Geophysical Prospecting - Volume 73, Issue 4, 2025

Amplitude versus offset attribute inversion primarily utilizes the change in amplitude with offset to extract lithologic information of the reservoir. We performed high resolution imaging using three‐dimensional seismic data and simulated four different models of gas hydrate and free gas to optimize the selection of sensitive attributes for gas hydrate characterization. We optimized the selection of sensitive attributes for gas hydrate characterization. Our research identified the gradient parameter G as a highly sensitive attribute for characterizing gas hydrate reservoirs. By comparing theoretical models with drilling site data, we predicted the saturation variations of gas hydrate based on the amplitude of G and summarized the amplitude versus offset characteristics and sensitive attributes of gas hydrate at different saturation levels. The results offer valuable insights for identification of gas hydrate in seismic data and provide a reference for their characterization.

In seismic exploration, particularly within the domain of oil and gas reservoirs, the accurate imaging of complex fault blocks and the identification of structural traps are important. Geological risk factors, including the implementation of structural traps, reservoir delineation, and precise target drilling, require immediate attention in practical exploration. Addressing these factors involves two primary challenges: ensuring imaging accuracy and minimizing structural distortions. This study introduces a high‐resolution velocity modelling technique with fault control, specifically developed to mitigate misties between seismic image and well‐log data and improve the accuracy of seismic depth imaging and well depth correlation. The method offers a targeted solution to the challenges of implementing structural traps, delineating reservoirs and executing precise drilling operations. By incorporating fault control, it accounts for the structural complexity of subsurface media, enabling an accurate inversion of velocity variations across fault blocks. This approach ensures that velocity models, constrained by geological and structural models, exhibit a high degree of consistency. Utilizing fault‐controlled travel time inversion, the method resolves mistier between seismic imaging and well‐log data, guaranteeing the precision of velocity models and imaging. The methodology provides reliable seismic data for target evaluation, effectively reducing exploration risks and improving the accuracy of velocity modelling.

Source time‐reversal imaging based on wave equation theory can achieve high‐precision source location in complex geological models. For the time‐reversal imaging method, the imaging condition is critical to the location accuracy and imaging resolution. The most commonly used imaging condition in time‐reversal imaging is the scalar cross correlation imaging condition. However, scalar cross‐correlation imaging condition removes the directional information of the wavefield through modulus operations to avoid the direct dot product of mutually orthogonal P‐ and S‐waves, preventing the imaging condition from leveraging the wavefield propagation direction to suppress imaging artefacts. We previously tackled this issue by substituting the imaging wavefield with the energy current density vectors of the decoupled wavefield, albeit at the cost of increased computational and storage demands. To balance artifact suppression with reduced computational and memory overhead, this work introduces the Poynting and polarization vectors mixed imaging condition. Poynting and polarization vectors mixed imaging condition utilizes the polarization and propagation direction information of the wavefield by directly dot multiplying the undecoupled velocity polarization vector with the Poynting vector, eliminating the need for P‐ and S‐wave decoupling or additional memory. Compared with scalar cross‐correlation imaging condition, this imaging condition can accurately image data with lower signal‐to‐noise ratios. Its performance is generally consistent with previous work but offers higher computational efficiency and lower memory usage. Synthetic data tests on the half‐space model and the three‐dimensional Marmousi model demonstrate the effectiveness of this method in suppressing imaging artefacts, as well as its efficiency and ease of implementation.

We compare classical and adjoint‐state first‐arrival tomography approaches in subsurface model reconstruction, focusing on pre‐salt updates with a circular‐shot ocean bottom node geometry. The investigation demonstrates that, whereas conventional tomography has faster convergence and better alignment with observed data, it produces significant noise artefacts and fails to adequately represent the reservoir section. In contrast, adjoint‐state tomography provides superior model reconstruction by taking into account the complete travel time volume, significantly lowering noise and boosting reservoir imaging despite its higher computational cost. A quantitative investigation of root mean squared errors for ultra‐long offsets confirms the efficacy of adjoint‐state tomography in minimizing data misfit and improving model fidelity. The findings emphasize the potential of adjoint‐state tomography in enhancing subsurface imaging and underscore the limits of conventional tomography in handling complex subsurface details with sparse acquisition geometry.

The pore structure of carbonate rocks is both complex and highly heterogeneous. Accurately assessing acoustic anisotropy is crucial for analysing and predicting the properties of carbonate reservoirs. Numerous experimental studies have investigated the acoustic anisotropy of carbonate rocks, and various fracture detection techniques have been developed. However, these studies have not adequately addressed the impact of the nonuniform distribution of rock pore structures on acoustic anisotropy. The pore structure of computed tomography scanning images of carbonate core can be obtained by using digital image processing techniques, and a method for evaluating pore distribution heterogeneity based on the box‐counting method of fractal theory was proposed. A numerical simulation of P‐wave azimuthal anisotropy was conducted, and the relationship between the pore distribution heterogeneity index and acoustic anisotropy parameters was analysed. This novel evaluation method for acoustic anisotropy provides a theoretical basis for predicting parameters in carbonate reservoirs.

Rock‐physics modelling provides theoretical basis for predicting elastic and anisotropy parameters from petrophysical properties. However, shale rocks usually develop complex pore structures, wherein isolated and connected pores or cracks may coexist. Conventional methods that assume either isolated or connected pores have limited applicability to shale reservoirs. To this end, this work proposes a shale rock‐physics modelling method to address pore complexities. In specific, the proposed method combines inclusion‐based and Brown–Korringa models to consider both isolated and connected pores in shales. Connected‐porosity coefficient is introduced in the modelling to balance the effects of the two pore types. To better handle pore complexities and improve modelling accuracy, the coefficient and pore aspect ratio are jointly estimated from measured vertical P‐ and S‐wave velocities with a global optimization algorithm. Numerical analysis is performed to analyse the general effects of connectivity and pore geometry on elastic properties of shales. The proposed method is applied to a well data from the Longmaxi shale reservoir in southwest China. The method is also compared with two other methods to show its capability of predicting elastic properties with satisfactory accuracy. The estimated connected‐porosity coefficient also facilitates the characterization of velocity anisotropy to some degree.

The resonance of coal reservoir induced by external excitation is a kind of environment‐friendly stimulation technology. In this work, an experimental system was established to carry out the coal resonance fracturing experiments, and the change trend of fracture structure was investigated by computed tomography scanning. The results showed that under the excitation of vibration within the resonance frequency band range, the coal fracture will gradually expand to failure. There are three forms of coal fracture expansion under the condition of resonant, which are the formation of slip zones in coal, the formation of fracture at phase interface and the formation of ‘void’. When the excitation frequency is constant, the natural frequency of coal decreases gradually with the expansion of fractures, resulting in the vibration frequency gradually deviating from the natural frequency of coal, and the fracture propagation behaviour of vibrating coal gradually changes from divergence to convergence.

The interlayer mesoscopic wave‐induced fluid flow and the squirt flow are two important mechanisms for seismic attenuation and dispersion in the fluid‐saturated porous layered rock. Although numerous studies have been conducted on these two mechanisms, their combined effects (especially the resulting frequency‐dependent anisotropy features) have not been sufficiently investigated. Hence, we propose a concise and rigorous theoretical model to quantify the combined effects of these two mechanisms. We first quantify the squirt flow effects through a wet rock frame for each layer that has frequency‐dependent and complex‐valued elastic properties. Then, we apply Biot's quasi‐static poroelasticity theory to derive the analytical solutions for the effective stiffness coefficients of the periodically layered rock. Using the derived rock stiffness coefficients, we calculate the seismic attenuation and dispersion, as well as the frequency‐dependent anisotropy. Two cases are studied, one with alternating water‐ and gas‐saturated layers (constant rock frame properties) and the other with periodically distributed fracture layers (constant saturating fluid properties). The P‐waves in these two cases are both influenced by the mesoscopic interlayer wave‐induced fluid flow and the squirt flow. However, the SV‐wave is solely affected by the squirt flow in the first case and primarily influenced by the mesoscopic interlayer wave‐induced fluid flow in the second case, respectively. The wave velocity and attenuation in the first case are isotropic, whereas those in the second case exhibit frequency‐dependent anisotropy (induced by the mesoscopic interlayer wave‐induced fluid flow). To validate our model, we compare our model to the measured extensional attenuation in a partially saturated sandstone sample under different effective pressures. The joint effects of the mesoscopic interlayer wave‐induced fluid flow and the squirt flow observed in the experiments are well predicted by our model. Our model has potential applications in the seismic characterization of reservoirs composed of layered rocks, such as shale reservoirs.

Estimating structure of pyroclastic deposits plays an important role in the interpretation of volcanic geology and evaluation of potential hazard. We aim to invert near surface seismic data to produce high‐resolution images of pyroclastic density current deposits resulting from the 18 May 1980 volcanic eruption at Mount St. Helens, Washington, USA. Elastic full waveform inversion is a popular data fitting method used to estimate seismic properties of the earth. Due to the great challenges in the convergence of elastic full waveform inversion when inverting high‐frequency and complex near‐surface land seismic data, we develop a specific workflow to improve the resolution of velocity models that progress from traveltime inversion and surface wave inversion to full elastic full waveform inversion. The final inverted models include fine‐scale feature structures that compare favourably to an adjacent outcrop. The final data fitting shows significant improvement with normalized waveform misfit decreasing from 1 to 0.4.

A thermoacoustics imaging system is investigated in this paper to enhance conventional imaging modalities in geological subsurface situational awareness applications. While thermoacoustics imaging has traditionally been used in biological scenarios like breast cancer detection, this work aims to extend thermoacoustics imaging to geophysical applications by demonstrating that water‐saturated sand can be distinguished from dry and oil‐saturated sand based on their amplitude differences. This breakthrough enables the feasibility of monitoring water distribution in these media. Moreover, to compensate for the low conversion efficiency from electromagnetic power to thermoacoustics amplitude, the signal modulation method is used by applying the frequency‐modulated continuous wave techniques. The experiment results show that the frequency‐modulated continuous wave can enhance the signal‐to‐noise ratio while maintaining a similar resolution as the pulse‐excited thermoacoustics wave. These findings pave the way for the future use of thermoacousticsimaging in subsurface sensing and imaging of fluid flow and transport in porous media.

Passive seismic tomography plays a significant role in monitoring subsurface structures and properties during hydraulic fracturing. In this study, we develop a new passive seismic tomography approach to jointly invert for event locations, 3D P‐wave velocity (Vp) and Poisson's ratio models, for downhole microseismic monitoring. The method enables to directly obtain the 3D Poisson's ratio or Vp/Vs ratio without the assumption of identical P‐ and S‐wave raypaths. The back azimuths of passive seismic events are incorporated into the proposed method to better constrain the event locations. The 3D cross gradients are further applied to the proposed method to assimilate the P‐wave velocity model with Poisson's ratio model in the same geological structure. The synthetic experiment demonstrates that the proposed tomographic method can recover the event locations and their adjacent 3D P‐wave velocity as well as Poisson's ratio models effectively. In the field experiment, microseismic events are relocated reasonably well compared with the grid search solutions in a calibrated layer model. The area with low Poisson's ratios may be utilized to estimate the stimulated reservoir volume and indicate a potential area associated with highly saturated hydrocarbon.

Multiple attenuation is an important step in seismic data processing, leading to improved imaging and interpretation. Radon‐based algorithms are commonly used for discriminating primaries and multiples in common depth point seismic gathers. This process implies a large number of parameters that need to be optimized for a satisfactory result. Moreover, Radon‐based approaches sometimes present challenges in discriminating primaries and multiples with similar moveouts. Deep learning, based on convolutional neural networks, has recently shown promising results in seismic processing tasks that could mitigate the challenges of conventional methods. In this work, we detail how to train convolutional neural networks with only synthetic seismic data for assessing the demultiple problem in field datasets. We compare different training strategies for multiples removal based on different loss functions. We evaluate the performance of the different strategies on 400 clean and noisy synthetic data. We found that training a convolutional neural network to predict the multiples and then subtracting them from the input image is the most effective strategy for demultiple, especially for noisy data. Finally, we test our model to predict multiples on an elastic synthetic dataset and four distinctive field datasets. Our proposed approach reports successful generalization capabilities predicting and eliminating internal and surface‐related multiples before and after migration while mitigating Radon challenges and relieving the user from any manual tasks. As a result, our effectively trained models bring a new valuable tool for seismic demultiple to consider in existing processing workflows.

Neelam–Heera fields are located in Mumbai offshore basin and are significant contributors to India's oil and gas production. These mature oil and gas fields are producing mainly from Eocene–Oligocene carbonate reservoirs. Key challenges of Neelam–Heera fields include seismic imaging issues in areas perturbed by shallow gas, discrepancy in time and depth structures caused by lateral and vertical velocity variations and delineation of karst/discontinuities in the Neelam–Bassein formation to improve well planning and avoid mud loss. The seismic interpretation study we conducted used recent ocean bottom node seismic data to provide new insights into these challenges and are discussed in this paper. An integrated study scaling from interpretive processing to high‐resolution mapping/attribute analysis was conducted to solve the key challenges of the Neelam–Heera fields. A significant improvement is seen in the imaging of the Neelam–Heera area because of prestack depth migration processing of ocean bottom node data, which led to detailed mapping and better understanding of various formations. One of the key challenges was to understand the structural changes in areas masked by shallow gas clouds in the Heera high area. The PP data could not provide clear imaging of areas masked by shallow gas; however, the PS data is not affected by fluid presence and provided imaging below the gas cloud. The gas‐bearing Bandra formation in the crestal part of the Heera field is mapped with the help of PS data. The Neelam–Heera area shows prominent trends related to a velocity anomaly at a shallow level (dissolution features), forming the valleys in the time cube that are not seen in the depth cube. The key formations are mapped carefully with the integration of a depth cube. The new depth maps provided insights in areas hindered by previous false structures and are key to future development planning. To gain insights on karstification and mud‐loss issues in Neelam field (Bassein formation), various structural attributes were derived and correlated with well data. Most positive/negative curvature attributes at different azimuthal stacks alongside variance maps provided detailed high‐resolution insights on faults/lineaments/fractures, and posting of well mud‐loss information validated the disturbed‐zone‐related mud‐loss areas. The availability of ocean bottom node data with high‐quality processing and imaging and subsequent detailed seismic interpretation using PP and PS data provided new insights to solve the key challenges of Neelam–Heera fields. The key outcome of this study includes an improved understanding of areas perturbed by shallow gas and better depth maps for integration into the future development model in areas with discrepancies caused by significant lateral and vertical velocity variations. In addition, we were better able to delineate karst/discontinuities in the Neelam–Bassein formation, which improved well planning and helped avoid mud loss.

Electrical resistivity tomography is a suitable technique for non‐invasive monitoring of municipal solid waste landfills, but accurate sensitivity analysis is necessary to evaluate the effectiveness and reliability of geoelectrical investigations and to properly design data acquisition. Typically, a thin high‐resistivity membrane is placed underneath the waste to prevent leakage of leachate. In the construction of a numerical framework for sensitivity computation, taking into account the actual dimensions of the electrodes and, in particular, of the membrane, can lead to extremely high computational costs. In this work, we present a novel approach for numerically computing sensitivity effectively by adopting a mixed‐dimensional framework, where the membrane is approximated as a two‐dimensional object and the electrodes as one‐dimensional objects. The code is first validated against analytical expressions for simple four‐electrode arrays and a homogeneous medium. Then it is tested in simplified landfill models, where a two‐dimensional box‐shaped liner separates the landfill body from the surrounding media, and 48 electrodes are used. The results show that electrodes arranged linearly along both sides of the perimeter edges of the box‐shaped liner are promising for detecting liner damage, with sensitivity increasing by 2 to 3 orders of magnitude, even for damage as small as one‐sixth of the electrode spacing in diameter. Good results are also obtained when simulating an electrical connection between the landfill and the surrounding media that is not due to liner damage. The configurations with the highest sensitivity directly beneath the liner are quadrupoles in which both the current and voltage dipoles have one electrode inside the liner and one electrode outside, and a two‐dimensional arrangement of the electrodes. The modelled sensitivity values beneath the liner are close to a minimum sensitivity threshold derived from arbitrary and simplified assumptions. We believe that direct current surveys have the potential to detect liner damage using electrode spreads positioned along the liner perimeter, both inside and outside the landfill. However, down‐scaled laboratory tests will be necessary to validate the modelling results and confirm whether the computed sensitivity values are sufficiently high to reliably detect liner damage.

The deep coalbed methane development is in its initial stage, with limited research on elastic anisotropy of coals in deep coalbed methane reservoirs. To study the elastic anisotropy of coals in deep coalbed methane reservoirs, three sets of cylindrical primary coals are collected from the relatively shallow area of the Taiyuan 8^# coal formation, a key site for deep coalbed methane extraction. The microscopic observation, basic physical property test, ultrasonic velocity measurement and theoretical modelling are constructed to study the pressure sensitivity of elastic properties and the factors affecting the elastic anisotropy of coal samples. The velocity increases rapidly with confining pressure increasing at confining pressures below 13 MPa but relatively stabilizes at higher pressures. The velocity perpendicular to the bedding plane is more sensitive to pressure than that parallel to the bedding plane. As confining pressure increases, the velocity anisotropy decreases but remains noticeable at the highest pressure. Based on the microstructure and ultrasonic experiment results, an anisotropic rock physics model for the deep coalbed methane reservoir is constructed to quantitatively analyse the effects of the clay and organic matter and pore structure on the elastic anisotropy of coals. The modelling analysis indicates that P‐ and S‐wave velocity anisotropies decrease as the degree of preferred orientation of organic matter and clay decreases, which increases with the increase of organic matter and clay content. The equivalent pore aspect ratio and lamination index of studied coals are obtained from a practical inversion scheme based on the proposed model. The results can support the anisotropic rock physics inversion of deep coalbed methane reservoirs and the accurate prediction of engineering sweet spot parameters.

This study introduces the application of the 2D weighted compact gravity inversion technique to model crustal thickness and intracrustal discontinuities in the Aegean region, encompassing both marine and terrestrial areas over a significant area of 430 km × 333 km. The method utilized advanced spectral analysis and upward continuation techniques to enhance the quality of Bouguer data, effectively mitigating surface noise. The findings reveal a remarkable correlation (over 99.9%) between observed and theoretical data, demonstrating the algorithm's robustness in accurately delineating crustal features. Depth estimates for the Conrad (2.7 g/cm³) and Moho (3.3 g/cm³) discontinuities were obtained, highlighting distinct density variations across discontinuity zones. Furthermore, the relationship between intracrustal discontinuities and seismicity was examined, revealing that earthquakes predominantly follow the Conrad boundary. Notably, this study uniquely produces 2D depth contour maps of Conrad and Moho discontinuities from specific density derived from gravity inversion sections. The results indicate that this method is a valuable tool for understanding crustal dynamics, suggesting potential applications for future tectonic assessments, especially in regional studies. The successful application of this novel technique emphasizes its significance in advancing geophysical modelling and enhances our understanding of isostatic evaluation in the Aegean region for further studies.

The gravity force, as a vector, has both magnitude (length) and direction. Previously, with the measurement methods employed, only the magnitude (length) of the vector was measured; there was no possibility in geophysical practice to measure the direction of the vector. Previously, the method of astronomic positioning was used to determine the vertical direction. This was extremely time‐consuming, taking several months, complex and expensive procedure, used in geodesy at some points to determine the values of vertical deflection. Recently, revolutionary changes have been made in this area, with the development of a new, quick and simple method for determining the local vertical direction, which is more accurate than ever before. In contrast to the previous extremely slow process, with the computer‐controlled, fully automated QDaedalus measurement system, a single person can perform multiple high‐precision measurements at various locations overnight, making the appropriate quantity and quality of measurement results suitable for geophysical structural research purposes. In this article, we briefly describe the operation of the QDaedalus system and draw attention to the geophysical applicability of the method in terms of detectable mass anomalies and useful station spacing using astrogeodetic field measurements. To demonstrate this, astrogeodetic measurements were also carried out along a 4 km long section, from which the deflection of the vertical values and geoid heights were determined. Our measurements were compared with the normal values calculated with the global gravity model of high resolution (Global Gravity Model plus). As the deflection of the vertical values calculated with this model includes the effect of the surface topographic masses at the resolution of the model, the difference between the measured and the modelled values is practically a function of subsurface density anomalies. At the same time, we performed model calculations showing how density anomalies of different extents assumed at different depths affect the value of the vertical deflections and the geoid anomalies.