The heterogeneity of subsurface media induces multipath scattering and dielectric loss in Ground Penetrating Radar (GPR) signal propagation, which results in wavefront distortion and signal attenuation. These effects degrade B-scan profiles by blurring target signatures, hindering automated feature extraction, and reducing the clarity of regions of interest (ROI). To address these issues, we propose the Adaptive Region Target Enhancement Algorithm (ARTEA), a multi-stage preprocessing framework. ARTEA integrates dynamic range compression, continuous-scale normalization guided by adaptive sigma maps, and a frequency-domain refinement step. By dynamically adjusting parameters according to local signal characteristics, ARTEA is designed to achieve an effective trade-off between artifact suppression and target preservation. Experiments on both synthetic and field GPR data demonstrate that ARTEA can enhance target contrast and structural fidelity while suppressing artifacts and preserving essential target features.

Data aggregation is essential in near-ground long chain sensing networks, as it ensures data freshness and stability while extending communication range through path planning and traffic scheduling. However, the dynamic and fragile nature of long-chain networks under near-surface interference—particularly the presence of time-varying network links—poses significant challenges, for which no effective aggregation solution currently exists. This paper focuses on a representative ground-based distributed seismic node long-chain network and introduces a Weight Agnostic Structure-Free (WASF) data aggregation method based on artificial neural networks. WASF regulates packet forwarding paths and waiting times via activation functions and employs shared connection weights to identify the globally optimal aggregation node, thereby enabling efficient data aggregation. Simulation results demonstrate that, compared with state-of-the-art aggregation methods, WASF reduces end-to-end latency and packet loss rate by 11.8% and 9.4%, respectively. Field deployment on GEIWSR-III seismic nodes further confirmed its effectiveness, yielding 84.3% lower energy consumption, 41% reduced latency, and 20.2% fewer packet losses. By enabling structure-free aggregation, WASF provides a reliable reference framework for efficient, robust, and scalable communication in long-chain IoT networks, such as tunnels, pipe galleries, rivers, and railways.

The quality of seismic images is often affected by the limitation of acquisition conditions and the interference of noises, which causes the low resolution of seismic images and misleads the following geological interpretation. Although the super-resolution method for seismic images based on convolutional neural network (CNN) has behaved well, the quality of weak events especially deep events is still need to be improved, due to CNN is limited by the receptive fields, which results in weaker ability to perceive relationships among pixels far apart. In this letter, we solve this problem by designing a combination network of CNN and transformer (CNCT). CNCT consists of three parts, edge feature fusion block (EFB), deep feature mining block (DMB), and feature enhancement block (FEB). The EFB aims to fuse the input low-resolution (LR) image and the corresponding edges obtained by the Sobel algorithm and performs preliminary shallow feature extraction. DMB mines deeper features by stacking residual blocks, and each residual block makes full use of its excellent perception of global and local information by combining transformer and CNN. Finally, the FEB uses subpixel convolution for upsampling to expand the size of feature maps. The experimental results on synthetic data and field data show that CNCT not only behaves better on perception effect and texture details than that of other deep learning (DL) methods but also can suppress noise and improve the dominant frequency.

The seismic data reconstruction techniques are primarily used to address data missing or damaged due to human or environmental factors under restricted acquisition conditions and thus enhance the accuracy of obtained stratigraphic information. Hence, seismic data reconstruction stands as a crucial preprocessing step and is necessary for the effective exploration of subsurface resources. The existing reconstruction methods often fall short of fully utilizing the information existed in seismic data, thereby impacting reconstruction accuracy of effective signals. To overcome the aforementioned limitation, we propose a hierarchical guided multicascade feedback network (HGMFN), which facilitates comprehensive interaction of seismic data across different resolutions by learning intricate features from clean and complete seismic data at various scales. The proposed network achieves progressive integration of features along with layer-by-layer guidance and multilevel feedback mechanisms, accomplishing the reconstruction task and improving the processing precision. Experiments conduct with both synthetic and field data have confirmed the accuracy and performance of HGMFN in complex seismic data reconstruction.

Distributed optical fiber acoustic sensing (DAS) is a rapidly developed seismic acquisition technology with the advantages of low cost, high resolution, high sensitivity, small interval, etc. Nonetheless, consecutively missing cases often appear in real seismic data acquired by the DAS system due to some factors, including optical fiber damage and inferior coupling between cable and well. Recently, some deep-learning (DL) seismic interpolation methods based on convolutional neural networks (CNN) have shown impressive performance in regular and random missing cases but still remain the consecutively missing case a challenging task. The main reason is that the weight sharing makes it difficult for CNN to capture enough comprehensive features. In this article, we propose a transformer-based interpolation method, called seismic interpolation transformer (SIT), to deal with the consecutively missing case. This proposed SIT is an encoder-decoder structure connected by some U-shaped swin-transformer (UST) blocks. In the encoder and decoder part, the multihead self-attention (MSA) mechanism is used to capture global features which is essential for the reconstruction of consecutively missing traces. The UST blocks are utilized to perform feature extraction operations on feature maps with different resolutions. Moreover, we introduce the SSIM loss to optimize the training process of SIT. In experiments, this proposed SIT outperforms A-Net and swin-transformer (ST). Moreover, ablation studies also demonstrate the advantages of new network architecture and loss function.

Full waveform inversion (FWI) is currently the most accurate technique for obtaining the properties of subsurface media. The absence of low frequencies in the observed data caused cycle-skipping phenomenon and poor initial model which affect the convergence of FWI. We propose a global-feature-fusion and multi-scale network (GM-Net) in a way of supervised learning to compensate for the absent low frequency components in the observed data trace by trace. The difficulty of extrapolating frequency is to achieve smoothness and continuity when changing from high frequency signals to low frequency signals, which is visually shown in the reduction and movement of the sidelobes in high-frequency signals and the overall oscillation of the signals is slowed down. For achieving better extrapolation, the encoder-decoder architecture with multi-scale feature extraction is designed as the backbone of the network. For avoiding the loss of information, we propose to perform 1/2 down-sampling on the original input signal separately based on the odd and even time samples, and then concatenate them along the channel dimension. Since 1-dimensional (1D) seismic data is a type of time-series signal and the wavelengths of low frequencies are long, we pay more attention to the relevance of contextual information. Thus, dilated convolution layers, gridding convolution blocks and non-local attention blocks are used to enlarger the receptive field both in time and channel dimensions to extract and fuse global features. Numerical tests both on synthetic data and different types of field marine data demonstrate the feasibility and generalization of our method.

Seismic exploration is a crucial method for studying underground geological structures and oil/gas resources. However, the presence of various noise sources during seismic wave propagation hinders accurate interpretation and imaging. To address this challenge, effective denoising methods are essential. In recent years, deep learning, particularly Convolutional Neural Networks (CNNs), has shown promise in seismic data processing. Nevertheless, CNNs have limitations in capturing long-range dependencies and global coherence. As an alternative, we propose a Transformer-based model called Seismic Data Denoising Transformer (SDT) for seismic signal processing. By leveraging self-attention mechanisms, the SDT model overcomes the limitations of CNNs and effectively captures long-range features for seismic signal reconstruction. We also introduce a novel self-supervised pretraining strategy using a large-scale dataset to further enhance performance. Experimental results demonstrate the advantages of SDT in complex seismic noise attenuation and preserving weak signal amplitudes. The proposed method exhibits promising potential for real-world seismic data applications.

Recorded seismic data often is characterized by a low signal-to-noise ratio (SNR) that can hinder subsequent imaging and interpretation tasks. Thus, it is necessary to explore a method to recover weak signals from strong background noise. While numerous studies have demonstrated the effectiveness of deep-learning methods in seismic noise attenuation, enhancing their capability to recover weak signals under low SNR conditions remains an area for further exploration. To address this issue, we propose an edge-feature-guided wavelet U-Net (EFGW-UNet). In this novel architecture, we utilize the discrete wavelet transform to replace the pooling operation deployed in the conventional U-Net, thereby maintaining more detailed information of effective signals. Meanwhile, we also design a dual decoder for edge detection to obtain the shape and edge information on the effective signals. Finally, to fuse multi-level image features and edge features, an attention feature fusion module is deployed. In the experimental part, we use synthetic and real data to illustrate the effectiveness of EFGW-UNet. Our results suggest better denoising performance than competitive methods, especially for weak signal recovery submerged in heavy noise.

Velocity and reflectivity models are fundamental outcomes obtained from seismic exploration, enabling the identification of subsurface structures and physical properties. Two primary methods for obtaining these high-resolution results are two-way wave equation-based full waveform inversion (FWI) and least-squares reverse time migration (LSRTM). Despite sharing a similar least-squares inversion framework, FWI and LSRTM present challenges in effectively combining them to generate both velocity and reflectivity models simultaneously due to their different modeling engines and dependencies on data components. In this study, we propose a novel joint migration inversion (JMI) approach by incorporating a two-way full-wavefield modeling engine parameterized by the subsurface velocity and reflectivity. By providing adjoint velocity and reflectivity sensitive kernels, this JMI can produce high-quality velocity and reflectivity models concurrently. The method developed in this study has the potential to directly process seismic data containing full-wavefield information, reducing data preprocessing complexity and avoiding potential damage to valid signals during the processing. Through a synthetic data test and a benchmark test, we demonstrate the effectiveness of our JMI approach. Moreover, when compared with conventional approaches for velocity building and imaging, our method yields velocity and reflectivity models with higher resolution and improved consistency.

In marine exploration, conventional migration algorithms based on primary reflections often have poor angular illumination, especially in shallow areas and complex salt boundaries. Source density is the primary controller of angular illumination. It is well-known that source density is relatively sparse in a typical marine seismic acquisition, such as in a towed streamer survey. In contrast, imaging of multiples treats each receiver as a virtual source, mimicking a high-density source survey. Imaging of multiples can provide additional angular illumination and help to improve subsurface imaging. The extra illumination provided by multiple reflections enhances angle-domain common-image gathers (ADCIGs), a component of velocity model building and amplitude versus angle (AVA) analysis. Our main objective is to illustrate the advantages of angular illumination from imaging of multiples in the angular domain. To achieve this purpose, we propose a workflow for calculating high-quality ADCIGs for imaging of multiples. The workflow mainly contains up-going and down-going wavefield decomposition and the stabilized Poynting vectors for calculating more accurate imaging angles. To illustrate the accuracy of the proposed workflow, a simple model is used to calculate angle gathers for imaging of multiples. Then, the Sigsbee2b model and a field dataset from the Gulf of Mexico are used to compute angle gathers by imaging primaries and multiples to demonstrate the improved angular illumination achieved when multiples are also used for imaging. The comparison results show that the angle gathers obtained by imaging of multiple has better angular illumination, especially in shallow areas and complex salt boundaries.

In seismic exploration, erratic noise is a type of intense and complicated interference with large-amplitude and non-Gaussian distributions. The presence of erratic noise has been demonstrated to corrupt reflection events, adversely affecting the identification of effective signals. Nonetheless, conventional and time-frequency thresholding denoising methods based on the least-squares scheme usually assume that the seismic random noise has a Gaussian distribution, which is not the case for erratic noise. Therefore, the attenuation for erratic noise is challenging, owing to the deviation from the assumptions of the conventional methods. The recent application of convolutional neural networks (CNNs) to seismic data processing has yielded promising results. However, these CNN-based frameworks always have limited feature-interaction capability, resulting in the degeneration in denoising performance when coping with intense erratic noise. To address this issue, a novel joint-guided denoising network (JGD-Net) is proposed in this study. Unlike conventional CNN frameworks, JGD-Net uses a joint-guided scheme and attention mechanism to enhance the denoising capability. We generate synthetic records using published geological models such as Marmousi and salt dome to compose our training dataset. Furthermore, a novel loss function based on L1 norm and hyperbolic tangent function is designed to further ensure the optimization process of the training procedure and ease the influence of abnormal energy of erratic noise. Both synthetic and field data are processed sfor the evaluation of denoising performance. Compared with other popular methods, JDG-Net shows advantages in attenuating intense erratic noise, particularly under extremely low signal-to-noise ratio (SNR) conditions.

The super-resolution reconstruction for seismic images obtained by multistep processing of field data is essential due to the noise contamination, sparse geometry, and low dominant frequency of events, which impairs the subsequent seismic interpretation. Deep learning-based methods show strong potential in super-resolution through supervised learning. Generative adversarial networks (GANs) have shown capability in super-resolution of different kinds of images; however, it is limited in enhancing the detailed geological structures of seismic images that are fatal for interpretation. To address this issue, we propose a super-resolution hybrid bilateral attention GAN (SHBGAN) to improve the recovery of weak signals and the reconstruction of geological structures. Specifically, the generator employs hybrid and bilateral attention modules (BAMs) to enhance the capture ability of global and local features. Meanwhile, we use dilated convolutional layers instead of batch normalization (BN) layers in the residual block to improve the generalization ability of the trained model. Meanwhile, the discriminator employs global average pooling and convolutional layers to score the authenticity of seismic images rather than the probability to enhance the stability of training. In addition, we add the mean structural similarity (MSSIM) term to the loss function of generator to improve the perception quality of predictions. The numerical tests on both synthetic and field data show that SHBGAN is more effective than competing methods in recovering weak signals and reconstructing subtle faults.

Distributed optical fiber acoustic sensing (DAS) is an emerging acquisition technology in seismic exploration. However, DAS records are always affected by the complex background noise, resulting in a low signal-to-noise ratio (SNR). In addition, the DAS background noise has different properties from the noise existing in conventional seismic data. Thus, conventional denoising methods may degrade the record when dealing with complex DAS data. To improve the denoising capability, a novel denoising network, called residual modular cascaded heterogeneous network (RMCHN), is proposed. In general, the network is based on the idea of heterogeneous convolution and modular convolutional neural networks. Specifically, different modules are designed to extract the discriminatory features of the DAS data through effective information integration. On this basis, heterogeneous convolution combined with long and short path feature learning strategy is employed to fuse the captured features, thereby improving the feature expression capability and avoiding the information loss. Both synthetic and field denoising results indicate that RMCHN can suppress the DAS background noise with excellent performance in signal restoration, even for the weak signals form deep strata.

Seismic data reconstruction is always an essential step in the field of seismic data processing. Effective reconstruction methods can obtain high-density information at low-cost and also recover missing seismic data. Due to the strong feature extraction ability, convolutional neural network (CNN) has shown remarkable performance in numerous fields of data processing and been gradually applied to seismic data reconstruction. However, most of CNN-based methods applied to seismic data reconstruction only consider features in single scale or just utilize simple interactions between different scales, which is likely to result in performance degradation when facing complex and extremely incomplete seismic data. To further promote the performance of CNN-based methods in seismic data reconstruction, a novel multiscale enhanced attention network (MSEA-Net) is proposed based on the self-enhanced scheme. In general, MSEA-Net has a multiscale architecture which can significantly improve the processing accuracy by fusing the potential features in different-resolution seismic data. From the basis, a parallel sparse residual block is designed and applied in MSEA-Net to enhance processing efficiency and avoid overfitting issues. In addition, a dense spatial attention block is also introduced to the network to further reinforce the effective features, thereby strengthening the reconstruction performance. Experimental results demonstrate that our proposed network can effectively reconstruct incomplete seismic data including regular missing data, irregular missing data, and even consecutively missing data with big gap, which is superior than exist interpolation methods including commonly used U-Net.

Recently, a number of deep learning (DL) methods are developed to attenuate the noise in seismic data. Most of them show good performance under a common precondition: the training and testing data are drawn from the same distribution. However, it is challenging to acquire sufficient noise training data whose distribution is same as the noise presented in the testing seismic data; we call such noise data with same distribution as target-domain noise. To address this issue, we propose a promising DL paradigm for seismic data denoising based on transfer learning and seismic noise modeling. We firstly utilize a Green-function-based modeling method for seismic noise to generate a massive amount of synthetic noise which is similar to real seismic noise. Secondly, the high-authenticity synthetic noise is used as the pre-training data in source domain. Finally, we utilize limited real target-domain noise data to fine-tune the partial trainable parameters of pre-trained model and thus transferring it into target domain. This proposed DL paradigm gets rid of the need for enough target-domain noise data, so as to extend the application scope of DL-based method in seismic data denoising. Moreover, we design a novel network architecture based on multi-cascade structure and attention mechanism. This DL paradigm shows extremely similar denoising performance to that of using a large amount of target-domain noise data in both synthetic and real examples, demonstrating its potential in mitigating the dependence of DL-based seismic denoising methods on target-domain noise data.

The formulation of conventional least-squares reverse time migration (LSRTM) starts with the forward modeling process; as a result, the migration operator of it presents as a migration process with a cross correlation imaging condition (CCIC). Since the imaging results produced by CCIC usually contain undesirable components (e.g., strong backscattering noise), it can be assumed that the primary target of the conventional LSRTM is to fit input data rather than produce high-quality imaging results; therefore, conventional LSRTM can be considered as a modeling-driven algorithm. To mitigate the desirable component in the imaging results, additional efforts should be spent in the process of modeling-driven LSRTM. To improve the performance of the LSRTM, we develop a migration-driven LSRTM by formulating the migration process using the inverse scattering imaging condition (ISIC) first. To guarantee the convergence of the algorithm, an adjoint modeling operator and a data precondition operator are incorporated in this migration-driven LSRTM. Since the ISIC can effectively eliminate the backscattering noise, this migration-driven LSRTM can produce high-quality images without the influence of that. After two synthetic data tests, this approach is applied to a 2-D streamer field dataset from the Gulf of Mexico. These tests indicate that, compared to the modeling-driven LSRTM, the migration-driven LSRTM approach can solve the inversion problem more robustly and efficiently.

Conventional least-squares reverse time migration (LSRTM) generally involves a de-migration operator based on the first-order scattering approximation (Born modeling), which can only simulate the seismograms containing the primary reflected wave. When the input observed seismograms contain “redundant information” (especially multiples), crosstalk may occur in the imaging results. Therefore, we develop a least-squares full-wavefield reverse time migration (LSFWM), which is implemented based on a two-way modeling engine with vector reflectivity and the corresponding adjoint sensitive kernel. This modeling engine is modified from the variable density acoustic wave equation and can simulate the subsurface wavefield containing the primaries and multiples only by giving the accurate or estimated subsurface reflectivity and velocity. Theoretically, this LSFWM approach can eliminate the influence of “redundant information” on imaging and provide higher-quality imaging results compared to conventional LSRTM. In addition, since the modeling engine is based on vector reflectivity, the imaging results produced by the LSFWM are also vectorized, which can give more information about the subsurface structures, especially steep structures. And the imaging results produced by the LSFWM can accurately depict the subsurface reflectivity. These are helpful to obtain the information on subsurface structure and physical properties more clearly.

Addressing the significant impact of undulating terrain on 3D Direct Current inversion, the unreliability of traditional linearized polarizability inversion for highly polarizable anomalies, and the non-uniqueness of separate resistivity and Time-Domain Induced Polarization inversion, this paper conducts a joint inversion study of 3D Direct Current resistivity and Time-Domain Induced Polarization with structural constraints in undulating topography. A regularly arranged deformed hexahedron mesh simulates undulating surface terrain, transformed into regular hexahedron elements for 3D undulating terrain DC resistivity modeling. Based on the exact inversion of polarization calculated from the inversion results of apparent resistivity and equivalent apparent resistivity data, a joint inversion of resistivity and polarization constrained by cross-gradient is implemented. Synthetic data examples show that the application of arbitrary hexahedron elements significantly reduces the influence of terrain on inversion, and the implementation of joint inversion markedly improves the recovery of high-polarization anomalies while enhancing both the model resolution and the inversion accuracy. The proposed algorithm is applied to the joint inversion of resistivity and polarizability in the lead-zinc mining area of Xiagalaiaoyi River in Huzhong area, the Great Khingan Mountains, northwestern Heilongjiang Province, achieving good results.

Three-dimensional gravity inversion is an effective method for extracting underground density distribution from gravity data. However, traditional deterministic gravity inversion methods suffer from problems such as skin effect, low computational accuracy, and poor efficiency. Therefore, we propose a three-dimensional gravity data focusing inversion algorithm based on the supervised descent method. Supervised descent method (SDM) is a non-linear optimization method based on the combination of machine learning and gradient descent method. In the offline phase, we construct a training set based on a priori information and iteratively learn a set of average descent directions between the initial model and the training model. In the online phase, we introduce a focused regularization into the prediction objective function. This addition aims to obtain a sharp boundary density model that conforms to the physical distribution. Additionally, we incorporate property boundary constraints in both the offline and online phases to control the upper and lower bounds of the density values to ensure consistency with reality. Model tests show that the proposed method can effectively overcome skin effect, improve the resolution of gravity inversion. Moreover, the construction of the training set of the proposed method is less affected by prior information, and it has strong generalization ability. Furthermore, the method does not require solving large-scale linear equations, accelerating the inversion computation speed and having strong noise resistance. Field examples demonstrate that this method has good potential for improving the accuracy and efficiency of actual gravity data inversion.

As complementary to the imaging of primaries, multiples can be used for imaging and produce additional angular illumination. These additional illumination angles can improve imaging resolution. There are several methods to calculate the angle gathers from imaging of primaries, among which the directional vector methods are matured and have been implemented in practice. For conventional directional vector algorithms using common-shot reverse time migration, usually, only one angle with the main contribution to each imaging point is collected, which is referred to as a one-shot, one-position, and one-angle mapping strategy. This strategy is appropriate for imaging of primaries; however, the principle fails in the imaging of multiples, where more than one angle exists at an imaging point even using a shot gather. Therefore, conventional directional vector approaches cannot separate different imaging angles produced by imaging of multiples. In this article, we introduce the Poynting vector method to calculate angle gathers for imaging of multiples. Based on the one-shot, one-position, and one-angle mapping strategy, we search for the peak of modulus of Poynting vectors in a time window to differentiate imaging angles at an imaging point. The ray path diagrams and angle-gather imaging conditions are used to demonstrate the improved angular illumination from imaging of multiples. Then, the arriva -time equation is derived mathematically to demonstrate wavefield arrival time differences for different imaging angles from imaging of multiples. Based on the arrival time differences, a Poynting vector algorithm is proposed to compute more than one angle at one subsurface location for imaging of multiples. Data examples are used to validate our approach.

Distributed acoustic sensing (DAS) has been considered as a breakthrough technique in seismic data collection owing to its advantages in acquisition cost and accuracy. However, the existence of complex background noise combined with a tough exploration environment always results in incomplete data with a low signal-to-noise ratio, posing a big challenge for the subsequent processing of DAS data. To improve the quality of DAS data, convolutional neural networks (CNN) have gradually been utilized to deal with the denoising and reconstruction tasks. Meanwhile, some successful applications have verified that CNN-based methods can significantly alleviate the impacts of DAS background noise and missing trace records, compared with conventional approaches. Nonetheless, in most researches, the denoising and reconstruction tasks are accomplished independently, severely affecting the processing efficiency. In this study, a multi-scale encoder-decoder network (MEDN) is proposed to simultaneously achieve the DAS background noise suppression and weak signal recovery through a unified model. Generally, MEDN can extract the different-scale features through both a multi-scale network architecture and a multi-scale residual (MSR) block. The captured different-scale features are then fused to enhance the effective feature. In addition, the encoder-decoder scheme is also utilized in the design of the network architecture to further enhance the reconstruction performance. Moreover, depthwise separable convolution (DSC) blocks are also utilized to ease the computational burden and improve the processing efficiency. Theoretical and field data processing results show that MEDN can provide better denoising and reconstruction performance than conventional methods and popular CNN-based frameworks.

Distributed fiber-optical acoustic sensing (DAS) is a new and booming technology in seismic exploration. DAS technology has been gradually applied to the exploration of vertical seismic profile (VSP) due to its strong resistance to high temperature and pressure, high sensitivity, high precision (trace interval can be accurate to about 1 m), and so on. However, real DAS-VSP data are always contaminated by both random and coherent noises, which greatly affects the quality of DAS-VSP data. In order to suppress the background noise and increase the signal-to-noise ratio (SNR), a convolutional neural network (CNN) based on leaky rectifier linear unit (ReLU) and forward modeling is proposed and named L-FM-CNN. In terms of network architecture, Leaky ReLU is adopted as the activation function of CNN, which can enhance the recovery ability of trained CNN denoising model to the weak effective signals. As for the training data set, we construct a high-authenticity theoretical pure seismic data set for DAS-VSP data through the complexity of forward models and the diversification of physical parameters. In addition, we propose a new mean square error (MSE) loss function combined with an energy ratio matrix (ERM). The ERM can adjust the SNR between the signal patch and noise patch during the network training and thus increase the robustness of trained CNN denoising model for the DAS-VSP data with different SNRs, especially the DAS-VSP data with extremely low SNR. Both synthetic and real experiments prove the effectiveness of the proposed L-FM-CNN.

For imaging and interpretation, high-quality seismic data are necessary. However, noise, which is strong in field desert seismic data, inevitably diminishes the quality of the data and reduces the signal-to-noise ratio. Moreover, the effective signals and noise in field desert seismic data are mostly distributed in the low-frequency band, which leads to severe spectral aliasing. Recently, some deep learning methods have improved the quality of desert seismic data in certain aspects. However, due to limitations of their networks and the serious spectral aliasing of desert seismic data, the denoising results usually show some false seismic reflections. To solve the above problems, we introduce Unsupervised Generative Attentional Networks with Adaptive Layer-Instance Normalization for Image-to-Image Translation (U-GAT-IT) to the denoising of desert seismic data in a semisupervised manner. U-GAT-IT is an unsupervised attentional generative adversarial network (GAN) combined with an attention module guided by the class activation map (CAM). The attention module guided by the CAM can guide the model to better distinguish between noise and effective signals. The experiment shows that the U-GAT-IT can effectively suppress desert seismic noise. Also, the denoising result has fewer false seismic reflections.

Recently, distributed optical fiber acoustic sensing (DAS) is regarded as a transformative technology in seismic exploration. However, both various complex background noise and weak desired signals significantly limit its practical application. To explore an effective denoising method for the vertical seismic profile (VSP) record received by DAS, we propose an improved residual encoder–decoder deep neural network (RED-Net) enhanced by deep iterative memory block (DMB) and channel aggregation block (CAB), called residual channel aggregation encoder–decoder network (RCEN). Here, DMB uses the weight accumulation theory to improve the feature extraction ability and achieve accurate noise elimination. Meanwhile, CAB, using the multi-channel analysis architecture, enhances the weak signal retention performance. In addition, we leverage both the synthetic data obtained by forward modeling and real DAS noise data to construct a sufficient training dataset with high authenticity, thereby meeting the requirement of network training. Both the synthetic and field DAS-VSP data processing results demonstrate the advantage of RCEN compared with competing algorithms, including singular value decomposition (SVD), conventional RED-Net, and feed-forward denoising convolutional neural network (DnCNN).

Seismic background noise often damages the desired signals, thereby resulting in some artifacts in the seismic imaging that follows. Since about 2016, some supervised-deep-learning methods have shown impressive performance in seismic data denoising, but they usually only consider single-scale features and neglect the multi-scale strategy. To further reinforce their denoising performance, a novel multi-scale convolutional neural network (CNN) combined with spatial attention mechanism, called multi-scale spatial attention denoising network (MSSA-Net), is proposed to tell weak reflected signals apart from strong seismic background noise. Unlike conventional single-scale CNNs, this proposed MSSA-Net can achieve the extraction of multi-scale features which is beneficial for the suppression of strong noise and the recovery of weak reflected signals. Specifically, MSSA-Net contains a principal denoising network and two auxiliary networks. The former utilizes the widen convolution composed of multiple parallel convolution layers with different kernel sizes to capture the informative multi-scale features; the latter two leverage up and down sampling to extract local fine and global coarse features, respectively. Furthermore, a spatial attention block is adopted to fuse these multi-scale features, thereby distinguishing weak reflected signals from strong seismic background noise. Multiple experiments of synthetic and real seismic records demonstrate the effectiveness of MSSA-Net. In addition, compared with two classical single-scale CNNs, MSSA-Net performs better in signal recovery, indicating the positive effect of multi-scale strategy.

Seismic prospecting is a common method used in oil and gas resource exploration. However, due to the limitations of current collection techniques, seismic records acquired in the field are typically contaminated by severe incoherent noise, which has negative implications for the subsequent processing and interpretation procedures. In addition, numerous traditional denoising algorithms have been applied in order to mitigate this problem, but further improvements are required, especially for the seismic data with spectral overlapping between effective signals and background noise. In recent years, feedforward denoising convolutional neural networks (DnCNNs) have been applied to suppress the complex random noise, and a series of essential insights have been gained. Nonetheless, conventional denoising networks always extract data features depending on single-scale information, resulting in impaired performance when coping with seismic records with a low signal-to-noise ratio (SNR). For solving this problem, a novel multiscale DnCNN (MSDCNN) is developed as an attempt for random noise suppression. Unlike conventional DnCNN, MSDCNN has a hierarchical structure capable of extracting features at different scales and capturing informative and discriminatory features through effective information integration. Meanwhile, the cross-scale feature interaction also increases the processing accuracy when confronted with weak reflection events. Experimental results derived from both synthetic and field data indicate that the proposed network can effectively suppress the random noise and accurately preserve reflection events, even under low SNR conditions.

Seismic background noise affects the recognition of reflection signals, thereby impeding the subsequent seismic data processing, such as seismic imaging and inversion. In addition, seismic background noise has relatively complex properties, such as non-stationarity and spectral aliasing, which can be further hampered with the deterioration of the exploration environment. Deep-learning methods have been successfully applied to effectively attenuate complex seismic noise and have shown significant improvements over conventional denoising methods. However, most denoising networks only utilize single-scale features, resulting in poor performance when confronted with seismic data in a low signal-to-noise ratio. To further enhance the denoising capability, a novel multiscale residual pyramid network (MRP-Net) was proposed to separate the desired signals and complex seismic noise. Compared with single-scale networks, MRP-Net can take advantage of the multiscale features, thereby improving noise attenuation capability. In general, the pyramid-like framework in MRP-Net can extract the potential features at different scales through down-sampling and up-sampling operations, and skip connections were applied to fuse the global-coarse and local-fine features. On this basis, a double-path spatial attention module was designed to enhance the desired features, further improving the processing performance of separating the desired signals from the intense seismic background noise. Comprehensive experiments on synthetic and field seismic data demonstrate that MRP-Net is effective for complex seismic noise attenuation, both for conventional geophone-acquired data and DAS records.

At present, almost no denoising method can effectively suppress the seismic random noise in different areas. This phenomenon is partially because of two reasons: 1) the variable dominant frequency (DF) distribution of random noise in different areas and 2) the different signal-to-noise ratios (SNRs) of the seismic data acquired from different areas. We have developed a deep-learning denoising method to suppress the random noise in different areas based on convolutional neural network (CNN). For a certain area, we leverage the wave equation and power spectrum analysis to construct a noise set whose DF distribution is close to that of the real random noise in this area, and then a CNN denoising model for this area can be obtained via the training of this noise set. In addition, an energy ratio factor is used to adjust the energy ratio of effective signal patch and noise patch in the training process, so as to improve the generalization ability of CNN denoising model to different SNRs. Experiments demonstrate that our method can effectively suppress the random noise in different areas and completely recover the effective events.st no denoising method can effectively suppress the seismic random noise in different areas. This phenomenon is partially because of two reasons: 1) the variable dominant frequency (DF) distribution of random noise in different areas and 2) the different signal-to-noise ratios (SNRs) of the seismic data acquired from different areas. We have developed a deep-learning denoising method to suppress the random noise in different areas based on convolutional neural network (CNN). For a certain area, we leverage the wave equation and power spectrum analysis to construct a noise set whose DF distribution is close to that of the real random noise in this area, and then a CNN denoising model for this area can be obtained via the training of this noise set. In addition, an energy ratio factor is used to adjust the energy ratio of effective signal patch and noise patch in the training process, so as to improve the generalization ability of CNN denoising model to different SNRs. Experiments demonstrate that our method can effectively suppress the random noise in different areas and completely recover the effective events.

The seismic data with high quality are the essential foundation of imaging and interpretation. However, the real seismic data are inevitably contaminated by noise, which affects the subsequent processing and interpretation of seismic data. In desert seismic data, the energy of noise is stronger. Also, the frequency-band overlap between noise and effective signals is more serious. Recently, some methods based on supervised learning can suppress the desert seismic noise to some extent. Generally, supervised learning-based methods use synthetic noisy data and paired pure data as training sets to train model. However, the difference between synthetic noisy data of training and real seismic data of testing leads to the degradation of the model, and the denoising results often have many false seismic events when dealing with field seismic data. To solve the above problem, we introduce Cycle-generative adversarial networks (GANs) into the denoising of desert seismic records. Cycle-GAN is an unsupervised learning-based method. It can learn the domain mapping from noisy data domain to effective signal data domain through unpaired data training. So we use unpaired real desert common-shot-point data and synthetic pure data to train Cycle-GAN, so as to effectively improve the denoising ability of the method for real seismic data. Finally, the denoising of desert seismic data is realized. The experiment shows that the Cycle-GAN with unpaired data training can effectively suppress desert seismic noise and retain the effective signal amplitude. Also, the denoising result has less false seismic reflection.

Distributed optical fiber acoustic sensing (DAS) is a new and rapid-developing detection technology in seismic exploration. Unfortunately, due to the weak energy of scattered optical signals and the inferior coupling between DAS cable and receiving interface, the seismic data received by DAS are often characterized by low signal-to-noise ratio (SNR); this low SNR is likely to affect some subsequent analysis, such as inversion, imaging, and interpretation. In addition, the noise caused by the inferior coupling is a new kind of noise not presented on conventional seismic data. To enhance the SNR of DAS seismic data and suppress the DAS noise effectively, we propose a convolutional adversarial denoising network (CADN) based on the basic strategy of generative adversarial network (GAN) and the usage of a denoiser to replace the original generator in GAN. In CADN, the performance of denoiser is significantly strengthened via its own mean square error (MSE) loss and the adversarial loss between it and the discriminator. To balance the two losses and thus ensure the optimization of denoiser, we construct a novel loss function, where the optimal ratio of MSE and adversarial losses is determined by quantifying the denoising performance. Both real and synthetic examples are included to testify the denoising performance of CADN. Experimental results have demonstrated that CADN can suppress most of the DAS noise and enhance the SNR of DAS seismic data; also, it can recover the effective signals completely, even the extremely weak effective signals reflected by deep layers.

Seismic exploration is a kind of exploration method for oil and gas resources. However, the disturbance of numerous random noise will decrease the quality and signal-to-noise ratio (SNR) of real seismic records, which brings difficulties to the following works of processing and interpretation. The seismic records of desert region pose a particular problem because of the strong energy noise and the spectrum overlapping between effective signals and random noise. Recent research works demonstrate that a convolutional neural network (CNN) can increase the SNR of seismic records. The optimization of denoising methods based on CNN is principally driven by the loss functions that largely focus on minimizing the mean-squared reconstruction error between denoising records and theoretical pure records. The denoising results estimated by the CNN model are often lacking the perfection of the signal structure. Therefore, when processing seismic records with low SNR, the denoising results often have a lack of effective signal in some traces, which leads to the poor continuity of events. In order to solve this problem, we adopt the strategy of generative adversarial network (GAN) to construct a GAN for denoising. It is divided into two parts: the generator (the denoising network based on CNN) is used to remove noise, while the discriminator is used to guide the generator to restore the structure information of effective signals. The generator and discriminator enhance the performance of each other through adversarial training, and the generator after adversarial training can greatly recover events and suppress random noise in synthetic and real desert seismic data.

Attenuating the random noise and improving the signal-to-noise ratio (SNR) for the seismic data are of great significance in industrial exploration. In recent years, fractal conservation law (FCL) has been proposed and applied to seismic random noise suppression successfully. However, in conventional FCL, the filtering parameter selection strategy is relatively simple and a fixed parameter set is always used for the whole seismic record. In addition, it is very difficult to make an excellent tradeoff in random noise attenuation and signal preservation only by fixed parameters especially under the low-SNR conditions. Thus, accurately recognizing the effective signals and adaptively choosing appropriate filtering parameters is a feasible approach to improve the performance of the conventional FCL. In this article, an adaptive FCL methodology is proposed by combining the seismic noise analyzing theory and stationarity testing techniques. It is known that the random noise and reflection signals have different properties in stationarity and thus, the signal and noise segments can be divided by stationarity testing. As a consequence, different filtering parameters can be adopted for signal and noise segments to achieve the noise suppression and signal preservation simultaneously. Synthetic and field data experiments demonstrate that the proposed method can remove the random noise from seismic record and effectively preserve the reflection events.

Lots of low-frequency noise including random noise and surface waves seriously reduces the quality of desert seismic data. However, the suppression for desert low-frequency noise faces three main problems: nonstationary and non-Gaussian of random noise; strong energy of low-frequency noise; a more serious frequency-band overlap between effective signals and low-frequency noise. Robust principal component analysis (RPCA) is a classical low-rank matrix (LM) recovery method which is very suitable for processing nonlinear noise. It can decompose noisy data to the optimal LM and sparse matrix (SM), which include most effective signals and noise, respectively. Therefore, the RPCA is introduced to suppress desert low-frequency noise. However, due to the low signal-to-noise ratio (SNR) and serious frequency-band overlap, much low-frequency noise still remains in the LM of desert seismic data after the decomposition of RPCA. Meanwhile, some nonnegligible effective signals are decomposed into the SM of desert seismic data. To solve this problem, the convolutional neural network (CNN) is introduced to extract effective signals from SM and LM. By constructing suitable training sets to guide the CNN's training, the CNN denoising models after training are used to predict the effective signals from these two matrices, respectively. In this article, to approach real desert seismic data, we use a variety of seismic wavelets to simulate different types of seismic events, and then use these synthetic seismic events and real desert low-frequency noise to construct training set. In experiments, our method can raise the SNR of synthetic noisy data from -8.69 to 9.63 dB.

High-quality seismic data are the basis for stratigraphic imaging and interpretation, but the existence of random noise can greatly affect the quality of seismic data. At present, most understanding and processing of random noise still stay at the level of Gaussian white noise. With the reduction of resource, the acquired seismic data have lower signal-to-noise ratio and more complex noise natures. In particular, the random noise in the desert area has the characteristics of low frequency, non-Gaussian, nonstationary, high energy, and serious aliasing between effective signal and random noise in the frequency domain, which has brought great difficulties to the recovery of seismic events by conventional denoising methods. To solve this problem, an improved feed-forward denoising convolution neural network (DnCNN) is proposed to suppress random noise in desert seismic data. DnCNN has the characteristics of automatic feature extraction and blind denoising. According to the characteristics of desert noise, we modify the original DnCNN from the aspects of patch size, convolution kernel size, network depth, and training set to make it suitable for low-frequency and non-Gaussian desert noise suppression. Both simulation and practical experiments prove that the improved DnCNN has obvious advantages in terms of desert noise and surface wave suppression as well as effective signal amplitude preservation. In addition, the improved DnCNN, in contrast to existing methods, has considerable potential to benefit from large data sets. Therefore, we believe that it can open a new direction in the area of seismic data processing.