Gravity is pivotal in fundamental geological surveys and resource and energy exploration, yet volumetric and superposition effects of the gravity field have long posed a technical bottleneck in achieving high vertical resolution. Although boundary resolution is enhanced by applying mathematical transformations to the density model within a regularized stabilization framework and by constructing a focusing reweighting matrix, traditional focusing algorithms still rely heavily on subjective judgment in setting the focusing interval and deciding whether piecewise focusing is required. Such subjectivity reveals a lack of quantitative analysis. An innovative adaptive focusing iterative algorithm based on the minimum-support principle is introduced, integrating four quantitative assessment methods—Mean Square Error, Edge Sharpness Operator, overall Gradient Evaluation Operator, and Variance Evaluation Operator—to adaptively determine an optimal focusing strategy. Numerical simulations indicate that the approach not only automatically adjusts the focusing interval but also achieves high-resolution detection of target anomalous bodies. Tests conducted on real data from the Dahongshan area confirm that the method can finely characterize anomalous bodies in that region.

To address the problem of imbalanced category distribution in remote sensing image target detection tasks, an efficient dual-branch remote sensing image object detection method based on simple, parameter-free attention is proposed. The method first designs a sampling strategy based on positive example category weighting. This allows the model to initially learn minority-category features and then gradually transition to majority-category features, thereby increasing its focus on minority-category features. Then, the Simple Parameter-Free Attention mechanism is incorporated into the backbone network to further improve feature extraction for minority categories. Finally, a dual-branch detection output network and an SPPF_ReLU module are designed to enable the model to effectively distinguish between regression and classification subtasks. Simultaneously, an Im_Ghost model was designed to replace the C3 module in the dual-branch detection output structure, thereby reducing the number of parameters and enhancing feature fusion for minority categories. Experimental results show that the model achieves state-of-the-art performance on the highly imbalanced DIOR dataset, but on the NWPU VHR-10 dataset, due to its small sample size and balanced class distribution, performance is not optimal. Comparative experiments with mainstream models demonstrate the superiority of the proposed method. The trained model was applied to disaster remote sensing images, achieving good detection results.

Abstract The Majiaoba area, on the northwestern margin of the Longmen Mountains within the Sichuan Basin, exposes a complete Silurian–Triassic succession. Silurian–Devonian strata are widespread, Carboniferous–Triassic formations are well developed, and metamorphic or igneous rocks are rarely exposed at the surface. With the exception of the weakly magnetic Lower Triassic Feixianguan Formation, most units are essentially non-magnetic. Despite this, the region exhibits high-amplitude magnetic anomalies (up to ~103 nT) whose genesis remains uncertain, and basin-scale magnetic highs in central-western Sichuan are commonly ascribed to rifting even though their driving mechanisms and lateral continuity are debated. Two high-precision profiles across Majiaoba were acquired, and gravity and magnetic datasets were inverted independently using a minimum-support, adaptive iterative focusing algorithm selected under multi-criteria evaluation, yielding high-resolution images of subsurface structure beneath both lines. Rock-magnetic measurements on Feixianguan samples, combined with the inversion results, reveal a pronounced vertical offset between the principal magnetic source and the surface-exposed reddish mudstone-shale of the Feixianguan Formation. Along profile L2, a closed high-susceptibility body is resolved at ~400–600 m depth. The anomaly pattern is inconsistent with a purely shallow sedimentary origin and is best explained by mafic (diabase) intrusions. At the regional scale, the deep magnetic source is interpreted as a product of sustained activity related to the large-scale magmatic episode along the western margin of the Yangtze Block during the Late Permian, with magmatism in the northern Longmen Mountain segment plausibly persisting into the Early Triassic (Feixianguan time). Considering gravity–magnetic anomaly patterns across northwestern Sichuan, the magmatic system was likely widespread rather than local in extent.

Panzhihua-type V–Ti magnetite deposits in the Panxi region are hosted in mafic–ultramafic intrusions, and their exploration potential depends strongly on the deep distribution of ore-bearing intrusions. High-resolution 3D magnetic inversion is an effective tool to image the geometry of these intrusions. Using 1:50,000 aeromagnetic data, we applied an unsupervised deep learning inversion to obtain the 3D magnetic susceptibility structure of related intrusions. The results show that magnetic anomalies are mainly NS and NEE trending, with minor NNW-trending features. NS-trending sources occur in the Baima–Miyi–Hongge zone between the Xigeda–Yuanmou and Anninghe faults, while NEE-trending anomalies lie west of the Xigeda–Yuanmou fault and east of the Chenghai fault. Integrated geological analysis reveals two Late Variscan rift systems: the Anninghe rift and the Panzhihua rift. Deep fault-controlled magma ascent and emplacement, forming the Emeishan large igneous province, are associated with strongly magnetic intrusions. Mantle plume-derived magmas, differentiated in shallow and deep magma chambers, generate well-differentiated layered complexes at depths < 10 km with magnetic intensities of 5–10 A/m. Shear structures within paleorifts provide favorable emplacement conditions and controlled ore localization. We propose a three-in-one ore-controlling mechanism involving rift systems, intrusive rocks, and shear structures for Panzhihua-type V–Ti magnetite mineralization.

Recent catastrophic tailings deposit incidents have caused significant casualties and environmental damage, raising widespread concern. Tailings, with high saturation, fine particles, and poor permeability, are vulnerable to failure during earthquakes. Although tailings in closed dams become compacted due to long-term consolidation and are generally considered stable under static conditions, they may still collapse due to co-seismic damage accumulation from multi-stage earthquakes, a process that remains understudied. This study uses a saturated physical model of dense tailings to simulate earthquake-induced progressive failure. Embedded sensors and particle image velocimetry (PIV) technology were used to monitor the seismic response and deformation evolution of the model. During the test, sensors recorded the excess pore water pressure (EPWP) and dynamic earth pressure. The results showed that the slope failure was attributed to cyclic mobility in the silty sand layer, which can be divided into three stages: initial stage, cyclic mobility-triggered stage, and failure stage. Cyclic mobility initiated when seismic loading exceeded a critical threshold, triggered by significant EPWP and large horizontal seismic forces during strong shaking period. In addition, a plastic displacement coefficient (D) is proposed to elaborate the co-seismic damage and the failure mechanism. Finally, the test results indicated that the damage accumulation from cyclic mobility increases the risk of later earthquake-induced instability, which provides insights into the failure mechanism of decommissioned tailings slopes.

The deep exploration potential of basic-ultrabasic rock masses associated with Panzhihua-type vanadium-titanium magnetite (VTM) deposits are closely tied to the occurrence of deep-seated rock bodies. In this study, we utilized newly acquired 1:50000 scale aeromagnetic data from the Panxi region to perform a 3-D magnetization inversion using an improved regularized focusing conjugate gradient approach to achieve high-resolution 3-D magnetic imaging of basic-ultrabasic rock masses within the "Panzhihua-type" VTM concentration zone at depths reaching 10 km. The inversion results reveal that the 3-D magnetic anomalies of strong magnetic sources correspond with the distribution of the NS fault zones in the study area. However, these anomalies are predominantly located within narrow zones between the fault zones rather than directly along the fault lines. It also suggests that during the Late Huashan period, two rift regions might have developed in the Panxi area: the Anninghe Rift and the Panzhihua Rift. The deep and large faults within these confined rift valleys likely controlled the eruption and intrusion of mantle-derived magma, facilitating the emplacement of basic-ultrabasic strong magnetic rock masses along these zones. Additionally, the local shear structures within the paleo-rift zones may have provided ample space and a relatively stable environment conducive to the formation of VTM deposits.

An appreciation of the geodynamics of the Cenozoic great collision between the India and Eurasia plates necessitates an understanding of the lithospheric structure beneath Tibet. Here, we utilise the World Gravity Map 2012 (WGM2012) gravity data to derive the lithospheric density variations and gravity anomaly beneath Tibet, employing the multi-scale inversion and regularised downward continuation methods. The findings indicate that the E-W striking structures persist to a depth of 90 km. The deep lithospheric mantle displays block-like density anomalies that exhibit a distinct N-S trend with increasing depth. The results clearly indicate a decoupling between the lithospheric crust and mantle. It is possible that changes may occur along the Indian plate's subduction front, which subducts to the south of 32 degrees N along the western portion of the subduction zone, as indicated by the apparent density of the lithospheric mantle. In contrast, in the central and eastern regions, the Indian lithosphere appears to subduct primarily north of the Indus-Tsangbo suture zone (30 degrees N), where it may undergo tearing and delamination beneath the Lhasa block. This phenomenon promotes the convection of soft materials and the upwelling of mantle material, which causes the Tibetan lithosphere to thin and stretch in an E-W direction. At a depth of 150 km beneath central Tibet, alternating lower-density belts trending north-south may be connected at greater depth. These structures likely reflect lithospheric deformation, which plays a crucial role in shaping the geodynamics of the region.

Gravity inversion plays a crucial role in mineral exploration and resource evaluation, yet conventional depth-weighting methods often impose uniform resolution across all depths and fail to effectively delineate anomaly boundaries. This study presents an innovative attentional depth-weighting matrix based on a regularized downward continuation (RDC) mechanism. First, the observed gravity data are projected to greater depths using RDC, which suppresses high-frequency noise amplification. Next, gradient extrema are extracted from each grid cell to identify anomaly boundaries, forming a constant weighting matrix that enhances the focus on target regions. This matrix is then integrated with traditional depth weighting and a minimum-support focusing factor to optimize the inversion process. The proposed method is validated through two synthetic models, demonstrating improved resolution of deeper targets and more accurate amplitude recovery compared to conventional approaches. Further application to the Dahongshan Copper–Iron Ore region in Yunnan, China, reveals a deep intrusive body at approximately 4–5 km depth, extending east–west with a distinct “U”-shaped geometry. These results, consistent with previous geological studies, highlight the method’s ability to enhance deep anomaly characterization while effectively suppressing shallow noise interference. By balancing noise reduction with improved resolution, this approach broadens the applicability of gravity inversion in geological, geothermal, and mineral resource exploration.

3D gravity inversion has been widely used in mineral resources exploration and research on deep density structures. However, traditional 3D gravity inversion methods in spatial domain suffer low depth resolution, severe non-uniqueness and low computational efficiency, which affect the accuracy and reliability of geological interpretation. Aiming at the above problems, this study proposes a 3D gravity inversion method in frequency domain based on a mixed norm regularization constraint. We firstly construct a model objective function based on depth weighting and L-1+L-2 mixed norm constraints, and apply the frequency domain method of 3D gravity forward modeling to each iteration of an inversion to update the model so that the problems of storing and computing the dense Jacobian matrix in traditional spatial domain inversion methods are transformed into frequency domain forward modeling, which greatly reduces the computational time and memory occupation. To ensure high accuracy of the frequency domain forward method, the high-precision Gaussian numerical integral is used instead of the rectangular integral in traditional Fast Fourier Transform (FFT) algorithm. Two synthetic inversion examples show that the proposed inversion method can effectively reduce the "skin effect" and smearing phenomenon and can recover more complex geological models compared with the traditional spatial domain inversion methods based on L-2 norm regularization and L-1 norm (focused) regularization. Finally, we apply this method to 3D density imaging of the Mobrun sulfide body in Noranda, Quebec, Canada. The results show that the depth of the sulfide body ranges from similar to 15 to 170 m, consisting with previous borehole data, which proves the effectiveness of the proposed method.

The "Lala-type" copper-iron deposits in the Lala area are primarily located in the central part of the Kangdian tectonic zone, distributed along an East-West (EW) trending Neoproterozoic gabbro-dolerite belt that is prominently controlled by regional EW-oriented tectonic structures. The genesis of these deposits is widely believed to be closely related to ancient volcanic structures, but their relationship with deep-seated basic intrusive rocks remains highly controversial. This article proposes an improved 3-D re-weighting regularized conjugate gradient (RRCG) focusing inversion method, constrained by the background field to recover the anomaly with a high resolution. The magnetic structure in a real case indicates the presence of a distinctly oriented basic intrusive rock mass in the deep part of the "Lala-type" Shilong copper deposit. Large-scale, high-precision magnetic profiles and magnetotelluric inversion results show that this strongly magnetic and high-resistance intrusive rock mass, which extends to a depth of 1.5 km, has transformed the ore-bearing basement into a clearly imaged anticlinal uplift and fold structure. Trenches and drill holes reveal that the ore bodies are primarily located in the fold and detachment spaces formed at the intersections of EW and North-South (NS) faults. The breccia formed by the basic rock intrusion along the faults provides favorable conditions for the occurrence and enrichment of ore bodies. The multiple intrusive thermal events in the Lala area not only supplied the fluids necessary for the enrichment of the deposits but also facilitated the transformation of the basement and the formation of ore-bearing spaces.

Salt domes are very important in hydrocarbon exploration and identification of potential drilling hazards. While seismic data are indispensable for detailed subsurface imaging, especially in delineating the geometry and properties of salt bodies and their boundaries, gravity inversion provides an additional layer of data by exploiting the density differential. However, traditional methodologies for tackling this problem are complicated by the ill-posedness of the inverse problems. The alternative approach to gravity image is based on machine learning (ML) algorithms. Despite the appealing attributes of convolutional neural networks (CNNs), they are not exempt from limitations, including diminished precision in pinpointing geological features, complications in managing the varying scales of geological structures, and inefficiencies in processing voluminous, high-dimensional data. These deficits can be mitigated by the proposed multiscale functional multiscale UNets (MS-UNets) network, which, through integration with squeeze-and-excitation (S-E) and strip pooling (S-P) modules, are designed to enhance the capture of detailed information about salt domes. These networks were subjected to rigorous testing using both synthetic and real gravity data, showcasing their robustness across diverse scenarios. This testing highlighted their significant potential for applications in geophysical data interpretation, structural modeling, and inversion processes.

In geophysical research, gravity-based inversion is essential for identifying geologic anomalies, mapping rock structures, and extracting resources such as oil and minerals. However, traditional gravity inversion methods face challenges, such as the volumetric effects of gravity fields and the management of large complex matrices. Unsupervised learning techniques often struggle with overfitting and interpreting gravity data. This study explores the application of various U-Net-based network architectures in gravity inversion, each offering distinct challenges and advantages. Nested U-Net, although effective, requires a high parameter count, leading to extended training periods. The its dynamic adaptability, whereas the attention U-Net's lack of research comprehensively analyzes the training processes, core functionalities, and module distribution of these networks, includ- ing the residual U-Net++. Our synthetic studies compare these networks with traditional focused regularized gravity inversion for reconstructing density anomalies. The results demonstrate that the nested U-Net closely approximates the actual model despite some redundancy. The recurrent residual U-Net indicates an im- proved alignment with minimal redundancies, and the attention U-Net is effective in density prediction but encounters difficulties in areas of low density. Notably, the residual U-Net++ excels in inversion modeling, achieving the lowest misfit percentage and accurately replicating density values. In practical applications, the residual U-Net++ impressively reconstructs the F2 salt diapir in the Nordkapp Basin with well-defined boundaries that closely match seismic data interpretations. These results underscore the capabilities of the residual U-Net++ in geophysical data analysis, structural reconstruction, and inversion, demonstrating its effec- tiveness in simulated settings and real-world scenarios.

The stark contrast in density between geological layers is a fundamental aspect in the examination of basic geological structures. The delineation between the crystalline basement and sedimentary layers, moreover, is pivotal in the pursuit of strategic energy resources, such as petroleum and natural gas. Traditional full space density inversion, however, is beleaguered by issues of stability and resolution, impeding the accurate characterization of the sharp density interface. To rectify these shortcomings, we introduce an innovative methodology for estimating 2-D depth-to-basement and overlying density distribution, employing a deep neural network with a leaky rectified linear unit as an activation function. Evaluation of the proposed method on simulated sedimentary basin models underscores its superior ability to discern complex geometries of basin boundaries and overlying density, despite the presence of various degrees of Gaussian noise. In practical application to the Poyang basin, the relief of the Cretaceous basement is proficiently recovered through vertical gravity field data, with validation provided by corresponding seismic sections and well-established stratigraphic markers.

Thermophysical properties determine the thermally upgraded area of low-maturity oil shales, which is of significance for restoring oil recovery but needs more in-depth investigations. We introduced a laser-flash NETZSCH LFA-467 analyzer to measure the dynamic response of thermal diffusivity (D), thermal conductivity (K), and specific heat capacity (c) at an in situ elevated temperature as well as their anisotropy in the orthogonal direction relative to shale bedding. The D and K values decrease with a rising temperature, while c values remain almost unchanged at a high temperature. The total organic carbon (TOC) content significantly affects the D and K values, mainly in the anisotropy coefficient (a newly proposed parameter), but sparingly influences the c values. The anisotropy coefficient of samples with a high TOC content is similar to 2 times greater than that with a low TOC content. The greatest D value decay rate of a high TOC content sample reaches similar to 75%, while that for a low TOC content sample is similar to 43%, found in the bedding-perpendicular samples. Results also suggest that the anisotropy of the thermophysical property derives from the lamellar structure of shale rock; the space (like pore and/or fracture) therein is averse to the heat movement. For the hypothetical vertical well mode or horizontal well mode, the arrangement of reasonable space between wells of heat injection and restoring oil production is of significance under the heating-induced dynamic evolution of thermophysical parameters. Hopefully, this work is helpful in enhancing the knowledge on the heat flow behavior in low-maturity oil shale when thermal upgrading is implemented.

On June 10, 2022, the MS6.0 Ma'erkang earthquake sequence occurred in the Bayan Har block. In this paper, the temporal and spatial distribution and attenuation characteristics of earthquake sequence is analyzed based on the regional structure, the temporal and spatial distribution of earthquake sequence, the focal mechanism solution, and the parameters of earthquake sequence, using the data of Sichuan Seismic Network and the temporary stations incorporated into the network. The results show that: (1) The MS6.0 Ma'erkang earthquake sequence is generally distributed in NW-SE direction, and the long axis of this series is basically consistent with the nearby Songgang fault. (2) As the earthquake sequence is formed by three large events with MS ≥5.0, the frequency of small earthquakes in the earthquake sequence area generally attenuates relatively slowly, while the activity level (magnitude) of aftershocks attenuates rapidly. (3) After the MS5.2 earthquake, the sequence parameters obtained show that the h-value is 1.09 and the p-value is stable at 1.02, indicating that the intensity and frequency attenuation of the earthquake sequence are gradually stable and tend to be normal. The b-value is 0.95, indicating that the maximum aftershock magnitude of the series is estimated to be ML5.1, and the b-value gradually tended to be stable, indicating that the stress in the region gradually tended to be balanced after the MS5.2 earthquake. The MS4.4 (ML5.0) earthquake that occurred at 4:37 on June 10 (local time) is the largest aftershock after the MS5.2 earthquake in the sequence. (4) The Songgang fault with NW-SE trend is presumed to be the main seismogenic tectonics, but the migration of three earthquakes with MS≥5.0 may also indicate that the Songgang fault is not a single seismogenic tectonics, which requires further field scientific investigation and analysis.

Numerous investigations have suggested that seismic source parameters and properties of the surrounding medium harbor valuable information regarding alterations in stress fields and medium properties at the focal depth. Monitoring the spatial and temporal evolution of these parameters can yield insights into variations in stress fields or medium properties within the seismogenic zone, offering a critical avenue to overcome the Earth's inaccessible nature. In recent years, modern seismic parameters such as seismic moment, focal mechanism solution, source stress drop, corner frequency, radiated seismic energy, rupture radius, and apparent stress have been increasingly used to characterize source characteristics. The Sichuan region is situated on the southeastern edge of the Qinghai-Tibet Plateau and is known for its strong eastward extrusion and structural deformation. The main fault zones in the area include the Xianshuihe Fault, Anninghe Zemuhe Fault, and Longmenshan Fault. This paper estimates source parameters of small to medium-sized earthquakes (ML 1.5-5.2) in the main fault zones and adjacent areas in Sichuan. A total of 4,310 earthquake stress drop measurements from 2019 to 2023 were analyzed to create a stress distribution image along the fault. This earthquake parameters training catalog provides insight into how coseismic stresses change and their relationship with other geophysical factors, as well as their spatial and temporal evolution.

The Red River fault is a Holocene active fault between the Indo-China plate and Yangtze plate, its great earthquake risk has gained attention. The middle-north section and nearby areas of the fault can be divided into three tectonic units. In the study, the crustal velocity structure images and accurate earthquake locating results were obtained for various zones by using the tomoDD method. The results showed that the northern and middle section of the region were characterized by low-velocity anomalies in NW and NS directions, and such feature became more obvious as the depth increased. On the west section, large-scale low-velocity anomalies were correlated with the Kainozoic volcanic zone. In addition, from the results of stress drop and velocity structure, the correlation between them is not obvious. Our study may help to understand seismogenic structures of complex fault systems in the Middle-north section of Red River Fault and adjacent areas.

The Litang fault (LTF), located in the southeast of the Qinghai-Tibetan Plateau, is known for its high level of present-day seismicity, whereas its Pleistocene activity has been scarcely documented. This study focused on a tract of banded travertine deposits precipitated from thermal waters along the NW-SE-trending LTF trace. The role of travertine deposits in recording neotectonic activity has been studied by identifying their internal structure. Typical soft-sediment deformation structures observed within the banded travertines include micro folds, liquefied breccia, and liquefied diapirs. These deformed structures, which are restricted to a single unit separated unconformably by undeformed layers, can be traced for tens of meters, indicating that they were formed by seismic shaking triggered by LTF activity. The deformation of the banded travertine layers is attributed to the combined effects of seismic shaking, liquefaction, and fluidization, and it can be related to a paleo earthquake event with a magnitude of M-S > 5. The U-series ages obtained from the banded travertine deposits perturbed by the earthquakes are in the range of 130.59-112.94 ka, indicating an important fault-assisted neotectonic activity that occurred during the Middle-Late Pleistocene. Analysis of such structures, in combination with the use of U-series dating methods, can yield a reliable timing of neotectonic activity and provide new evidence for understanding the seismotectonic setting of the Litang area.

The Dahongshan deposit is influenced by pronounced structural controls and intrusions, and its model of mineralization resulting from volcanic sedimentation has faced criticism for an extended period. To describe the deep structure and distribution characteristics of the ore deposit, and to investigate its ore-forming process and metallogenic model, this study uses the gravity, magnetic, and controlled source audio-magnetotelluric (CSAMT) data with different scales to independently recover density, magnetization intensity, and electrical resistivity for shedding light on the deep structure and mineralization distribution of the deposit. The inversion results show the presence of a giant intrusion extending up to 6 km deep within the deposit. The distribution of iron-rich ore bodies and deposit is controlled by the basement tilting and faults, with the deposit exhibiting a U-shaped distribution and the mineral body occurring in a lens-like shape. In addition, the electrical results indicate the presence of high-resistance magma along faults that intrude the deposit. We propose that the deposit is a magmatic-related deposit, with a deep intrusion believed to be the residual source body from early rift intrusion, providing the source for mineralization of the deposit. The characteristics of the deposit controlled by east–west and north–south structures and the lenticular orebody distribution indicate that the deposit is closely influenced by regional structure and magmatism of fault intrusion. The mineralization model is proposed to be a result of the coordinated actions of structural, magmatic, and regional dynamic background for the breakup and convergence of supercontinents.