Suppressing random noise is imperative to recover effective signals from microseismic data. However, the weak energy and poor frequency-domain concentration pose a fundamental challenge on the signal, leading conventional denoising to suffer from significant distortion or residual noise. To address this, this study focuses on a specific space by utilizing the second-order spectrum in the complex time-frequency domain and pioneers a novel noise suppression method via sparse decomposition within the space. By incorporating the complex frequency-domain orthogonal decomposition into the classical two-stage Fourier-transform framework, the complex time-frequency second-order spectrum is directly constructed from the single-sideband real or imaginary part of the signal. Microseismic signal spectra are decoupled in the designed space, facilitating targeted noise suppression solely in the amplitude domain. The signals also maintain their time-domain energy concentration without sidelobes, yielding a highly sparse representation that stands in sharp contrast to the dispersed, low-rank structure of random noise. Given this structural disparity, the robust principal component analysis (RPCA) algorithm enables high-precision separation of the sparse signal from the low-rank noise. Theoretical analysis and experimental data demonstrate that the proposed method achieves an average signal-to-noise ratio (SNR) improvement of approximately 28 dB under strong background noise. Compared to traditional methods such as wavelet threshold denoising, it provides an SNR enhancement of over 50%. Moreover, the method maintains good adaptability across diverse noise models and multisignal scenarios. This study offers a novel strategy for extracting microseismic signals under strong background noise, thereby supporting advances in deep resource exploration and safety monitoring.

ObjectiveThe conventional inversion of surface direct-current (DC) resistivity data faces challenges such as limited exploration accuracy and depth. This study aims to enhance the accuracy and reliability of exploration in complex engineering and geological settings through the joint inversion of borehole-surface DC resistivity data. MethodsFirst, different geoelectric models were constructed through three-dimensional (3D) forward modeling of full-space DC resistivity. Then, the potential response characteristics under surface and borehole observation methods were systematically analyzed to determine the differences in detection sensitivity between surface and borehole observations. Accordingly, a joint 3D inversion technique for borehole-surface DC resistivity data was developed using the Gauss-Newton method. Finally, the effectiveness and superiority of the joint inversion technique were verified using theoretical models and actual data. Results and Conclusions The inversion of surface observations yielded high lateral resolution but a limited vertical resolution, while the inversion of borehole observations exhibited a high vertical resolution. In contrast, the joint inversion of borehole-surface data provided more accurate resistivity structures and physical property values. The effectiveness of the joint inversion technique was verified using both theoretical models and actual data. The verification results demonstrate that the joint inversion technique combines the advantages of the surface and borehole observation methods, thereby enhancing both the spatial resolution of inversion results and the capability for identifying deep anomalous body. Therefore, the developed joint inversion technique provides technical support for improving the exploration accuracy in complex engineering and geological settings.

The remanent magnetization reflects crucial geoscientific information about the Earth’s internal structure, formation environment, and evolutionary processes. It is extensively employed in resource exploration and deep geological investigations, making it a central focus in both magnetic exploration and rock magnetism. However, remanent magnetization alters the total magnetization direction of rocks, posing a significant challenge in obtaining the magnetization distribution when the magnetization direction is unknown. In this study, we propose a sparse inversion method for magnetic anomaly amplitude based on nonconvex optimization. By transforming magnetic anomalies into their amplitude form, the influence of the magnetization direction is mitigated. We construct the inversion objective function using the L0 norm as a sparsity constraint, enabling the recovery of magnetization models with clear boundaries while also achieving scale sensitivity and control within the inversion process. The objective function incorporating the L0 norm constraint represents a typical NP-hard problem, making stable optimization a challenge task. To address this, we employ an improved iterative hard thresholding method that ensures both a good fit to the magnetic anomaly amplitude and asymptotically linear convergence of the L0 norm of the magnetization matrix. Furthermore, we introduce an expected model scale as the stopping criterion during iterations, which allows us to obtain magnetization models that conform to the expected scale. This method not only facilitates the reconstruction of magnetization models in scenarios where remanent magnetization is present, but also effectively controls the volume of the inversion model without relying on predefined physical property bounds.

The integrated fusion and inversion of seismic data and marine controlled-source electromagnetic (MCSEM) data can identify the gas hydrate distributions. However, due to differences in observation systems and scales between seismic and MCSEM data, current fusion methods have failed to effectively address the critical issue of physical property variation within gas hydrate reservoirs. This research seeks to consolidate the two datasets into a cohesive observational framework. By transforming the MCSEM data into low-frequency constraints applicable to seismic impedance inversion, it is possible to realize an effective integrative interpretation that combines both seismic and MCSEM data. Using a 3 km-long seismic dataset and MCSEM data from the Shenhu Sea area in the South China Sea as a case study, we apply the Poisson blending algorithm to integrate seismic and MCSEM data, enabling precise identification and characterization of gas hydrate reservoirs and underlying gas-bearing fluids. The gas hydrate saturation results, derived from seismic inversion constrained by MCSEM data, demonstrate strong consistency with well logging and geological interpretation. This concordance validates the efficacy of the integrated fusion and inversion methodology and highlights its advantages in accurately predicting the spatial distribution of gas hydrate enrichment. The developmental positions of deep gas-bearing fluid pathways, coupled with the fault locations within the free gas zone and gas hydrate-bearing layer, play significant roles in the heterogeneous enrichment of gas hydrates. This research provides important technical and theoretical support for the precise and efficient prediction of gas hydrate reservoirs.

The magnetotelluric method has large depths of investigation and can provide important structural information in many exploration problems. The one-dimensional magnetotelluric inversion also has been applied to extract boundary information and provide the constraints for the interpretation of complementary datasets. Traditional smooth inversion based on the L(2 )norm only provides a single smooth model that it is difficult to detect the location of the geological boundary. Trans-dimensional inversion provides an effective means to determine the boundaries with uncertainty quantification but incurs significant computational costs. We present an efficient method to detect distinct interfaces from one-dimensional blocky magnetotelluric inversions using an Ekblom norm. The method leverages the Ekblom norm to assess the change in the recovered resistivity model with the threshold parameter as a means to delineate the significant boundaries in the subsurface. The threshold parameter specific to the Ekblom-norm inversion is then used to probe the variability of the inversion to obtain a more robust interface detection. Once the interfaces are detected, we calculate the average resistivity value between detected interfaces to form a final conductivity model. As a demonstration, we apply this method to a synthetic example and the field data from East Tennant in Australia. The results show that the method is effective in obtaining boundaries.

As a rapidly developed seismic observation technology, distributed acoustic sensing (DAS) has shown great potential in many fields such as near surface wave imaging and oil and gas exploration. Taking a landfill site as an example, this research uses DAS seismic data to carry out the study of near surface wave imaging based on DAS. By analyzing the spectral characteristics of DAS ambient noise data, an ambient noise data screening method based on power spectral density (PSD) spectrum is proposed to obtain high-quality virtual shot gather and dispersion spectrum images. A joint active and passive dispersion curve (DC) inversion flow is designed to obtain the shear wave velocity profile. The result indicates the low-velocity layer at the top of the landfill site and the transverse discontinuous interface inside are found. At the same time, the resistivity profile is obtained by inversion of the high-density resistivity data collected from the landfill site, which shows the high-resistivity region on the surface and the transverse discontinuous interface of the internal resistivity. The spatial positions of the abnormal regions in the inversion profiles of the two physical property parameters are very consistent, which proves the effectiveness of the proposed method. Our research can lay a foundation for structural inspection of landfill based on DAS.

We present a study of marine controlled-source electromagnetic (CSEM) inversion using joint constraints based on structural and petrophysical properties. Marine CSEM can directly evaluate the oil and gas distribution within a reservoir and improve drilling success rates. However, marine CSEM inversion suffers from significant nonuniqueness. Constrained inversion is an effective method to reduce nonuniqueness and improve resolution. There are two main approaches to constrained inversion: one involves applying constraints based on subsurface structural information, and the other relies on known petrophysical information. Current constrained inversion techniques typically utilize only one of these methods in isolation. In this study, we develop a joint constraint method that combines structural constraints with petrophysical property constraints for marine CSEM inversion. First, we obtain a high-precision velocity structure using seismic full-waveform inversion and introduce the velocity structure constraint into the objective function of marine CSEM inversion using a cross-gradient function. This enables marine CSEM inversion based on seismic velocity structure constraints. In addition, during the iterative inversion process, we apply petrophysical property constraints using the fuzzy C-means (FCMs) clustering method. The joint constraint method integrates the advantages of both types of constraints, effectively improving inversion accuracy. Synthetic model analyses and field data applications demonstrate that marine CSEM inversion, guided by joint constraints from seismic velocity structures and petrophysical properties, not only enhances the imaging accuracy of resistivity structures but also recovers realistic resistivity properties. Furthermore, the FCMs clustering method sharpens the boundaries of inversion results, making them clearer and more interpretable for geological analysis.

The distribution of water in the mantle plays a critical role in deep earth dynamic processes, including plate subduction, earthquake generation, and magmatic activity. Seismic tomography studies have consistently shown that the subducted Pacific plate beneath eastern China stagnates within the mantle transition zone. The release of water from the slab can substantially modify the surrounding electrical properties, producing pronounced high-conductivity anomalies. Resolving the regional conductivity structure is therefore essential for constraining the geometry and dehydration state of the stagnant slab, as well as for improving our understanding of the strong seismicity and widespread Cenozoic volcanism in eastern China. Conventional magnetotelluric (MT) surveys are constrained by their frequency range and cannot resolve deep structures on thousand-kilometer scales. Geomagnetic depth sounding (GDS), which utilizes ultra-long-period signals (>100 days), can image the mantle down to ∼1,600 km, making it a powerful approach for investigating mantle water content and partial melting. Here, we compile long-term geomagnetic records from an array covering eastern China and apply a three-dimensional unstructured finite-element GDS modeling and inversion scheme. The inversion employs a limited-memory quasi-Newton optimization strategy to improve computational efficiency and model stability. The resulting electrical resistivity model delineates the mantle transition zone and lower mantle beneath eastern China, clearly imaging the stagnant Pacific slab and indicating possible dehydration and melting processes. These findings offer new constraints and insights into the structure and dynamics of the deep earth.

SUMMARY Distributed acoustic sensing (DAS) enables high-density sampling of seismic wavefields at low cost compared to conventional geophones. This capability facilitates structural detection of a municipal solid waste (MSW) landfill, which is important for protecting the surrounding ecosystem. However, processing the vast amount of data from DAS array for ambient noise imaging can be computationally intensive. To address this, we employed the common-midpoint two-station (CMP-TS) analysis to enhance the efficiency of ambient noise imaging in the MSW landfill. CMP-TS analysis involves selecting pairs of traces at equal distances on both sides with the subarray midpoint as symmetry, which reduces the number of DAS array recordings for cross-correlation calculations. After positioning the DAS arrays linearly on top of the MSW landfill to automatically collect ambient noise, we used the CMP-TS analysis in the cross-correlation calculations to speed up the measurement of dispersion. The S-wave velocity structure of the study region was obtained quickly by inverting the extracted dispersion curves using the gradient optimization method. Ambient noise imaging based on CMP-TS analysis with DAS was applied to a test of an area-type MSW landfill. The resulting S-wave velocity section revealed a discontinuous low-velocity zone, validated by the high-density resistivity method. This low-velocity zone was interpreted as containing leachate from waste decomposition, and its discontinuity may be caused by excessive differences in the waste residues settling rates under compaction. Employing CMP-TS analysis in ambient noise data collected by DAS offers more cost-effective monitoring and a reliable basis for environmental pollution prevention and control.

The magnetotelluric (MT) method plays an important role in the exploration of oil and gas, geothermal resources, and deep structures due to its large exploration depth and low cost. MT data inversion is a key step in data interpretation. The geophysical inversion based on a deep learning (DL) method does not need to calculate the sensitivity matrix, and the inversion results do not depend on the initial model, which has also received much attention. The feature extraction capability of a conventional convolutional neural network (CNN) is limited. The self-attention mechanism can select the data information with relatively large contribution to the result from a lot of input information and improve the ability of key feature extraction. We present a 2-D MT CNN inversion study based on the self-attention mechanism. We first use a blocky 1-D MT inversion to quickly build the representative sample model, and then we perform a 2-D MT forward algorithm by finite element method to build the training set samples. We introduce the self-attention mechanism into the CNN and place the self-attention mechanism behind the convolutional layer. After the data are extracted from the convolutional layer, feature extraction is performed again to improve the reliability and efficiency of the prediction. We tested the algorithm using different geoelectric models and field data from the Dayangshu Basin. Our results show that the inversion accuracy is improved, especially the inversion results of field data are consistent with borehole data.

The basin environment is a widely studied subject in both geology and geophysics for its economic significance in energy and mineral explorations. However, the estimation of the basement depth is often a challenging task given the complexity of the basement relief and lateral physical property change. Previous works simplify the problem by only inverting for the depth to the basement, and more recent studies have suggested the need to incorporate the variation of physical properties to improve basement structure imaging. In this study, we develop an inversion method with the associated workflow to simultaneously recover both the depth to a magnetic basement and a laterally varying magnetic susceptibility in the basement rock. To achieve this, we employ a set of constraints on the inverse problem. Particularly, both the recovered susceptibility and basement depth models are bounded below a possible maximum value, and the depth model is guided by a few depth points obtained from the resistivity models that are obtained from the one-dimensional blocky inversions of magnetotelluric (MT) data. In addition, we apply the fuzzy C-means (FCM) clustering to the susceptibility model during the inversion and use the inverted cluster centers to differentiate for different geological units in the basement. To show the effectiveness of our work, we compare the existing approaches and our method using two test inversions on one synthetic model resembling the basin-basement environment before demonstrating our method on a field data example with magnetic data collected by the U.S. Geological Survey (USGS) over the Illinois Basin. Our results show improved recovery in both basement relief and susceptibility in the basement rock, and inversion with field data is able to identify three different susceptibility zones in basement rock below the Illinois Basin.

Based on the superposition principle for potential field of Green's equivalent layers and frequency cutoff filtering by successive layer optimization, a new imaging method of normalized downward continuation of gravity and magnetic data by successive layer optimization (NDCSLO) is developed. Owing to noise interference is unavoidable in field gravity and magnetic data, the affect of noise interference is studied by NDCSLO tests with sphere and slab model data of different noise intensity, measurement grid spacing and anomaly strength. Ii is shown that the most affected interval is the first 1-2 grid spacing of continuation depth. Too small or big grid spacing will result in field energy dissipation and spurious imaging. With increasing source physical property and thus increasing anomaly strength, the signal-noise ratio will be enhanced and benefit improving the resolving power and accuracy of the downward continuation imaging. From NDCSLO test of field gravity and magnetic data in Southwest Sichuan, it is verified that the affect of measurement error is mainly on the first 2km of the continuation depth, and the middle to deep continuation imaging is less influenced. Therefore, it is suggested that if the field data is apparently affected by noise disturbance, a priori noise filtering be used before NDCSLO; otherwise the NDCSLO can be utilized directly, although the continuation imaging of less than the first 2km depth is not recommended to interpret qualitatively or quantitatively.

井中重力矢量测量可以近距离感知地下目标体,获得地质体不同方向的重力异常特征,提高地下介质的纵向分辨率.这里首先基于点元法实现了井中重力矢量正演,通过模型正演计算,分析了井中三分量重力异常响应特征,然后基于相关性搜索的黄金分割算法实现了井中重力矢量联合反演,通过模型反演分析了井中重力不同分量数据联合反演的效果,验证了反演方法的可靠性.结果表明,井中三分量重力异常从不同方面反映了异常体分布特征,不同分量的峰值变化对异常体位置分布有较好的指示作用.反演中井中重力三分量同时联合反演效果最好,针对复杂模型,单井观测信息有限,多井多分量联合反演可以获得准确的反演结果.

Magnetotelluric (MT) data can provide important structural information in many exploration problems. We present an efficient method to detect distinct interfaces from 1D blocky magnetotelluric (MT) inversions using an Ekblom norm and to quantify the associated uncertainty. The method leverages the Ekblom norm to assess the change in the recovered resistivity model with the threshold parameter as a means to delineate the significant boundaries in the subsurface. The threshold parameter is then used as a means to probe the inversion and quantify the uncertainty for use in subsequent interpretation. Once the interfaces are detected, we calculate the average resistivity value between detected interfaces based on inversion results using different Ekblom parameters to form a final model. As a demonstration, we apply this method to the East Tennant MT data from Australia and construct the regolith thickness and major interfaces.

Magnetotelluric (MT) inversion algorithms are an essential tool in exploration geophysics because they provide us with resistivity models of the subsurface. Inverse problems are ill-posed and non-unique. Resistivity models produced by geologically unconstrained MT inversions are typically smooth, which means that the geological interpretations obtained based on the resistivity models will be ambiguous and uncertain. To solve the above problems, geophysicists have made many efforts over the years to introduce any prior information (such as geological structure information and physical property data) into models, which could help to better constrain the inverted model with the inversion algorithms. In the last few years, there has been renewed interest in machine learning techniques. Various machine learning methods have been applied to inversions. Using the fuzzy c-means (FCM) clustering method, geophysicists proposed new inversion algorithms that are capable of building petrophysical information into the inversion. Although the FCM has been used and advanced in previous work, geophysicists concluded that a priori information of the correct number and value of cluster centers is very important for a successful FCM inversion. In fact, it is difficult to obtain the appropriate clustering information in some geophysical survey areas. In this study, we present an effective way to build the petrophysical information for the MT inversion based on FCM clustering algorithm. When the actual petrophysical information is insufficient, considering the characteristics of MT data, we perform a one-dimensional blocky inversion that can clearly identify the interface with distinct electrical property contrast; then, we obtain the number and value of cluster centers from 1D blocky inversion for the further MT inversion with FCM clustering. The algorithm uses guided FCM clustering to improve the model within the iterative minimization during MT inversion. In our MT inversion method, we integrate the geophysical inversion and geological differentiation into a unified scheme, which interact and enhance each other. The resistivity models obtained from the inversion respect the geophysical and petrophysical data and are easy to geologically interpret. We tested the algorithm using two synthetic examples and a field data example.

随着待勘探目标地质体越来越复杂且埋深增大,单一地球物理勘探方法的片面性和局限性 日益突显,综合利用多种地球物理技术及相应数据已成为现今必然趋势.为此,提出基于多元地质统计学的交叉—变差函数建立速度与电阻率之间的岩石物理关系,并在此基础上利用机器学习中的引导模糊C均值聚类算法进行基于岩石物理关系的多重约束反演,实现电—震联合建模.大杨树盆地南部坳陷实际资料的应用结果表明,该电—震联合建模约束反演可逐步降低单一地球物理方法的多解性,提高对目标地质体的识别能力.非地震与地震方法所得结果相互印证,展示了该联合建模约束反演技术具有良好应用潜力.

针对井的数量和分布空间有限,磁法三维反演存在计算量大,反演多解性严重的问题,开展基于模糊C均值聚类约束的井-地磁法联合数据空间反演.采用数据空间反演方法,提高三维反演计算效率,然后在反演中引入模糊C均值聚类对反演迭代结果进行物性约束,提高反演结果可靠性.通过理论模型和冬瓜山矿区模型进行反演分析.结果表明:数据空间反演能够减少反演计算时间;井-地磁异常联合反演综合了地面磁测和井中磁测的优点,提高反演结果的空间分辨率;基于模糊C均值聚类约束反演使反演物性值更接近真实值,并且反演异常体边界清晰.

We have used the integrated interpretation of gravity, magnetotelluric (MT) data, and seismic data to improve the structural imaging of the Dayangshu Basin. The Dayangshu Basin is mainly composed of clastic and volcanic rocks. The logging data in the basin show different degrees of direct hydrocarbon indication, suggesting that the Dayangshu Basin has good potential for exploration. However, the widely distributed volcanic rocks attenuate seismic waves and lead to poor seismic imaging. Thus, the seismic signal is weak in the Ganhe Formation (K1g) and reliable seismic images cannot be obtained below that formation. MT data can accurately obtain images of deep structures because the resistivity of volcanic rocks is significantly higher than that of sedimentary rocks. Therefore, to obtain a more reliable geologic model, we combine the traditional 3D MT inversion result with logging and seismic data to establish an initial model. The 3D MT fuzzy constrained inversion (FCI) produces a more reliable geophysical model and geologically meaningful results. The resistivity model inverted from FCI shows that volcanic rocks are widely distributed in the Ganhe Formation, and the resistivity value of the lower section of the Longjiang Formation is greater than that of the upper section of the Longjiang Formation. Finally, the 3D gravity inversion with structural constraints from 3D MT FCI method was performed to improve the model resolution in depth and to highlight the density variations within the Jiufengshan Formation, which can further optimize the geologic model. We have determined how the effective integration of gravity, MT, and seismic data can improve the structural imaging of the Dayangshu Basin.

We carried out a multigeophysical data joint interpretation to image volcanic units in an area where seismic imaging is difficult due to complicated and variable volcanic lithology. The gravity and magnetic methods can be effective in imaging the volcanic units because volcanic rocks are often strongly magnetic and have large density contrasts. Gravity and magnetic data have good lateral resolution, but they are faced with challenges in defining the depth extent. Although seismic data make for poor imaging in volcanic rocks, they can provide a reliable stratigraphic structure above volcanic rocks to improve the vertical resolution of the gravity and magnetic method. We have developed an integrated interpretation method that combines the advantages of seismic, gravity, magnetic, and well data to generate a 3D quasigeology model to image volcanic units. We first use seismic data to obtain the stratigraphic boundaries, and then we apply an anomaly stripping method based on a seismic-derived structure to extract residual gravity and magnetic anomaly produced by volcanic rocks. We further perform the 3D gravity and magnetic amplitude inversion to recover the distribution of the density and effective susceptibility. We perform geology differentiation using the inverted density and effective magnetic susceptibility to identify the spatial distribution of four groups of volcanic units. The results show that the integrated interpretation of multigeophysical data can significantly decrease the uncertainty associated with any single data set and yield more reliable imaging of lateral and vertical distribution of volcanic rocks.

We present a case study on imaging volcanic units in gas exploration by constraining magnetic amplitude inversions using magnetotelluric (MT) sounding data at sparse locations. Magnetic data can be effective in mapping volcanic units because they have remanent magnetization and significant susceptibility contrast with surrounding rocks. Although magnetic data can identify the lateral distribution of volcanic units, they often have difficulties in defining the depth extent. For this reason, additional structural constraints from other geophysical methods can often help improve the vertical resolution. Among the independent geophysical methods, MT data can provide the needed structural information at a low cost. We have investigated an approach to combine a set of sparse MT soundings with magnetic amplitude data to image the distribution of volcanics in a basin environment. We first use a blocky 1D MT inversion based on Ekblom norm to obtain the structural constraint, and then we perform a constrained 3D magnetic amplitude inversion to recover the distribution of effective susceptibility by incorporating the structural information from MT soundings. We determine that even a small number of MT stations (e.g., 20) in a [Formula: see text] area is sufficient to drastically improve the magnetic amplitude inversion. Our results indicate that magnetic amplitude inversion with structural constraint from MT soundings form a practical and cost-effective means to map the lateral and vertical distribution of volcanics.