Geophysical Prospecting - Volume 74, Issue 1, 2026

Accurately detecting the locations of fractures and the permeability structure within a subsurface reservoir can significantly improve the optimization of production performance. Achieving peak performance in subsurface groundwater or hydrocarbon reservoirs depends on creating an accurate map that details the reservoir's characteristics derived from the history‐matching process. However, this process involves repeated forward modelling simulations until the results align with historical production data, often consuming significant resources and potentially yielding non‐unique reservoir models. An integration approach between bottom‐hole pressure data and surface self‐potential measurements was used to perform simultaneous inversion for the permeability structure. The self‐potential method, a cost‐effective geophysical technique, allows for the inversion of subsurface self‐potential sources based on the underlying resistivity structure. Through a series of synthetic experiments, we demonstrate that combining borehole pressure data with surface self‐potential measurements significantly enhances reservoir characterization, providing more robust and accurate subsurface models. By using the resolution matrix, we further confirm that the solution achieves higher accuracy when both data sets are integrated. This approach not only improves the precision of reservoir mapping but also reduces the uncertainty typically associated with traditional methods, offering a more efficient and reliable tool for optimizing production performance.

As demand for large‐scale seismic data interpretation tasks increases, machine learning‐based horizon autotracking methods have gained attention in the geological and geophysical fields. Although such methods have demonstrated time‐ and cost‐efficiency in large‐scale data interpretation, studies on the expansion of interpreted horizons into the reservoir characterization process are relatively limited. Hence, a reservoir characterization process that can incorporate the machine learning‐interpreted horizons and their structural uncertainties into the reservoir uncertainty assessment is necessary for an efficient reservoir modelling process. The proposed workflow consists of various modelling processes, including horizon construction where machine learning‐interpreted horizons are used instead of manually interpreted horizons, facies modelling and petrophysical modelling. The modelling algorithms are based on stochastic methods: sequential indicator simulation for facies models and Gaussian random function simulation for petrophysical properties. Each modelling process incorporates variables such as variogram parameters, facies ratios and modified porosity values. The results show promising performance in incorporating machine learning‐interpreted horizons into the uncertainty quantification process and analysing their impact by capturing the influence of structural uncertainties of horizons in the final reservoir pore volume.

During seismic data processing, strong single‐frequency interference noise often affects data quality. Traditional methods for single‐frequency interference identification are typically conducted in the frequency domain, primarily by searching for abnormal peaks in the frequency spectrum of each trace. However, when the interference amplitude in the frequency domain is weak relative to the entire seismic frequency, the identification process becomes significantly more challenging. To address this issue, this article proposes a time‐domain approach for identifying single‐frequency interference. First, frequency analysis is performed on seismic data containing single‐frequency interference to obtain the initial frequency , and sine and cosine signals with an amplitude of 1 at this frequency are then generated. Next, the seismic data are normalized to balance amplitude differences across different datasets in the time domain. After normalization, deep‐time seismic data are cross‐correlated with the generated sinusoidal and cosine signals, and correlation coefficient R is computed to determine whether suppression is necessary. On the basis of theoretical simulations and field data analysis, suppression is considered necessary when R > 0.001. Finally, for identified single‐frequency interference, a hierarchical approximation method based on a cross‐correlation objective function is employed to search for interference frequencies with finer step sizes near the initial frequency and calculate the corresponding amplitudes. The interference signal is subsequently subtracted in the time domain to achieve interference suppression. Through synthetic experiments and the application analysis of various field seismic data (including single‐shot and stacked profiles), the proposed method demonstrates high efficiency and accuracy in identifying and suppressing single‐frequency interference.

The Marchenko method is a powerful tool for reconstructing full‐wavefield Green's functions using surface‐recorded seismic data. These Green's functions can then be utilized to produce subsurface images that are not affected by artefacts caused by internal multiples. Despite its advantages, the method is computationally demanding, primarily due to the iterative nature of estimating the focusing functions, which links the Green's functions to the surface reflection response. To address this limitation, an optimization‐based solver is proposed to estimate focusing functions in an efficient way. This is achieved by training a network to approximate the forward modelling problem on a small subset of pre‐computed focusing function pairs, mapping final up‐going focusing functions obtained via the conventional iterative scheme to their initial estimates. Once trained, the network is fixed and used as the forward operator within the Marchenko framework. For a given target location, an input is initialized and iteratively updated through backpropagation to minimize the mismatch between the output of the fixed network and the known initial up‐going focusing function. The resulting estimate is then used to compute the corresponding down‐going focusing function and the full Green's functions based on the Marchenko equations. This strategy significantly reduces the computational cost compared to the traditional Marchenko method based on the conventional iterative scheme. Tests on a synthetic model, using only 0.8% of the total imaging points for training, show that the proposed approach accelerates the imaging process while maintaining relatively good imaging results, which is better than single scattering imaging. Application to the Volve field data further demonstrates the method's robustness and practicality, highlighting its potential for efficient, large‐scale seismic imaging.

The solution of magnetotelluric equations is used to determine the apparent resistivity and to model the electromagnetic field's behaviour within the Earth. In this paper, we extend the scaled boundary finite element method (SBFEM) to compute the solutions of magnetotelluric equations. The salient features of the proposed framework are that internal features and boundaries are captured through a quadtree decomposition. The SBFEM handles the resulting hanging nodes as a part of local refinement efficiently without needing additional constraints or shape functions. Further, we employ patterns to speed up the computations of the essential matrices without compromising accuracy. The results from the present approach are compared with other approaches, and it is seen that the SBFEM framework is not only efficient but also accurate. The efficacy and robustness are demonstrated with a few examples.

High‐resolution seismic data are essential for interpreting thin‐layered stratigraphy and subtle structures within hardrock media, as this information can lead to better exploration decisions in the mining sector. Conventional resolution enhancement techniques, such as spectral broadening and supervised deep‐learning techniques, often rely on oversimplified assumptions or require high‐resolution training labels. These limitations restrict their applicability in real seismic data processing, especially for hardrock seismic data, which are typically characterized by high velocities, strong heterogeneity and short reflector continuity due to complex emplacement contacts and geological settings. We propose an unsupervised seismic resolution enhancement framework that integrates physics‐guided and attention‐based mechanisms. The framework is designed to address the progressive loss of high‐frequency information in seismic exploration and the limitations of conventional resolution enhancement methods, which struggle to balance imaging fidelity with geological interpretability. The proposed network incorporates coordinate attention blocks and the lightweight Vision Transformer, enabling more effective capture of spatial dependencies and long‐range significant features in complex geological settings. Specifically, our approach utilizes a physics‐constrained deconvolutional loss function, where the predicted reflectivity is regularized by an adaptive sparsity prior and convolved with a wavelet to synthesize seismic traces that are consistent with the observed data. In addition, a robust Charbonnier penalty ensures stable physical fitting, while anisotropic total variation regularization improves lateral continuity. Following this design, the model achieves end‐to‐end recovery of high‐resolution seismic information without requiring high‐resolution labels, thereby explicitly embedding physical constraints into the learning process. Testing results on synthetic and field datasets from two different regions demonstrate that the proposed method significantly enhances vertical resolution, reflector sharpness and lateral continuity, enabling more precise delineation of subtle stratigraphic features within the target intervals. Compared with spectral enhancement and conventional deep‐learning methods, our approach achieves higher seismic reconstruction fidelity and more interpretable reflectivity, providing an option that combines robustness with interpretability for high‐resolution imaging in complex geological conditions.

Accurate estimation of seismic velocity and density models is essential for subsurface imaging and characterisation. We propose a global optimisation framework to estimate non‐linked P‐wave velocity (Vp) and density (ρ) for post‐stack seismic data, and Vp, ρ and S‐wave velocity (Vs) for pre‐stack data, by minimising the misfit between non‐corrected normal moveout (non‐NMO) observed and synthetic seismic gathers. We demonstrate that a non‐linear, underdetermined and complex problem can be addressed by introducing an additional constraint on one of the parameters, using non‐linear forward modelling combined with global optimisation algorithms. The synthetic seismic gathers are iteratively generated by randomly and simultaneously updating initial models. Updates are accepted on the basis of the simulated annealing method, a global optimisation technique that helps to prevent entrapment in local minima. Optimisation is performed by minimising an L2‐norm misfit function. For the pre‐stack case, the observed seismic data are a real gather. For the post‐stack case, the observed gather is formed from the full stacked seismic trace, taken as a near trace, and the reflectors are spread along moveout curves that are computed from the smoothed log of Vp. A two‐parameter search (Vp, ρ) is launched in the post‐stack case, whereas a three‐parameter search (Vp, ρ and Vs) is used when pre‐stack data are available, with both starting from smoothed initial models. To reduce the number of iterations by at least one order of magnitude and to increase computational efficiency, initial estimates of the Vp and ρ models can first be obtained from the post‐stack process using well logs. Then, the full three‐parameter search is performed, starting with initially estimated models for Vp, ρ and smoothed Vs. The proposed methodologies offer an approach for estimating elastic properties and for overcoming the limitations of conventional seismic inversion. The approach eliminates reliance on regression techniques, which often oversimplify the complex, nonlinear relationships between seismic data and subsurface properties; avoids linearisation of the inversion process, which can introduce errors again due to the inherently nonlinear nature of geological properties; and bypasses the need for normal moveout (NMO) correction, which can cause amplitude stretching and distort critical amplitude information, because the algorithm utilises the moveout. By working with fully migrated 1D gathers, we assume a constant velocity across all offsets, which allows us to bypass full wave‐equation modelling while maintaining acceptable accuracy. Additionally, the framework supports the implementation of alternative global optimisation strategies.

Seismic inversion can effectively establish the connection between seismic data and underground reservoir parameters. Aiming at the current problem of low accuracy of deterministic inversion, a seismic attribute‐oriented deterministic inversion method is proposed. The method is similar to most inversion algorithms and consists of two parts: modelling and inversion. The innovation of the model lies in extracting seismic attributes and learning the mapping between the seismic attributes of the well‐side and the parameters to be inverted based on the support vector regression (SVR) algorithm. Then, the mapping relationship is used to realize the modelling of the parameters to be inverted in the well‐free area. Under the constraints of this model, seismic inversion is implemented through the Markov Chain Monte Carlo (MCMC) approach, yielding inversion results that exhibit strong consistency with the corresponding seismic responses. As the multi‐trace structural attributes contain more high‐frequency information, the resolution of the parameters to be inverted based on this model is also higher. We continue to conduct inversion tests using post‐stack seismic data. The results show that seismic attribute‐oriented inversion has a significant advantage in inversion resolution over partially deterministic inversion algorithms (model‐based inversion, sparse spike inversion, etc.).

We present deep learning (DL) networks for three‐dimensional (3D) joint inversion of active seismic full waveform and passive seismic traveltime data to image reservoirs and their properties and quantify imaging uncertainties. Active seismic full‐waveform data can provide high‐resolution monitoring images but are collected only intermittently because of their high acquisition cost. In contrast, passive seismic data can be gathered at relatively low cost between regular active surveys, although their imaging quality can be compromised by factors such as low signal‐to‐noise ratios and limited ray coverage of the target. Although these datasets are routinely acquired together at CO2 storage sites, their combined inversion within a 3D DL framework has not been previously demonstrated. To our knowledge, this is the first study to address this gap, combining the strength of both data types. For efficient data storage and DL training with large 3D seismic datasets, we use a 3D data matrix in which a random number of passive seismic traveltime data are stored as parabolic envelopes using one‐hot encoding and a 3D full‐waveform data matrix in which multiple shot gathers are summed. Two network architectures are evaluated: a single‐encoder U‐Net for single‐data type inversion and a dual‐encoder U‐Net for joint inversion of active and passive seismic data. We also evaluate the single‐encoder U‐Net for joint inversion by concatenating full‐waveform data and traveltime data. We propose a systematic approach for selecting an optimal dropout rate that balances regularization during training and Monte Carlo dropout‐based uncertainty quantification during prediction by examining the correlation coefficient between standard deviation and prediction error, along with the training misfit, across a range of dropout rates. 3D DL inversion experiments include five different network configurations, with evaluations under ideal, noisy and dropout‐enabled conditions. Both model and data uncertainties are assessed, as well as their combined effects. Across all conditions, the networks consistently predict accurate CO2 saturation models with low prediction errors, such as a structural similarity index of 0.993 and CO2 difference of 1.1%. Uncertainty estimates show strong spatial correlation with prediction errors, confirming the effectiveness of the proposed dropout selection approach. The results demonstrate that our DL approach, utilizing compact data representations and appropriate uncertainty quantification, yields accurate subsurface images under various inversion conditions and provides valuable insights into the reliability of predictions.

Multiples are generally considered coherent noise in conventional seismic migration. If they are not appropriately separated or eliminated from the observed reflection data, it can result in significant artefacts of the migration image, which will adversely affect subsequent structural interpretation and reservoir description. Inspired by the state‐of‐the‐art model‐driven deep learning methods, we have proposed a self‐supervised deep‐learning approach for multiple elimination. The proposed method contains two parts. The first one is that we use the conventional multiple prediction approach to predict the initial surface‐related multiple reflections. The second part is to build a multiple‐elimination model based on a deep neural network. In the deep neural network, the input is set as the predicted initial surface‐related multiples from a conventional method and the label is set as the observed reflection data with primaries and multiples. Therefore, the proposed approach is a kind of self‐supervised deep‐learning multiple‐elimination model. The deep neural network component of our proposed approach can be interpreted as a corrector in conventional methods that performs amplitude and phase correction on predicted multiples. Moreover, we combine the advantages of L1 and L2 loss functions and introduce the attention mechanism to improve the inversion efficiency and accuracy of self‐supervised deep networks. Through some experiments with synthesized and field data, we demonstrate that the proposed self‐supervised deep‐learning approach can effectively and efficiently eliminate multiples from the observed data. It excels in both accuracy and efficiency compared to traditional method.

Enhancing seismic data resolution is a crucial step for geological interpretation and imaging. Deep learning‐driven resolution enhancement primarily depends on sophisticated network architectures and extensive datasets. A lightweight seismic super‐resolution model based on contrastive learning and knowledge distillation is proposed. Knowledge distillation is implemented by training a compact student network to mimic a powerful teacher model, thereby reducing reliance on extensive datasets and complex architectures. Contrastive learning is leveraged to align the bottleneck features encoded from the teacher network with the ones from the student network across different noisy inputs. The student network's total loss comprises a supervised loss with ground‐truth labels, a distillation loss with the teacher's pseudo‐labels and a feature‐matching loss derived from the bottleneck features of both networks. The comparative experiments were conducted on four field datasets and 3200 pairs of slices extracted from 800 pairs of synthetic three‐dimensional seismic cubes. Experimental results demonstrate that the proposed model achieves similar to or better performance than the comparison models for noise suppression and weak signal recovery, even with only parameters and training data compared to the reference model.

We propose a novel physics‐guided deep learning framework for geophysical inversion that integrates Langevin Monte Carlo (LMC) sampling to quantify uncertainties in model parameters. A statistical sampling strategy is employed to enhance computational efficiency by reducing the number of required samples while preserving diversity and informativeness. The training data for the supervised learning networks are iteratively expanded with outputs from a stochastic sampler and their corresponding observed responses, ensuring representative coverage of the model space. The Jensen–Shannon divergence is adopted as the loss function for training the network model, in which the Gaussian assumption is applied to enable analytical computation. The developed workflow is evaluated on reservoir porosity inversion, where it successfully reconstructs porosity patterns in the subsurface, yielding results that closely match the reference model. Compared to traditional LMC algorithm applied to the entire data cube, the proposed approach attains substantial computational efficiency by leveraging an active learning strategy that identifies and utilizes a limited yet representative subset of the observations. The results demonstrate the effectiveness of the proposed method, highlighting its potential for application to a wide range of geophysical inverse problems.

Seismic inversion quantitatively extracts reservoir properties from seismic data, which has gained increasing attention in assisting the exploration and evaluation of deep coalbed methane (DCBM) reservoirs. However, the accuracy of seismic prediction is limited because the DCBM reservoirs exhibit complex pore geometries governed by a dual‐porosity system. To address this limitation, the study presents a dual‐porosity parameter seismic inversion method based on decoupled equivalent medium theory. A dual‐porosity rock physics model is constructed and then decoupled to derive a linear forward operator that links matrix porosity, crack porosity and crack aspect ratio to the corresponding elastic parameters. To account for lithological variability, a Gaussian mixture model is employed to describe the joint prior probability distribution of dual‐porosity parameters. Well‐log data are applied to invert matrix porosity, crack porosity and crack aspect ratio, which serve as prior constraints in the iterative Bayesian inversion framework, thereby enhancing the stability and accuracy of the forward operator. By explicitly treating dual‐porosity parameters as inversion targets, the proposed method effectively captures the spatial heterogeneity of pore geometries in DCBM reservoirs. Borehole‐side synthetic seismic gather validation results demonstrate that the proposed approach significantly enhances inversion accuracy compared with conventional equivalent‐porosity inversion methods. The application to pre‐stack seismic data demonstrates the ability of the method to capture the dual‐porosity geometry.

Characterization of cracks is a key issue in shale oil and gas that have become increasingly important in the hydrocarbon industry. Seismic exploration is frequently employed for the characterization of cracks in shale reservoirs. However, the accurate interpretation of seismic data for characterizing cracks in shale reservoirs remains a significant challenge, primarily due to an insufficient understanding of how subsurface pressure affects the anisotropic elastic properties of cracked shales. To address this knowledge gap, this study systematically investigates the effects of confining pressure on the anisotropic elastic properties of cracked artificial shales, with a specific focus on decoupling the distinct roles of background porosity and crack porosity. The five anisotropic elastic velocities were measured on manufactured shale samples with varying crack and background porosity, respectively, and the corresponding anisotropic parameters, Young's moduli and Poisson's ratios were derived as a function of confining pressure. The results demonstrate that the influence of crack porosity on reducing the velocities and on enhancing the elastic anisotropy is significantly more pronounced than that of background porosity. Notably, the velocities across the cracks, Vp(0°) and Vsh(0°), exhibit the greatest sensitivity to pressure changes, especially in samples with high crack porosity. Consequently, all the anisotropic parameters reduce exponentially with increasing confining pressure, with the reduction being most significant in shales with either the lowest background porosity or the highest crack porosity. The pressure‐dependent geomechanical properties (Young's moduli and Poisson's ratios) reveal that the direction parallel to cracks remains the most favourable path for hydraulic fracturing, particularly under low confining pressure and in rocks with high crack porosity. These findings provide critical insights for improving the quantitative interpretation of seismic data for characterizing cracks and for optimizing hydraulic fracturing design in shale reservoirs.