Seismic Data Processing Research Articles

A deep learning framework for suppressing prestack seismic random noise without noise-free labels

Random noise attenuation is significant in seismic data processing. Supervised deep learning-based denoising methods have been widely developed and applied in recent years. In practice, it is often time-consuming and laborious to obtain noise-free data for supervised learning. Therefore, we propose a novel deep learning framework to denoise prestack seismic data without clean labels, which trains a high-resolution residual neural network (SRResnet) with noisy data for input and the same valid data with different noise for output. Since valid signals in noisy sample pairs are spatially correlated and random noise is spatially independent and unpredictable, the model can learn the features of valid data while suppressing random noise. Noisy data targets are generated by a simple conventional method without fine-tuning parameters. The initial estimates allow signal or noise leakage as the network does not require clean labels. The Monte Carlo strategy is applied to select training patches for increasing valid patches and expanding training datasets. Transfer learning is used to improve the generalization of real data processing. The synthetic and real data tests perform better than the commonly used state-of-the-art denoising methods.

Hypercomplex Processing of Vector Field Seismic Data: Toward vector-valued signal processing [Hypercomplex Signal and Image Processing

Hypercomplex Processing of Vector Field Seismic Data: Toward vector-valued signal processing [Hypercomplex Signal and Image Processing

Seismic data reconstruction based on a multicascade self-guided network

Due to inherent limitations in data acquisition, seismic data reconstruction is an important procedure to recover missing data or improve observation density. Many conventional methods exist to solve the reconstruction task. Reconstruction is challenging, especially in the case of complex seismic data. Recently, convolutional neural networks (CNN) have been applied in seismic data processing. In most cases, the architectures of these CNN-based methods are relatively simple, without sufficient feature interaction, limiting their performance. To improve reconstruction results, a multicascade self-guided network (MSG-Net) is presented. In general, MSG-Net is inspired by the self-guided scheme, and a multicascade architecture is designed to extract informative features within the analyzed seismic data at different resolutions. Following this, a parallel spatial attention module is used to further refine and enhance the primary features, thereby improving reconstruction accuracy. To test and verify the new approach, a training data set is generated, based on the synthetic records obtained by forward modeling methods. Experimental results demonstrate that MSG-Net is a promising approach for performing seismic data interpolation.

Sparse Regularization Based on Orthogonal Tensor Dictionary Learning for Inverse Problems

In seismic data processing, data recovery including reconstruction of the missing trace and removal of noise from the recorded data are the key steps in improving the signal-to-noise ratio (SNR). The reconstruction of seismic data and removal of noise becomes a sparse optimization problem that can be solved by using sparse regularization. Sparse regularization is a key tool in the solution of inverse problems. They are used to introduce prior knowledge and make the approximation of ill-posed inverses feasible. It deals with ill-posedness by replacing an ill-posed inverse problem with a well-posed problem that has a solution close to the true solution. In the last 2 decades, interest has shifted from linear toward nonlinear regularization methods even for linear inverse problems. In inverse problems, regularizations serve as stabilizing the solution of ill-posed inverse problems and give a solution that adequately fits measurements without producing unjustifiably complex artifacts. In this paper, we present a novel sparse regularization based on a tensor-based dictionary method for inverse problems (seismic data interpolation and denoising). This regularization avoids the vectorization step for sparse representation of seismic data during the reconstruction process. The key step in sparsifying signals is the choice of sparsity-promoting dictionary learning. The learning-based approach can adaptively sparsify datasets but has high computational complexity and involves no prior-constraint pattern information for the dataset. Many existing dictionary learning methods would be computationally infeasible for the high dimensional seismic data processing. These methods also destroy the essential information as well as it reduces the discriminability and expressibility of the signal, since they deal with vectorization. In this paper, the orthogonal tensor dictionary learning that learns a dictionary from the input data by employing orthogonality and separability is proposed as sparse regularization for the inverse problems. The performance of the proposed method was validated in seismic data interpolation and denoising individually as well as simultaneously. Numerical examples of synthetic and real seismic datasets demonstrate the validity of the proposed method. The SNR of the recovered data confirms that the proposed method is the most effective method than K-singular value decomposition and orthogonal dictionary learning methods.

Seismic communication data processing based on compressed sensing algorithm

AbstractGeophysical prospecting signals encompass subsurface structural information and incorporate textual messages generated in accordance with a specific pattern. These signals can be employed in places without radio access to ensure public and worker safety. Therefore, the use of seismic signals to transmit information through the earth has attracted the attention of researchers in the last decade. Presently, achievements in seismic communication are mainly in the coding, generation, and propagation of seismic signals. Little work has been done on methods to convert seismic signals generated by vibroseis sources into text and broadcast them vocally. Therefore, we built a seismic communication system with 6‐bit code using the American Standard Code for Information Interchange. To better recommend the seismic communication system scheme, the origin, state and principles of seismic communication system are illustrated in detail. Then, a seismic transmitting system is devised with a 500 N vibroseis source, which compiles seismic signals through amplitude modulation. After seismic signals propagate through the earth, they are received by geophones and recorded in seismographs. Through data acquisition based on a compression algorithm, seismic signals are converted into text and voice signals, which significantly reduces the storage and transmission of seismic data.

Sparse seismic data regularization in shot and trace domains using a residual block autoencoder based on the fast Fourier transform

The increasing use of sparse acquisitions in seismic data acquisition offers advantages in cost and time savings. However, it results in irregularly sampled seismic data, adversely impacting the quality of the final images. In this paper, we develop the residual block fast Fourier transform-convolutional autoencoder (ResFFT-CAE) network, a convolutional neural network with residual blocks based on the Fourier transform. Incorporating residual blocks allows the network to extract high- and low-frequency features from seismic data. The high-frequency features capture detailed information, whereas the low-frequency features integrate the overall data structure, facilitating superior recovery of irregularly sampled seismic data in the trace and shot domains. We evaluate the performance of the ResFFT-CAE network on the synthetic and field data. On synthetic data, we compare the ResFFT-CAE network with the compressive sensing method using the curvelet transform. On field data, we conduct comparisons with other neural networks, such as the CAE and U-Net. The results demonstrate that the ResFFT-CAE network consistently outperforms other approaches in all scenarios. It produces images of superior quality, characterized by lower residuals and reduced distortions. Furthermore, when evaluating model generalization, tests using models trained on synthetic data also exhibit promising results. In conclusion, the ResFFT-CAE network indicates great promise as a highly efficient tool for regularizing irregularly sampled seismic data. Its excellent performance suggests potential applications in the preconditioning of seismic data analysis and processing flows.

Seismic adaptive multiple subtraction using a structure-oriented matched filter

Multiple removal is a crucial step in seismic data processing prior to velocity model building and imaging. After the prediction, adaptive multiple subtraction is used to suppress multiples (considered noise) in seismic data, thereby highlighting primaries (considered signal). In practice, conventional adaptive subtraction methods fit the predicted and recorded multiples in the least-squares sense using a sliding window, formulating a localized adaptive matched filter. Subsequently, the filter is applied to the prediction to remove multiples from the recorded data. However, such a strategy runs the risk of over-attenuating the useful primaries under the minimization energy constraint. To avoid damage to valuable signals, we develop a novel approach that replaces the conventional matched filter with a structure-oriented version. From the predicted multiples, we extract the structural information to be used in the derivation of the adaptive matched filter. Our structure-oriented matched filter emphasizes the structures of predicted multiples, which helps to better preserve primaries during the subtraction. Synthetic and field data examples demonstrate the effectiveness of our structure-oriented adaptive subtraction approach, highlighting its superior performance in multiple removal and primary preservation compared with conventional methods on 2D regularly sampled data.

Modeling and numerical simulation of a type of pure-viscoacoustic-wave equation in attenuated transversely isotropic media

The anisotropy and attenuation features of subsurface media significantly affect seismic data processing. Ignoring anisotropy and attenuation in seismic wave propagation may result in inaccurate reflector positions, dimming amplitudes, and reduced spatial resolution in the imaging results. Therefore, accurate seismic wave modeling of anisotropy and attenuation is essential for understanding wave propagation in the earth’s interior. This paper derives three pure-viscoacoustic-wave equations from characterizing the earth’s frequency-independent Q behavior in transversely isotropic (TI) media. First, we develop three time-space domain pure-qP-wave equations in TI media based on different approximation methods, whose coefficients can be determined by different approximation methods and Thomsen’s anisotropic parameters [Formula: see text], [Formula: see text]. Subsequently, we introduce the Kelvin-Voigt attenuation model into our derived three time-space domain pure-qP-wave equations and then obtain three pure-viscoacoustic-wave equations. To further demonstrate the effectiveness and accuracy of our methods, we provide some 2D and 3D numerical tests. The numerical results indicate that the wavefield generated by pure-qP-wave equations and pure-viscoacoustic-wave equations have accurate kinematic characteristics of qP-wave in TI media and attenuated TI media and are free of S-wave artifacts while remaining stable under Thomsen’s anisotropic parameters [Formula: see text], so our methods have broader applicability compared with some existing methods. At the same time, the simulation results of pure-viscoacoustic-wave equations also reflect the absorption and attenuation characteristics of qP-waves in attenuated TI media.

Deep-Towed Array Geometry Inversion Based on an Improved Particle Swarm Optimization Algorithm

When marine deep-towed multichannel seismic data are processed, the description of the receiving array geometry significantly impacts the quality of the imaging profile. Therefore, achieving a highly precise description of the receiving array geometry is very important for the fine imaging of such data. While basic particle swarm optimization (PSO) is known for its ease of implementation and efficiency, it often exhibits a low convergence accuracy. Consequently, the PSO algorithm is improved by modifying the inertia weight and incorporating Gaussian mutation. In combination with the actual motion of the towing streamer during surveys, a strategy for inheriting particle positions is introduced. When each seismic shot is solved sequentially, the results from the previous shot can serve as the initial particle positions for the next shot. The results indicate that this strategy achieves superior fitness values and outperforms the basic PSO algorithm. This method exhibits simplicity, rapid optimization, and a favorable solution quality, thereby offering a valuable approach to deep-towed array geometry inversion. It enhances the efficiency of deep-towed seismic data processing and serves as a reference for similar applications.

Study Investigates Seismic Monitoring for Carbon Storage Leak Detection

_ This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper IPTC 22980, “Carbon Storage Leak Detection Through Seismic FWI and RTM: Different Survey Analyses,” by Sajjad Amani, Kyoto University. The paper has not been peer reviewed. Copyright 2023 International Petroleum Technology Conference. Reproduced by permission. _ In the complete paper, marine seismic data processing is investigated as a tool for monitoring possible leakages in geological carbon storage. Because of the great importance of storage permanence, a precise leakage-monitoring strategy is crucial. The proficiency of seismic monitoring solutions for leakage monitoring can be affected by shallower layers as a result of structure, seismic wave attenuation, and leak size. The authors explore two popular seismic monitoring methods used in this application in different scenarios: full waveform inversion (FWI) and reverse-time migration (RTM). Introduction Among the various carbon capture and storage (CCS) options, underground storage in saline aquifers is the best-understood solution. To assure storage consistency and permanence, finding the best strategy to precisely detect possible carbon leaks is essential. A perfect method must demonstrate the difference between stored CO2 and injected CO2 to detect potential fast- and slow-leakage areas. The seismic monitoring technique is the most efficient approach in this respect. Two popular tools for seismic monitoring are FWI and RTM. Several studies of their use in this application have been conducted. However, previous research did not analyze different survey arrays for carbon-leak detection using FWI and RTM. The current research aims to investigate quantitative aspects of CCS monitoring to conduct sensitivity analysis of the three different survey arrays [vertical seismic profile (VSP), crosswell, and surface] for different amounts of CO2-storage leakage in a saline aquifer reservoir. The capability of seismic-imaging methods for small amounts of leakage was tested. Comparison of these three arrays using monitoring methods such as RTM and FWI reveals their pros and cons in providing detailed information about the reservoir. In this research, a simple synthetic model was built that closely fits actual reservoirs characterized by suitable physical features such as velocity and density. Subsequently, elastic wave propagation simulation was implemented by use of a finite-difference scheme over a physical model of the reservoir. Then, FWI was applied to enhance the accuracy of the model parameters. With an efficient forward-modeling and inversion scheme, RTM was used as a powerful imaging tool that provides final high-resolution results for monitoring CO2 migration and possible leakage. This process examined the synthetic model for different amounts of CO2 leakage in saline aquifers to evaluate the performance of CO2 leakage monitoring using the seismic method. The results of the different receiver and source arrays also were compared to establish their effectiveness.

Seismic resolution enhancement with variational modal-based fast-matching pursuit decomposition

Enhancing vertical resolution and signal-to-noise ratio (S/N) is a key objective in seismic data processing. Considering the underground medium is inhomogeneous and incompletely elastic, seismic wave energy attenuation occurs during underground propagation, which has a significant impact on seismic data resolution and S/N. Traditional fast-matching pursuit (FMP) algorithms make it difficult to separate valid signals and noise effectively while reconstructing the noisy signals. Therefore, an improved FMP algorithm that combines the variational-mode decomposition (VMD) strategy is developed. The VMD algorithm is used to obtain intrinsic mode functions with varying amplitudes, frequencies, and center times. It can achieve a multiscale decomposition of nonstationary seismic data. Based on the intrinsic mode functions of different scales, the FMP algorithm can reconstruct prior information of the amplitude, frequency, and center time of valid signals and noise signals in the mode functions. Thus, the high-resolution sparse representation of intrinsic mode functions is achieved. The numerical results indicate that our method not only separates the effective signal and noise but also preserves the valid signal as much as possible. In addition, the feasibility of the method is further verified by field exploration data. The results indicate that this strategy can enhance the resolution of seismic data while restoring the attenuated energy using multiscale seismic data.

Multi‐model stacked structure based on particle swarm optimization for random noise attenuation of seismic data

AbstractRandom noise attenuation is a fundamental task in seismic data processing aimed at improving the signal‐to‐noise ratio of seismic data, thereby improving the efficiency and accuracy of subsequent seismic data processing and interpretation. To this end, model‐based and data‐driven seismic data denoising methods have been widely applied, including f–x deconvolution, K‐singular value decomposition, feed‐forward denoising convolutional neural network and U‐Net (an improved fully convolutional neural network structure), which have received widespread attention for their effectiveness in attenuating random noise. However, they often struggle with low‐signal‐to‐noise ratio data and complex noise environments, leading to poor performance and leakage of effective signals. To address these issues, we propose a novel approach for random noise attenuation. This approach employs a multi‐model stacking structure, where the parameters governing this structure are optimized using a particle swarm optimizer. In the model‐based denoising method, we choose the f–x deconvolution method, whereas in the data‐driven denoising method, we choose K‐singular value decomposition for shallow learning and U‐Net for deep learning as components of the multi‐model stacking structure. The optimal parameters for the multi‐model stacking structure are obtained using a particle swarm optimizer, guided by the proposed novel hybrid fitness function incorporating weighted signal‐to‐noise ratio, structural similarity and correlation parameters. Finally, the effectiveness of the proposed method is verified with three synthetic and two real seismic datasets. The results demonstrate that the proposed method is effective in attenuating random noise and outperforms the benchmark methods in denoising both synthetic and real seismic data.

Seismic survey of steeply inclined coal seams using the vertical data acquisition system

Prior to the establishment of a coal mine, a high-resolution three-dimensional (3D) seismic survey was carried out. For seismic exploration of steep coal seams with an inclination angle more than 45°, the selection of the data gathering system is essential. The time disparity that remains when the traditional acquisition technology acquires 3D seismic data will be rather significant. The relationship between coal seam inclination, remaining time difference, and reflection coefficient is investigated in light of the presence of steeply inclined coal seams. When the direction of the generation surveying line is parallel to the coal seam, it is suggested as a vertical data acquisition system. To create a vertical data gathering system, the point is situated on the down-dipping side of the coal seam and a predetermined distance is kept between shooting points and detection points. A vertical data collection system was used to implement 3D seismic data acquisition in the Muli Coalfield in northern Qinghai. As a solid foundation for the future processing and analysis of 3D seismic data, valuable observation data with high resolution were successfully obtained. The study came to the conclusion that a vertical data collecting system is helpful in promoting effective coal mine output.

Geology-constrained dynamic graph convolutional networks for seismic facies classification

Knowing a land’s facies type before drilling is an essential step in oil exploration. In seismic surveying, subsurface images are analyzed to segment and classify the facies in that volume. With the recent developments in deep learning, multiple works have utilized deep neural networks to classify facies from subsurface images. Unlike natural images, seismic data have different patterns and structures, which means that although general deep learning architectures can work with seismic data, it would be more effective if these architectures were optimized and refined specifically for such types of data. Most of the works in the seismic domain focus on convolution neural networks as the main backbone for the architectures, and more recently transformers started becoming more common in seismic data processing. Proposing a different approach that can capture unique correlations in the data, we introduce the use of dynamic graph convolutional networks as a method for capturing long-term dependencies for seismic facies classification. The proposed architecture combines the use of convolution neural networks and graph convolution networks to capture both global and local structures of the data. The performance of the model was evaluated on a facies classification dataset, and the proposed method provided state-of-the-art results while significantly reducing the number of parameters in the model compared to other architectures. Code is available at https://github.com/swaidan/Geology-Restricted-DGCNN.

Generalization of Snell's Law for the propagation of acoustic waves in elliptically anisotropic media

<abstract> <p>In seismic data processing, both in inversion (Inverse Processing) and modeling (Direct Processing), it is essential to consider anisotropy to unravel the geological structure of the subsoil. Besides, in most cases, the macroscopic model of anisotropy in 2D seismic surveys is elliptical and weak, with ratios of anisotropy close to one. Therefore, it is crucial to have at disposal the analytical formulas for acoustic wave propagation in elliptical anisotropic media. We presented the generalization of the Snell's Law for the case of acoustic wave propagation in elliptically anisotropic media. The generalization of the Snell's Law for acoustic anisotropic media had different applications in digital processing, raytracing, and acoustic inversion to properly consider elliptical anisotropy.</p> </abstract>

A decade of technology advancement in seismic processing: A case study from reprocessing legacy sparse OBC data in Kashagan Field

Ceaseless advancements in seismic data processing technology and workflow enable the extraction of increasing amounts of subsurface information essential for derisking and optimizing natural resource development. The Kashagan oil field, one of the world's largest carbonate reservoirs, is faced with significant development optimization challenges due to a combination of complex geology and suboptimal seismic data coverage. The latest processing technologies were applied to the 2001–2002 vintage sparse ocean-bottom cable data and produced step-change improvements over the 2010–2011 legacy processing and imaging results. Such significant image uplift is representative of technology advancements over the span of a decade fully leveraging the newly preprocessed input data, including (1) an integrated model building workflow, (2) adaptive application of full-waveform inversion, (3) geologically constrained reflection tomography, and (4) least-squares imaging. The new seismic results improved the structural and stratigraphic imaging of multiple layers of shallow carbonates, mitigated fault shadow and other complex overburden effects, increased the resolution of subsalt carbonate reservoir features such as karst and internal fractures, improved well marker depth misfit, and ultimately influenced the placement of upcoming wells and reservoir development plans. Further improvements in seismic imaging would be feasible with a modern acquisition of more densely sampled seismic data, which would allow the full potential of the latest seismic processing technologies and workflows to be unlocked.

Using 3D Seismic Technology to Accurately Identify Faults and Reservoir Characteristics

With the accelerating development of the Chinese economy, the demand for oil and gas energy is increasing day by day. The accuracy and efficiency of fault interpretation are crucial for the exploration and development of oil and gas reservoirs. The presence of oil and gas in reservoirs will inevitably cause changes in the geophysical response characteristics, and different types of reservoirs have different response characteristics in logging curves such as sound, discharge, and electricity. 3D seismic technology is currently one of the most effective techniques for obtaining fault structural features and identifying small faults, but conventional seismic data processing techniques are unable to meet the accuracy requirements of current seismic exploration. With the development of computer technology, more and more scholars are applying deep learning (DL) algorithms to the identification of faults and reservoirs to further improve the recognition effect. This article is based on the DL algorithm and utilizes 3D seismic technology to design a method for automatic and accurate recognition of fault and reservoir features. Obtain labeled seismic amplitude faults, reservoir data, and seismic amplitude data to be identified from the 3D seismic amplitude data volume, and construct a CNN training model to identify the fault and reservoir data. The experimental results show that the method designed in this article can improve the efficiency and accuracy of fault and reservoir identification.

Generative interpolation via a diffusion probabilistic model

Seismic data interpolation is essential in a seismic data processing workflow, recovering data from sparse sampling. Traditional and deep-learning-based methods have been widely used in the seismic data interpolation field and have achieved remarkable results. In this paper, we develop a seismic data interpolation method through the novel application of diffusion probabilistic models (DPMs). DPM transforms the complex end-to-end mapping problem into a progressive denoising problem, enhancing the ability to reconstruct complex situations of missing data, such as large proportions and large-gap missing data. The interpolation process begins with a standard Gaussian distribution and seismic data with missing traces and then removes noise iteratively with a UNet trained for different noise levels. Our DPM-based interpolation method allows interpolation for various missing cases, including regularly missing, irregularly missing, consecutively missing, noisy missing, and different ratios of missing cases. The ability to generalize to different seismic data sets is also discussed in this paper. Numerical results of synthetic and field data indicate satisfactory interpolation performance of the DPM-based interpolation method in comparison with the f- x prediction filtering method, the curvelet transform method, the low-dimensional manifold method (LDMM), and the coordinate attention-based UNet method, particularly in cases with large proportions and large-gap missing data. Diffusion is all we need for seismic data interpolation.

Intelligent reconstruction for spatially irregular seismic data by combining compressed sensing with deep learning

Data reconstruction is the most essential step in seismic data processing. Although the compressed sensing (CS) theory breaks through the Nyquist sampling theorem, we previously proved that the CS-based reconstruction of spatially irregular seismic data could not fully meet the theoretical requirements, resulting in low reconstruction accuracy. Although deep learning (DL) has great potential in mining features from data and accelerating the process, it faces challenges in earth science such as limited labels and poor generalizability. To improve the generalizability of deep neural network (DNN) in reconstructing seismic data in the actual situation of limited labeling, this paper proposes a method called CSDNN that combines model-driven CS and data-driven DNN to reconstruct the spatially irregular seismic data. By physically constraining neural networks, this method increases the generalizability of the network and improves the insufficient reconstruction caused by the inability to sample randomly in the whole data definition domain. Experiments on the synthetic and field seismic data show that the CSDNN reconstruction method achieves better performance compared with the conventional CS method and DNN method, including those with low sampling rates, which verifies the feasibility, effectiveness and generalizability of this approach.

Characteristic analysis and data comparative of linear and nonlinear low-frequency sweep in vibroseis

Abstract Low-frequency seismic data plays a crucial role in seismic data processing and seismic wave inversion. At present, there are two methods to realize the low-frequency excitation of vibrators: one is exciting low-frequency vibrators by linear sweep signals, and the other is exciting conventional vibrators by nonlinear low-frequency sweep signals. The cost of exploration using low-frequency vibroseis is high, and it is challenging to obtain sufficient low-frequency information using traditional vibrators. To this end, this paper comparatively studies the low-frequency sweep signal characteristics and data effects of low-frequency and traditional vibrators. Therefore, three kinds of linear and nonlinear low-frequency sweep signal are designed. Theoretical analysis shows that there are certain differences between linear and nonlinear signal in design methods, signal shapes, etc., but after correlation calculations the signal spectra reflecting the seismic response and the related wavelet shapes are basically consistent. Besides, the actual force signal data shows that the linear and nonlinear harmonic distortion are basically equivalent. Finally, based on the forward simulation of three sweep signals and the comparative analysis of field test data, it can be considered that the linear and nonlinear low-frequency sweep signals of vibrators have almost the same denoising ability under the basic conditions of the spectrum and wavelet. Both can achieve low-frequency excitation and obtain rich low-frequency information, and the quality of seismic data is basically the same, so they can be applied in practical production.