Affected by the fluvial-deltaic sedimentary system, the Jurassic coal seam roof sandstone in the Ordos Basin is characterized by rapid lithological changes and high argillaceous content, resulting in great difficulties in the fine prediction of its water abundance. To address this problem, taking the roof sandstone aquifer of No.3 coal seam in Bayangaole Mine, Inner Mongolia as an example, based on logging data, the Simandoux formula was used to quantitatively analyze the water saturation distribution of the roof sandstone; the quantitative prediction results of resistivity logging were verified through λρ-μρ (product of Lame constant λ and density ρ - product of shear modulus μ and density ρ) crossplot analysis; integrating sandstone thickness, porosity and water saturation, a water abundance evaluation index was constructed; and on the basis of variogram analysis, the sequential Gaussian simulation (SGS) method was adopted to predict the planar water abundance distribution of the roof sandstone. The results show that the roof sandstone of No. 3 coal seam is generally characterized by high water saturation, large thickness and high porosity, but exhibits strong vertical and horizontal heterogeneity; the λρ-μρ crossplot can effectively distinguish the differences between sandstone porosity and water saturation, thereby indirectly verifying the reliability of the water abundance prediction results from resistivity logging; the water-rich areas of the roof sandstone are mainly distributed in the northwest and northeast of the study area, which are in good agreement with the low-resistivity anomaly areas predicted by the transient electromagnetic method. Compared with the traditional Archie formula, the Simandoux formula is more applicable in predicting the water saturation of Jurassic coal seam roof sandstone. Therefore, the water abundance evaluation method based on the combination of resistivity and acoustic logging is not only a useful supplement to the existing technologies, but also provides an effective approach to solving such evaluation problems, and can offer reliable technical support for groundwater resource protection and mine roof water hazard prevention and control.

Gallium (Ga) in coal is a nationally emerging strategic mineral resource, yet research on using petrophysical methods to detect the spatial variation in critical metals in coal seams remains limited. Analyzing the distribution characteristics of Ga-rich coal using geophysical well-logging methods is of great significance for the development and utilization of Ga. This study introduces a quantitative method for predicting Ga-rich laminations in ultra-thick bituminous coal seams by integrating: (i) wireline-log-based lithofacies classification, (ii) lithofacies-constrained mineral inversion, and (iii) lithofacies-constrained and laboratory-established Ga-mineral correlations. The coal seam was first classified into four distinct lithofacies types-(i) parting, (ii) medium-ash coal (MA), (iii) low-ash coal (LA), and (iv) extra-low-ash coal (ELA)-through integration of conventional wireline log interpretation, cluster analysis, and XGBoost machine learning. Second, lithofacies-constrained Ga-host mineral associations were established by integrating core sample analysis, correlation analysis, and linear regression modeling. Third, mineral content predictions for each lithofacies were obtained through wireline-log-based mineral inversion, constrained by petrophysical boundaries. Finally, prediction uncertainties were evaluated using Markov Chain Monte Carlo (MCMC) simulation, while Ga-rich laminations were predicted by integrating log-derived mineral inversion results with regressed Ga prediction models. The results demonstrate strong agreement between mineral inversion and XRD analyses within uncertainty ranges, achieving a prediction accuracy of 73.6% for Ga. This validated methodology presents a novel approach for quantifying Ga concentrations in coal, as demonstrated through a case study.

Water in roof sandstone constitutes a significant concealed geological hazard, threatening the safe and efficient mining of coal resources. Geophysical prospecting is a key technical means for advanced detection of such hazards. Conventional electrical methods, while advantageous in detecting low-resistivity anomalies, are limited by their inability to identify complex structures, as well as by constrained investigation depth and spatial resolution. Meanwhile, traditional seismic approaches often overlook complex fluid-related dissipation mechanisms within strata. To address these limitations, this study proposes an integrated quantitative evaluation framework based on frequency-dependent rock physics modeling and seismic dispersion attribute inversion. First, a seismic rock physics model applicable to roof sandstone aquifers in Permian coal-bearing strata was established by integrating effective medium theory with Chapman's multiscale fracture model. Second, Sobol global sensitivity analysis was employed to quantify the contributions of key physical parameters to P-wave velocity (VP) and inverse quality factor (1/Q). Subsequently, a forward model with four horizontal layers was constructed to comprehensively simulate reflection and interference effects in viscoelastic media, thereby establishing a theoretical relationship between water saturation (SW) and the dispersion attribute (Dp). The forward modeling employed the propagator matrix method, using a 45 Hz Ricker wavelet as the source. In practical applications, high-resolution seismic spectra were obtained through the smoothed pseudo-Wigner-Ville distribution (SPWVD) combined with spectral balancing, and the dispersion attributes in the roof sandstone of the No. 4 coal seam were extracted using a frequency-scanning strategy. Sensitivity analysis indicates that, within the seismic frequency band, variations in VP are primarily controlled by porosity, whereas 1/Q exhibits pronounced sensitivity to water saturation and fracture density. Seismic forward modeling further reveals a positive correlation between dispersion attributes and water saturation. Field results demonstrate that the spatial distribution of Dp attribute exhibits strong lateral heterogeneity, and Dp values show a significant positive correlation with hourly water yield measured at hydrogeological boreholes. These results confirm that the proposed approach enables the effective evaluation of sandstone aquifer water abundance using stacked seismic data, providing robust support for roof water hazard prevention. Moreover, for CO2 geological storage (CCS) applications, the developed frequency-dependent rock-physics framework offers valuable constraints for reservoir characterization, site evaluation, and dynamic monitoring.

The development of weathered and oxidized zones in coal seams seriously affects the mining of shallow coal seams. Aiming at the economical and efficient detection of weathered and oxidized zones in shallow coal seams, taking the detection of the weathered and oxidized zone of shallow-buried No. 6 coal in the Ningdong Mining Area as an example, this paper analyzes the feasibility of detecting coal seam weathered and oxidized zones by using the noise autocorrelation method, and proposes a delineation method for weathered and oxidized zones based on the subarray-assumed noise autocorrelation technique. The research results show that the superposition of cross-correlation functions within subarrays can significantly suppress strong surface wave interference in passive seismic records and enhance the body wave reflection signals of target strata. The reflected waves of No.6 coal reconstructed from passive noise signals are highly consistent with the geological structures and the lower boundary of the weathered and oxidized zone of No.6 coal in the study area, and the scope of the weathered and oxidized zone can be delineated according to waveform variations. Compared with seismic profiles in low-frequency and high-frequency bands, the waveform difference caused by the weathered and oxidized zone is the most significant in the medium-frequency band (10-40 Hz). The lower boundary of the weathered and oxidized zone in the study area is distributed between the coal seam contour lines of 1 240 m and 1 260 m, being shallower in the east and deeper in the west, which is consistent with the exposure data of two boreholes in the study area. The Yuanyanghu-Fengjigou anticline exerts a remarkable influence on the development of the weathered and oxidized zone, and the boundary of the weathered and oxidized zone in the anticline core is near the 1 250 m contour line. Compared with the geological inference results, the measured weathered and oxidized zone is approximately 100 m shallower on the west side and 30-50 m deeper on the east side, with a minimum distance of about 160 m from the southern boundary of the work area. The noise autocorrelation method based on the subarray assumption can economically and efficiently detect weathered and oxidized zones in shallow-buried coal seams, and is suitable for wide popularization and application.

Although Mercury Intrusion Porosimetry (MIP) and Nuclear Magnetic Resonance (NMR) are widely used for pore characterization, their effectiveness is fundamentally constrained by theoretical limitations. This study investigated the pore structure characteristics of coal-bearing sandstones from the northeastern Ordos Basin using an integrated approach combining experimental measurements and model-based inversion. The experimental measurements comprised a stress-dependent acoustic velocity test (P- and S-wave velocities), X-ray diffraction (XRD) mineralogical analysis, and NMR relaxation T2 spectra characterization. For model-based inversion, we developed an improved Mori-Tanaka (M-T) theoretical framework incorporating stress-sensitive pore geometry parameters and dual-porosity (stiff/soft) microstructure representation. Systematic analysis revealed four key findings: (1) excellent agreement between model-inverted and NMR-derived total porosity, with a maximum absolute error of 1.09%; (2) strong correlation between soft porosity and the third peak of T2 relaxation spectra; (3) stiff porosity governed by brittle mineral content (quartz and calcite), while soft porosity showing significant correlation with clay mineral abundance and Poisson’s ratio; and (4) markedly lower elastic moduli (28.78%–51.85%) in Zhiluo Formation sandstone compared to Yan’an Formation equivalents, resulting from differential diagenetic alteration despite comparable depositional settings. The proposed methodology advances conventional NMR analysis by simultaneously quantifying both pore geometry parameters (e.g., aspect ratios) and the stiff-to-soft pore distribution spectra. This established framework provides a robust characterization of the pore architecture in Jurassic sandstones, yielding deeper insights into sandstone pore evolution within the Ordos Basin. These findings provide actionable insights for water hazard mitigation and geological CO2 storage practices.

Detection and identification of faults are crucial in the process of coal exploration and mining, and the traditional manual method of fault interpretation can no longer meet the needs of actual production, and the deep learning-based seismic fault interpretation method performs better in the field of fault segmentation. Conventional convolutional neural network (CNN) has limited sensory field and cannot make good use of the global information, which will lead to some predicted faults with insufficient continuity and missing faults, etc. Transformer has the advantage of extracting global information, and introduces the TransUNet network which is a fusion of CNN and Transformer to construct a CBAM- based seismic fault identification method. TransUNet seismic fault identification method to identify 2D seismic fault images. Firstly, the CBAM-Block attention module is integrated into the TransUNet network, and the module is added into the CNN tomography encoder part and the 3-layer jump connection part connecting the tomography encoder and the tomography decoder, respectively, to enhance the recognition ability of the seismic tomography image from two dimensions, namely, the channel and the space; secondly, the loss function optimised jointly by the Dice loss function and the cross-entropy loss function is selected to make the segmentation of the tomography image more accurate. function and cross-entropy loss function to make the fault image segmentation more accurate, and the DICE and IOU values obtained by the CBAM-TransUNet fault identification network on the synthetic seismic dataset are increased to 0.84 and 0.75, respectively, and the experimental results show that the continuity of the fault identification is stronger, which is obviously superior to other classical segmentation methods; finally, the constructed model is used to interpret the faults on the real seismic dataset of the F3 block of the North Sea, off the coast of the Netherlands. Finally, the constructed model was used to interpret the faults in the real seismic data set of Block F3 in the North Sea off the Netherlands. The experimental results show that the seismic fault identification method based on CBAM-TransUNet can effectively identify the faults while removing the redundant fault information, and performs well in terms of fault identification accuracy and fault identification continuity, and the identified faults are richer in details, which improves the accuracy of fault identification, and can be effectively applied to identify the faults in the real seismic data.

ObjectiveThe identification and assessment of sandstone aquifers in coal seam roofs play a vital role in the safe mining of mines. Investigating the rock physics and amplitude variation with incidence and azimuth (AVAz) responses of these sandstone aquifers is critical to the prevention and control of water hazards in mines. MethodsBy integrating rock physical models Voigt-Reuss-Hill (VRH), differential equivalent medium (DEM), Hudson, and Wood, as well as Gassmann’s anisotropic fluid substitution theory, this study proposed a rock physics modeling method for fractured sandstones of the horizontal transversely isotropic (HTI) media type (hereafter referred to as HTI sandstones). Using this method, this study explored the impacts of fracture parameters and water saturation on seismic rock physical responses of the sandstones. Accordingly, it constructed a two-layer theoretical forward model, calculated the reflection coefficients of HTI sandstones, and analyzed the relationships of the reflection coefficients with the fracture density and water saturation. Results and Conclusions The rock physics modeling results indicate that a higher fracture density corresponded to lower compressional and shear wave (also referred to as P- and S-wave) velocities and stronger anisotropy. As the water saturation increased, the P-wave velocity decreased initially and then increased, whereas the S-wave velocity decreased slightly. Concurrently, with an increase in the water saturation, anisotropy coefficients \begin{document}$ {\varepsilon ^{({\text{v}})}} $\end{document} and \begin{document}$ {\delta ^{({\text{v}})}} $\end{document} increased, while anisotropy coefficient \begin{document}$ {\gamma ^{({\text{v}})}} $\end{document}remained unchanged. The AVAz forward modeling results indicate that a higher fracture density was associated with more pronounced azimuthal anisotropy of the sandstones’ reflection coefficients and larger differences in P-wave reflection coefficients between saturated and dry sandstones. The saturation state of sandstones was the most distinguishable in the case where the angle of incidence and azimuth were 40° and 0°, respectively. Indicators for sensitivity to the fracture density and water saturation of sandstone aquifers included the fitted slope and intercept of the AVAz curves, as well as the isotropic and anisotropic components of the amplitude versus offset (AVO) gradients. The results of this study provide a theoretical basis for the identification and assessment of sandstone aquifers.

Time–depth conversion is a crucial step in 3D seismic interpretation of coalfields. Fast and accurate time–depth conversion is essential for ensuring safe and efficient coal production. However, conventional methods often struggle to balance accuracy with efficiency, which makes it difficult to achieve good application results in the coalfield. To address this problem, we proposed a new coal seam time–depth conversion method based on machine learning and seismic velocity inversion. Firstly, a high-precision time-domain layer of the coal seam floor was obtained. Subsequently, the average velocity of the coal seam floor was calculated from boreholes. Following this, post-stack seismic inversion was performed to obtain velocity volumes, and the velocity volumes were subjected to median filtering. Next, machine learning models were trained using the average velocity of the coal seam floor, extracted from the inverted velocity volumes processed with different median filter windows, and two-way travel times of the coal seam floor as inputs, with actual coal seam floor elevations as the outputs. Finally, different machine learning methods and conventional methods were compared and analyzed for time–depth conversion in coalfield. The results indicate that the Bayesian-SVR model achieved the highest accuracy in time–depth conversion, with a maximum absolute error of only 2.86 m and a mean absolute error of 1.79 m at verification boreholes. In summary, this study introduces a machine learning-based coal seam time–depth conversion method that does not require complex velocity models, enhancing efficiency while maintaining high accuracy, which holds significant importance for advancing intelligent coal mining and achieving transparent working faces.

Petrophysical properties are critical for shale gas reservoir characterization and simulation. The Wufeng-Longmaxi shale, in the south-eastern margin of the Sichuan Basin, is identified as a complex reservoir due to its variability in lithification and geological mechanisms. Thus, determining its characteristics is challenging. Based on wireline logs and pressure data analysis, a shale reservoir was identified, and petrophysical properties were described to obtain parameters to build a reservoir simulation model. The properties include shale volume, sand porosity, net reservoir thickness, total and effective porosities, and water saturation. Total and effective porosities were calculated using density method. Shale volume was estimated by applying the Clavier equation to gamma-ray responses. Sand porosity and net reservoir thickness were evaluated using the Thomas-Stieber model, and the Simandoux equation was used to compute water saturation. The results indicate that the reservoir is characterized by a relatively low porosity and high shale content, with shale unequally distributed in its laminated form (approximately 75%), dispersed (about 20%), and structural form (5%). This research workflow can efficiently evaluate shale reservoir parameters and provide a reliable approach for future reservoir development and fracture identification.

Quantifying the seismic attenuation of wave propagation in the earth's ' s interior is essential for studying subsurface structures. Previous approaches for attenuation simulations (e.g., the standard linear solid and the fractional derivative model) are mainly based on the frequency-independent quality factor Q assumption. However, seismic attenuation in high-temperature and high-pressure regions usually exhibits power-law frequency-dependent Q characteristics. To simulate this Q effect in attenuative media, we derive a new viscoacoustic wave equation with decoupled fractional Laplacians in the time domain. Unlike the existing methods using relaxation functions to fit the power-law relationship in a specific frequency band, our equation is directly derived from the approximated complex modulus, which explicitly involves the reference quality factor and fractional exponent parameters. Furthermore, this equation contains two fractional Laplacians, which can easily simulate decoupled amplitude dissipation and phase distortion effects, making it amenable to Q-compensated reverse time migration. In the implementation, a Taylor-series expansion and a pseudospectral method are introduced to solve the fractional Laplacians with variable fractional exponents. Numerical experiments demonstrate the effectiveness of our method for power-law frequency-dependent Q simulations. As a forward-modeling engine, our derived viscoacoustic wave equation is a good supplement to the current Q simulation methods and it could be applied in many seismic applications, such as Q-compensated reverse time migration and full-waveform inversion.

Seismic attenuation is a basic physical property of the earth, which significantly affects the characteristics of seismic wavefields. Accurately simulating wave propagation in the earth is essential to image subsurface structures. Some prevailing methods (e.g., the standard linear solid and fractional Laplacian equation) to describe seismic wave propagation in attenuating media are mainly based on the constant- Q model (CQM), which is valid at room temperature and pressure. However, laboratory measurements suggest that the quality factor Q is a function of frequencies in some regions. To simulate the frequency-dependent Q effect, we derive a viscoacoustic wave equation from the stress-strain relationship of the fractional Zener model (FZM) with variable fractional orders. During the implementation, we separate the real and imaginary parts of the modulus and introduce a low-rank decomposition method to solve the FZM equation. Because the amplitude dissipation and phase dispersion are decoupled, we establish a compensated reverse time migration ( Q-RTM) algorithm to mitigate adverse effects caused by seismic attenuation and improve the quality of seismic migration in frequency-dependent attenuating media. A two-layer model and the BP gas chimney model are used to perform Q-RTM tests. A low-pass filter with a Tukey window function is applied to suppress numerical instability during the compensation. Numerical results demonstrate that our FZM Q-RTM approach can produce high-resolution images with corrected reflector positions and amplitudes. Because the CQM equation ignores the frequency dependence of Q, it may lead to overcompensation in Q-RTM.

The seismic attributes of water-rich sandstone contain much information about the rock's physical properties and seismic wave parameters. They are commonly used to predict the rock's physical properties (e.g. porosity). However, the seismic attributes of water-rich sandstone are affected by porosity, water saturation and thickness. To eliminate the influence of thickness on the porosity prediction of water-rich sandstone and improve the accuracy of the porosity prediction, the authors propose a Lambert W-R transform method to isolate the contribution of thickness and porosity from seismic attributes. First, a rock physical model is used to calculate the equivalent elastic parameters of water-rich sandstones with different porosity values and water saturation levels. Second, the seismic attribute dataset of water-rich sandstone is established by forward modelling the seismic response of the wedge-shaped water-rich sandstone model, and the selection of sensitive physical properties is completed. Then, the transformation parameters (zeta R-Ah(s) and eta R-Ah(s)) are obtained by Lambert W-R transformation, which is exponentially related to instantaneous amplitude. zeta R-Ah(s) and eta AhRs are sensitive to thickness and porosity, respectively. Finally, an interpretative template for porosity prediction of water-rich sandstone is established by cross-plot analysis (zeta R-Ah(s) and eta R-Ah(s)) and verified by a practical case. The verification results show that the porosity predicted by the interpretation template is consistent with drilling fluid consumption. However, it is lower than the porosity of logging constrained P-wave impedance inversion.

Integrating petrophysical and geomechanical parameters is an efficient approach to evaluating shale gas reservoir potential. The high cost of corings and their limited number, coupled with time-intensive investigation, led researchers to use this alternative combination approach. In the Jiaoshiba area, from single-pilot well core data and log measurements, petrophysical and geomechanical parameters such as shale volume, total organic carbon, gas content, as well as pore pressure, stress components, and mineral brittleness were first estimated using established methods. In the second phase, based on logging curves, the reservoir electro-facies (EF) classification was performed using the unsupervised multi-resolution graph-based clustering method on a series of twenty wells, identifying five EF with different intrinsic characteristics. Unsupervised analyses were developed using the multilayer artificial neural network while incorporating the K-nearest neighbors and graphical classification algorithms. The results from the first and second phases indicate reservoir richness in organic matter, with the best reservoir exhibited by EF2 and EF3. In addition, effective stress components (SV, SH, and Sh) evaluation shows a normal stress regime with hydraulic fracture systems perpendicular to the minimum horizontal stress at each measured depth of the reservoir (Sv > SH > Sh). This research workflow can efficiently evaluate shale reservoirs with a realistic approach for identifying favorable fracturing positions while reducing errors due to human interference.

Seismic wave propagation inside the earth is usually accompanied by amplitude dissipation and velocity dispersion. It is essential to accurately characterize these effects and make the corresponding compensations in seismic reverse-time migration (RTM) and the following interpretation. Due to the approximation of omega approximate to kv0 in the derivation, the conventional fractional viscoacoustic wave equation in strongly attenuative media is unsatisfactory. This paper develops a modified viscoacoustic wave equation by combining the relationship between angular frequency and complex wavenumber. The numerical simulation demonstrates its advantage of high accuracy in describing constant-Q effects, especially in strongly attenuative media with small Q values. A truncated Taylor-series expansion algorithm and a pseudospectral method are developed to solve this equation during wave propagation. To address the issue of numerical instability in Q-compensated RTM (Q-RTM), a stabilized technology with regularization is further developed. Unlike the conventional filtering method implemented in the wavenumber domain, our approach only adds an explicit regularization term, which can be directly computed in the space domain. Furthermore, this regularization depends on velocities and quality factors, significantly improving its applicability in complicated structures. Noise-free data tests demonstrate that our Q-RTM method can effectively correct phase distortion, compensate for amplitude attenuation, and produce high-resolution migrated images without any cutoff-frequency artifacts caused by filtering. Noisy data experiments further verify the antinoise performance and robustness of our method.

ABSTRACT Velocity and attenuation (Q) anisotropy are widely distributed in the Earth’s interior, significantly affecting the kinematic and dynamic characteristics of seismic-wave propagations. Previous studies developed to simulate these effects are mainly restricted to the constant-Q assumption. However, seismic attenuation in high-temperature and high-pressure regions is demonstrated to be frequency-dependent and usually follows a power-law formulation. To simulate this Q effect in transversely isotropic (TI) attenuating media, we derive a new pure-viscoacoustic wave equation with decoupled fractional Laplacians, which can simultaneously simulate amplitude dissipation and velocity dispersion effects. Based on the wavenumber relationship between the observation and physical coordinate systems, the tilted TI (TTI) wave equation is further derived. Compared with the pseudoviscoacoustic wave equation, the proposed pure-viscoacoustic equation can simulate stable P wavefields in complex geological structures without S-wave artifacts. To solve this new equation, two low-rank decompositions are introduced to approximate the real and imaginary parts and avoid the separation of wavenumbers and dip angles, making it much simpler in programming and implementation. We further use this equation to perform Q-compensated reverse-time migration to generate high-resolution migration images in anisotropic attenuating media. Numerical examples demonstrate the effectiveness of the proposed method for pure-viscoacoustic wavefield simulations and migrations in TTI attenuating media with power-law frequency-dependent Q effects.

This paper presents the zonal geochemistry and elasticity characteristics of gallium- and lithium-rich No. 6 coalbed in the Haerwusu mine and discusses interpretation methodologies of coal-hosted gallium and lithium resources using lab-measured samples and field-measured wireline logs. The results demonstrate that both coal-composition-based and elastic-parameter-based classifications yield similar results, categorizing the coalbed into subzones related to coal quality. Material compositions, elastic properties, critical metals, and host minerals exhibit zonal distribution characteristics within the ultrathick No. 6 coalbed. Three-class classifications significantly enhance correlations among host minerals, elastic parameters, and critical metals, albeit with differing trends among classes. In classes II and III (ultralow- and low-ash-yield coals), boehmite and kaolinite primarily host gallium and lithium, respectively. In class I (medium-ash-yield coal), gallium is associated with kaolinite, while lithium lacks specific mineral associations. Constrained by wireline logs, a rock physics modeling strategy is proposed to link mesoscale coal compositions to macroscale elastic responses. Moreover, explicit correlations between host minerals and critical metals are established, connecting macroscale elastic responses to microscale gallium and lithium enrichments and exploring interpretation methods of coal-hosted critical metals. Preferred lithium interpretation methods include compositional ternary plots and elastic parameter cross plots, while preferred gallium interpretation methods involve boehmite-gallium and elastic parameter-gallium fitting. These findings may contribute to understanding the enrichment mechanisms and interpretation technologies of coal-hosted critical metals in ultrathick low-rank coalbeds.

In order to further improve the identification accuracy of small-scale faults in seismic interpretation, Bayesian optimized extreme gradient boosting (XGBoost) model was constructed to recognize small-scale faults across coalbeds using reduced seismic attributes based on the theory of information value(IV). Firstly, the seismic attribute data of the mining area were preprocessed to remove abnormal samples and large noise samples. Secondly, chi-square bins were performed for each feature of the processed model, the weight of evidence (WOE) was calculated in each container, and the information value of each element was obtained, which is used as the importance of each feature. Features with low information values were reduced to remove high-noise feature attributes. At the same time,a certain degree of noise is added to the seismic data of small-scale faults to enhance the anti-noise ability of the model. Finally, the Bayesian optimized XGBoost model was constructed. The method to improve the XGBoost objective function was proposed to balance the training weights of the positive and negative examples. As the acquisition function of the Bayesian optimized algorithm quickly falls into the local optimum, it does not easily balance the “exploit” and “explore” approach. Therefore, this paper proposes an adaptive balance factor change algorithm, which dynamically ground balances the process of “mining” and “exploring” the pi acquisition function to improve the robustness of the parameter optimization process. Comparing the identification outcomes, the new XGBoost model framework (SAPI-Bay-ImpXGBoost) has a higher prediction accuracy than BP neural network, Support Vector Machine(SVM), K-nearst neighbors(KNN) and Adaptive Boosting(AdaBoost). In summary, the proposed method can further strengthen the identification of small-scale faults in coal mining areas.

煤层精准定位是无人采煤的关键技术,煤层厚度预测是煤田地震资料解释的重要研究内容之一.参考实际地层厚度及物性参数,构建含楔形煤层的正演模型,通过地震剖面正演和地震属性提取、优化,对比分析信噪比和多种回归方法对煤层厚度预测的影响.研究结果表明:部分地震属性与煤厚相关性较强,可以用于煤厚预测;地震属性间的信息冗余不可忽略,但基于主成分分析和多维标度的地震属性优化结果无本质区别;当信噪比较低(10 dB)时,随机森林回归算法的均方根误差最小(1.07),支持向量机回归算法的误差居中(1.15),多元线性回归算法的误差最大(1.84);当信噪比较高(25 dB)时,支持向量机回归算法的误差最小(0.05),随机森林回归算法的误差居中(0.11),多元线性回归算法的误差最大(0.20);输入数据信噪比对煤厚预测有明显影响,信噪比越高、预测效果越好.基于地震属性优化及支持向量机回归的煤厚预测方法,是实现薄煤层厚度高精度解释的一种有效途径.

The acoustic behavior in fluid attenuating media can be effectively simulated using a fractional Zener model (FZM). Because of the fractional time derivatives of both stress and strain in the constitutive relationship, this mechanism is very realistic and flexible in describing seismic attenuation. However, using conventional FZM wave equations to propagate seismic waves requires storing large amounts of previous wavefield information to calculate the fractional time derivatives, which is unacceptable in practice. In this paper, we derive a new time-domain viscoacoustic wave equation in the framework of the FZM. This new equation does not contain any fractional time derivatives; thus, it is more economical in computational costs. Furthermore, the amplitude attenuation and phase dispersion effects are separated in the newly proposed equation, which is very favorable to compensate for energy loss and correct phase dispersion in reverse-time migration. To improve the accuracy, we incorporate a wave number (k)-space operator into the decoupled FZM wave equation to compensate for temporal dispersion errors caused by the second-order finite-difference discretization. Therefore, a high-temporal-accuracy viscoacoustic wave equation is derived to simulate nearly constant-Q wavefields in attenuating media. In the implementation, a low-rank decomposition method is introduced to solve the mixed-domain operators. Numerical analysis and modeling results demonstrate the effectiveness and applicability of the proposed method for simulating the decoupled viscoacoustic wavefield with high accuracy.

Previous studies demonstrated that seismic attenuation and anisotropy can significantly affect the kinematic and dynamic characteristics of wavefields. If these effects are not incorporated into seismic migration, the resolution of the imaging results will be reduced. Considering the anisotropy of velocity and attenuation, we derive a new pure-viscoacoustic wave equation to simulate P wave propagation in transversely isotropic (TI) attenuating media by combining the complex dispersion relation and modified complex modulus. Compared to the conventional complex modulus, the modified modulus is derived from the optimized relationship between angular frequency and wavenumber, which can improve the modeling accuracy in strongly attenuating media. Wavefield comparisons illustrate that our pure-viscoacoustic wave equation can simulate stable P wavefields in complex geological structures without S-wave artifacts and generate similar P wave information to the pseudo-viscoacoustic wave equation. During the implementation, we introduce two low-rank decompositions to approximate the real and imaginary parts and then use the pseudo-spectral method to solve this new equation. Since the proposed equation can simulate decoupled amplitude attenuation and phase dispersion effects, it is used to perform Q-compensated reverse-time migration (Q-RTM). Numerical examples demonstrate the accuracy and robustness of the proposed method for pure-viscoacoustic wavefield simulations and migration imaging in transversely isotropic attenuating media.