Geological Model ◽  
Reverse Time ◽  
Reverse Time Migration ◽  
Energy Propagation ◽  
Time Migration ◽  
Low Illumination

We propose a deep-learning-based illumination analysis and efficient local imaging method. Based on the wavefield forward modeling, seismic illumination can intuitively express the energy propagation of direct waves, reflected waves, and transmitted waves, while it requires high calculation costs. We use a series of convolution operations in deep learning to establish the nonlinear relationship between the model and the illuminations to realize single-shot illumination result of the model. Stacking the single shot illumination results obtained by the network prediction can further help determine the target area. For the target area, we use a deep learning method to obtain the low illumination area of the geological model. Each shot has contribution to the low illuminated area; single shot is selected based on the contribution of the shot being greater than the average illuminance, and the low illumination area is imaged by reverse time migration on the selected shot gather. The trained convolutional neural network can help us quickly obtain the single shot illumination result of the model, which is convenient to analyze the energy distribution of various areas of geological model, and do further imaging for target areas. Using part of the shot gathers to image a local area can recover the complex geological structure of the area and improve the efficiency of reverse time migration especially for 3D problems. This method has universal applicability and is suitable for local imaging of various complex models such as subsalt areas and deep regions.

scholarly journals Deep learning for enhancing multi-source reverse time migration

2022 ◽  
Author(s):  
Yaxing Li ◽  
Xiaofeng Jia ◽  
Xinming Wu ◽  
Zhicheng Geng
Keyword(s):  
Deep Learning ◽  
High Resolution ◽  
Plane Wave ◽  
Computational Efficiency ◽  
Reverse Time ◽  
Reverse Time Migration ◽  
Numerical Examples ◽  
Time Migration ◽  
High Resolution Images ◽  
Wave Migration

<p>Reverse time migration (RTM) is a technique used to obtain high-resolution images of underground reflectors; however, this method is computationally intensive when dealing with large amounts of seismic data. Multi-source RTM can significantly reduce the computational cost by processing multiple shots simultaneously. However, multi-source-based methods frequently result in crosstalk artifacts in the migrated images, causing serious interference in the imaging signals. Plane-wave migration, as a mainstream multi-source method, can yield migrated images with plane waves in different angles by implementing phase encoding of the source and receiver wavefields; however, this method frequently requires a trade-off between computational efficiency and imaging quality. We propose a method based on deep learning for removing crosstalk artifacts and enhancing the image quality of plane-wave migration images. We designed a convolutional neural network that accepts an input of seven plane-wave images at different angles and outputs a clear and enhanced image. We built 505 1024×256 velocity models, and employed each of them using plane-wave migration to produce raw images at 0°, ±20°, ±40°, and ±60° as input of the network. Labels are high-resolution images computed from the corresponding reflectivity models by convolving with a Ricker wavelet. Random sub-images with a size of 512×128 were used for training the network. Numerical examples demonstrated the effectiveness of the trained network in crosstalk removal and imaging enhancement. The proposed method is superior to both the conventional RTM and plane-wave RTM (PWRTM) in imaging resolution. Moreover, the proposed method requires only seven migrations, significantly improving the computational efficiency. In the numerical examples, the processing time required by our method was approximately 1.6% and 10% of that required by RTM and PWRTM, respectively.</p>

Least-squares reverse time migration of multiples in viscoacoustic media

Geophysics ◽  
2020 ◽  
Vol 85 (5) ◽  
pp. S285-S297
Author(s):  
Zhina Li ◽  
Zhenchun Li ◽  
Qingqing Li ◽  
Qingyang Li ◽  
Miaomiao Sun ◽  
...  
Keyword(s):  
Least Squares ◽  
Signal To Noise Ratio ◽  
Real Data ◽  
Velocity Model ◽  
Reverse Time ◽  
Reverse Time Migration ◽  
Time Migration ◽  
Q Model ◽  
Standard Linear Solid ◽  
Standard Linear

The migration of multiples can provide complementary information about the subsurface, but crosstalk artifacts caused by the interference between different-order multiples reduce its reliability. To mitigate the crosstalk artifacts, least-squares reverse time migration (LSRTM) of multiples is suggested by some researchers. Multiples are more affected by attenuation than primaries because of the longer travel path. To avoid incorrect waveform matching during the inversion, we propose to include viscosity in the LSRTM implementation. A method of LSRTM of multiples is introduced based on a viscoacoustic wave equation, which is derived from the generalized standard linear solid model. The merit of the proposed method is that it not only compensates for the amplitude loss and phase change, which cannot be achieved by traditional RTM and LSRTM of multiples, but it also provides more information about the subsurface with fewer crosstalk artifacts by using multiples compared with the viscoacoustic LSRTM of primaries. Tests on sensitivity to the errors in the velocity model, the Q model, and the separated multiples reveal that accurate models and input multiples are vital to the image quality. Numerical tests on synthetic models and real data demonstrate the advantages of our approach in improving the quality of the image in terms of amplitude balancing and signal-to-noise ratio.

Automatic waveform-based source-location imaging using deep learning extracted microseismic signals

Geophysics ◽  
2020 ◽  
Vol 85 (6) ◽  
pp. KS171-KS183 ◽  
Author(s):  
Omar M. Saad ◽  
Yangkang Chen
Keyword(s):  
Clustering Algorithm ◽  
Signal To Noise Ratio ◽  
Reverse Time ◽  
Reverse Time Migration ◽  
Time Frequency ◽  
Time Migration ◽  
Frequency Representation ◽  
Convolutional Autoencoder ◽  
Migration Method

We have used an automatic unsupervised technique to extract waveform signals from continuous microseismic data. First, the time-frequency representation (scalogram) is obtained for the input microseismic trace. Second, the convolutional autoencoder (CAE) is used to extract the significant scalogram features related to the waveform signals and discard the rest. Third, the extracted features from the CAE encoder are considered as the input for the k-means clustering algorithm, in which the input samples are classified into waveform and nonwaveform components. The proposed algorithm is evaluated using several synthetic and field examples. We find that the proposed algorithm successfully extracts the waveform signals even in a noisy environment with a signal-to-noise-ratio as low as −10 dB. We compared the proposed algorithm to benchmark algorithms, for example, simple k-means and short-term and long-term average ratio methods, and find that the proposed algorithm performs best. We find that the detected waveform signals can enhance the resolution of microseismic imaging using a waveform-based reverse time migration method.

Sign in / Sign up

Export Citation Format

Share Document