Many cities or urban residential areas in central and western China reside in alluvial plains formed from piedmont alluvial fans. Hence, revealing the stratigraphic architectures and stability of alluvial fans holds critical significance for urban construction planning and rational land use. The alluvial fan in the eastern piedmont of the Liupan Mountains hosts the urban residential areas and villages of Guyuan City, with a dense population. Moreover, the alluvial fan develops several event deposits recording the activity of the alluvial fan under tectonic movements and climatic changes. Through field geological survey, optically stimulated luminescence dating, controlled source audio-frequency magnetotellurics (CSAMT), and conventional radon measurement, this study revealed the stratigraphic architecture of the alluvial fan and its two-phase event deposits (~43.33 ka B.P. and 22.92~20.72 ka B.P) since the Late Pleistocene. As indicated by the CSAMT and conventional radon measurement results, the alluvial fan still exhibits high activity under the influence of the Haiyuan and Qingshuihe faults. The results of this study provide fundamental data for crustal stability assessment, prevention and control of geologic hazards, and engineering construction in the Liupanshan area.

The method of structural superimposed halo to find blind ore is a new method to find blind ore, which is based on the study of the theory of primary halo to find blind ore-the axial zoning of primary halo, and puts forward two new theories of ‘primary superimposed halo theory’ and ‘structural superimposed halo theory’. The accuracy of blind ore prediction by superimposed halo of structure is determined by the correctness of prediction marks and indexes. Based on the summary of seven common signs of structural superimposed halos in the prediction of deep blind ore deposits in more than 100 hydrothermal deposits, four important quantitative qualitative prediction signs are upgraded to quantitative prediction indicators. The structural superimposed halo prediction signs and indicators of 17 different combinations of accurate prediction of blind ore and discrimination of ore body denudation degree in the deep and periphery of the mining area are described in detail, and a practical ideal model of structural superimposed halo for predicting blind ore and discriminating ore body denudation degree is established. The common prediction signs and indicators of the model have important guiding significance for the prediction of deep blind ore in typical hydrothermal deposits, and have achieved remarkable prospecting results in more than 100 mine applications.

Using seismic reflection profiles and drilling data, this study determined the types and styles of inverted structures in the Tabei Sag, Erlian Basin. Accordingly, this study explored the genetic mechanisms of the inverted structures and their relationships with sandstone-hosted uranium mines. The results indicate that the eastern and western parts of the Tabei Sag experienced different degrees of normal and inverted tectonism. As a result, the eastern part exhibits typical inverted structural styles. In contrast, in the western part, the Saihan Formation on the hanging wall of inverted fault F₁ in the Bayanwula and Manglai area was uplifted and almost completely eroded, with the original geometries of inverted structures being destroyed. During the late depositional stage of the Early Cretaceous Saihan Formation (113~98.9 Ma) and from the Late Cretaceous to the Early Paleocene (66~42 Ma), the subduction direction of the Paleo-Pacific Plate shifted from NW to NWW. This change altered the stress regime in the Tabei Sag from extension to compressional inversion, leading to the formation of a series of compressional and compressional-torsional structures. Consequently, the Saihan Formation and the Upper Cretaceous strata were uplifted and eroded, resulting in the formation of regional angular unconformities. The structural inversion transformed the sedimentary system of the Saihan Formation in the Tabei Sag from lacustrine to fluvial facies. Meanwhile, it caused differential uplift and uneven erosion of the Saihan Formation and its overburden, leading to the formation of erosion windows. This facilitated the infiltration and migration of uranium- and oxygen-bearing fluids toward deep parts. These processes controlled the morphologies and development of interlayer oxidation zones, thereby promoting the enrichment of uranium deposits.

The Salt Lake geothermal field in Yuncheng City, Shanxi Province, China lies beneath a densely populated urban area, posing significant challenges to further geothermal exploration and extraction. Based on the distribution of geothermal gradients in the geothermal field, which are higher in the southwest and lower in the northeast, a NE-trending microtremor survey profile was arranged in the southern part of Yuncheng City, aimed at investigating the deep geothermal reservoir structure and NW-trending structures in the geothermal field. The 2D velocity structure profile reveals a pronounced low-velocity anomaly in the eastern part, which is supposed to be induced by the fault fracture zone formed by multiple NW-striking tensional faults. Spatially, this concealed fault zone roughly corresponds to the low-geothermal gradient anomaly in the northeastern Salt Lake geothermal field, suggesting that this fault fracture zone might facilitate the rapid infiltration of cold surface water, thereby lowering the temperature of deep rocks in the northeastern part, leading to the formation of a large-scale low-temperature anomaly zone. Additionally, the faults identified by the microtremor survey can be traceable and confirmed in a controlled source reflection seismic profile in the study area, demonstrating the complementary nature of the two methods. This study further reveals the deep geothermal structures of the Salt Lake geothermal field based on previous efforts. This study provides more valuable bases and guidance for future exploration and resource evaluation of geothermal fields in the region while also demonstrating the effectiveness and superiority of the microtremor survey method in research on urban geothermal resources.

Areas with low exploration degrees in the Tarim Basin exhibit intricate seismic-geologic conditions and extensively developed multiples.The internal multiples generated by strong reflection interfaces involving Paleozoic gypsum rocks undermine the imaging of the Ordovician stratigraphic information,resulting in vague images for Ordovician fractured-vuggy carbonate reservoirs and thus affecting stratigraphic interpretation.Accurate prediction and suppression of internal multiples in the Yubei area is challenging,thus restricting the deployment of oil and gas exploration in the area.The conventional prediction method for internal multiples is computationally intensive and ineffective.Hence,this study proposed an internal multiples suppression technique combined with the modeling method and adaptive matching subtraction.Based on the principles of internal multiples prediction and the forward modeling results,the proposed technique takes the main reflection interfaces generating internal multiples as the model layers and follows the approach of layer-by-layer prediction and suppression.Finally,it employs adaptive matching subtraction for the suppression of internal multiples.As demonstrated by the processing results of actual data,the proposed technique can effectively predict and suppress the dominant internal multiples to eliminate their interference effects.It solves the problems of layer-crossing phenomenon and flat formation occurrence,improving the imaging accuracy of formation structures,faults,and fractured-vuggy reservoirs.

Seismic data processing reveals that internal multiples exhibit highly complex formation mechanisms,making their accurate identification and effective suppression a persistent challenge in geophysical exploration.Traditional methods for internal multiple suppression frequently require manual identification of subsurface reflection interfaces,making them difficult to apply to complex underground medium structures.Moreover,these methods are not only computationally cumbersome but also typically ensure only temporal and positional consistency when predicting internal multiples.The amplitude often deviates from observed values,requiring adaptive matching subtraction algorithms for effective suppression.This study developed an internal multiple suppression method based on Marchenko theory.Specifically,the wavefield relationships between the focusing function and Green’s function were constructed using the convolution and correlation reciprocity theorem of the Green’s function during data processing.The Green’s function was then solved using the focusing function,yielding either multiples or primaries constructed from the Green’s function.This method requires only the background velocity or the original data as operators during the iterative multidimensional correlation and convolution process,rendering it simple and computationally efficient. This study constructed an expression for the primary wave field using the Green’s function and the Marchenko equation.The tests using the SMAART model and actual data from the Gulf of Mexico demonstrate that the Marchenko method can effectively suppress internal multiples under the conditions of complex subsurface media.The iterative process requires no velocity information,providing significant advantages over conventional methods and great potential for application in complex underground environments where layers are difficult to distinguish.

Conventional inversion and forward modeling of large-scale potential field data from gravity and magnetic exploration, demanding high computer performance, exhibit low efficiency. Hence, this study defined a footprint determination method for potential fields, analyzed the influencing factors, and innovatively proposed a footprint-FFT strategy for forward modeling of potential fields. The footprint-FFT algorithm improved the forward modeling process from three aspects: (1) Kernel matrices were calculated based on the potential field-derived properties, significantly reducing their size; (2) A footprint concept for potential fields was introduced and defined, decoupling data scales from kernel matrix sizes, thus improving the kernel matrix computing efficiency and reducing the hardware cost; (3) Based on the above, the computing area was divided into subspaces, and the footprint-FFT strategy was first proposed for the batch computing of potential fields in subspaces, accelerating the forward modeling process. By reducing the computational complexity and storage of the kernel matrix, the method proposed in this study significantly improved the operational speed while ensuring computational accuracy. This method enabled the fast forward modeling of potential fields with more than 1 billion grids on a laptop computer within a few minutes. Theoretical examples demonstrate that this method has high efficiency and moderate requirements for computer configuration, manifesting considerable potential in the forward modeling and inversion of large-scale potential field data.

Traditional resistivity inversion methods tend to rely on the initial inversion model selected, get stuck in local minima, and be time-consuming. To address these issues, this study proposed a real-time resistivity inversion method based on the Res-UNet neural network. First, a significantly expanded forward response dataset was generated using the Gmsh software. Then, inversion experiments were carried out based on appropriate network parameters determined according to data characteristics. The experimental results indicate that the Res-UNet algorithm can fully dig the data characteristics and rapidly produce resistivity images that align with the electrical properties of strata. The experiments on the dataset for resistivity forward modeling yielded a mean squared error between the predicted values and the forward responses of 0.019 44, and those on the test set yielded a mean squared error of 0.075 8, suggesting improved imaging results compared to traditional inversion methods. Furthermore, the proposed method achieved encouraging results in the inversion calculations of simulation models, enabling rapid and accurate inversion of the location and morphologies of subsurface anomalies while exhibiting a strong noise resistance. This study provides a new method and philosophy for mapping the relationship between resistivity data and the actual geoelectric structures.

The tipper vector,a significant parameter in magnetotelluric(MT) sounding,is applicable to infer fault structures that cause lateral inhomogeneity of media.Subsurface faults typically exhibit three-dimensionality and complexity.To reveal the MT tipper response characteristics in 3D fault models,this study conducted numerical simulations of the MT tipper response in 3D fault models based on the vector finite element method.First,the validity of the 3D tipper forward modeling program was verified through theoretical model calculations and comparisons with previous finite element results.Subsequently,four typical 3D models for vertical,normal,reverse,and strike-slip faults were employed for forward modeling,obtaining the response characteristics of the real part,imaginary part,amplitude,and phase of the tipper.The simulation results are as follows:(1) In two polarization modes,the response characteristics of the real part,imaginary part,and amplitude of the tipper effectively reflect the properties,strikes,and dip directions of the four different faults while indicating the location of the laterally inhomogeneous boundaries,thus serving as a significant basis for discriminating fault types and characteristics;(2)In contrast,the relatively complex response characteristics of the phase fail to effectively mirror the fault characteristics.

In thereconstruction of actual subsurface structures, strong noise limits the accuracy of the magnetotelluric (MT) method,causing adverse effects on later data interpretation. Given this and the characteristics of the MT time series,this study analyzed different types of noise in the MT time series,proposing a signal denoising technique based on variational mode decomposition (VMD) and long short-term memory (LSTM) predictive reconstruction. First, baseline drift correctionwas performed for the original MT datausing the VMD signal decomposition algorithm. Then, the time series was further decomposed into multiple different intrinsic mode functions (IMFs) through VMD. The LSTM time series detection model was trained using interference-free data in the RSE component, which was then identified. Afterward, the time intervals containing noise weremarked, the increasement of noise was calculated, and the noise information wastransmitted to the original signal for truncation and removal. Finally, an LSTM multi-dimensional prediction model was trained for the IMFs, followed by the prediction of missing values under various modes. The predicted results under all modes were combined to obtain the final predicted MT signals. After signal reconstruction, a secondary signal-noise separationwas performed for spike-pulse noise that was not effectively identified through VMD. TheVMD-LSTM-based signal denoisingtechnique can accurately identify strong noise in MT signals by merely processing the time series intervals containing noise, thuseffectively preserving interference-free data. Moreover, its prediction errors can berestricted within the allowable error range of the data processing for MT signals. Therefore, this technique enjoys significant denoising effects.

Effectively separating target anomalies while minimizing over- or under-separation remains challenging in gravity and magnetic field separation. In this study, the low-rank decomposition was employed to separate gravity and magnetic anomalies. Additionally, to determine the balance parameters that affect potential field separation, this study proposed an optimal estimation method based on the minimum correlation coefficient. Tests of various separation methods based on theoretical gravity and magnetic anomaly models demonstrate that the proposed method allows for effective separation, significantly reducing under- or over-separation caused by the sliding window average and wavelet analysis methods. The proposed method was applied to the Bouguer gravity anomaly data from a mining area in western China. The separated local anomalies clearly reflected the presence of rock and/or ore bodies with low magnetic susceptibility and high density. Model experiments and field applications demonstrate that the proposed method can enhance the accuracy and practicality of potential field separation.

The inversion of petrophysical data can image the microscopic fracture structures inside rocks, revealing the evolutionary patterns of fractures within rocks and soil with changes in external environments. Hence, it is an intuitive and reliable method for investigating the mechanisms of deep geotechnical disasters. This study presented a petrophysical data acquisition system and resistivity-based forward modeling and inversion algorithms. Based on the above, this study conducted numerical simulations of 2D and 3D inversion imaging of petrophysical data. As indicated by the numerical simulation results, 2D inversion imaging can characterize millimeter-scale rock fractures with high/low resistivities, whereas 3D inversion imaging can accurately locate and effectively identify millimeter-scale fractures and vugs with high/low resistivities. Moreover, data measurement and inversion imaging were conducted on rock samples subjected to microwave-induced fracturing in three states: heated sandstone before failure, sandstone heated to a molten state, and molten sandstone in a cooled state, preliminarily revealing the variation patterns of sandstone fractures under microwave heating. Overall, this study provides a novel method for exploring the mechanisms of deep geotechnical disasters.

With the advancement of social economy and science and technology, the demand for oil and gas resources has been increasing in daily life and industry. Tight sandstone reservoirs have been the priority targets for the exploration and production of oil and gas resources. However, there still exist many challenges in assessing the parameters and quality of tight sandstone reservoirs. This study conducted experiments on the physical properties, pore structures, and electrical properties of rock samples from the Taiyuan Formation in the Shenmu gas field of the Ordos Basin. Based on this, it established a porosity-permeability relationship model, a capillary pressure prediction model, and a classification saturation assessment model. Besides, it obtained the relative permeability of gas and water phases, which varied point by point, from wells based on the I-Kr model. This study proposed the factors for assessing reservoir quality, which were applied to the target interval in the study area considering the physical properties, pore structures, and multiphase seepage capacity, yielding satisfactory assessment results. Therefore, the method of this study provides a reliable basis for the log-based assessment of the quality of tight sandstone reservoirs.

The topographic factor significantly influences the 3D inversion results of magnetotelluric data. Despite extensive research results previously obtained in suppressing topographic effects, the gridding of complex terrains (with significant elevation changes) is still challenged by grid design complexity and difficulty in correcting data elevation points. Based on the mainstream 3D inversion module ModEM for magnetotelluric data, this study proposed a novel method for rapid automatic grid design and partitioning of terrains based on unsupervised learning, primarily involving the K-means++ algorithm and the assessment of clustering effects. Compared to the uniform and equal proportion-based hierarchical methods ignoring the topographic factor, the proposed method shows the following advantages: (1) The terrain grid generated by the clustering-based hierarchical method manifested higher terrain approximation, reducing the average error between the terrain grid and the actual terrain by 25%; (2) The matching calculation for terrain correction based on the digital elevation model was somewhat avoided; (3) The rapid design of terrain grids can be achieved, and the hierarchical characteristics can be referenced for gridding in other modeling software. The proposed method was employed to demonstrate the whole process of partitioning the elevation data of a complex terrain in a mining area, generating a resistivity structure model more representative of the actual terrain characteristics. Based on this model, finer-scale 3D inversion results were obtained. Theoretical and practical applications illustrate that the proposed method can significantly improve the topographic adaptability of gridding, holding critical significance for suppressing topographic effects on the 3D inversion of magnetotelluric data.

AVO inversion, based on the Zoeppritz equation, extracts various hidden petrophysical parameters from pre-stack seismic data. In seismic data, angle information is recorded in the form of offset values, and converting between offset values and angles is prone to generate errors. In addition, using the same approximate formula for different acreage types may lead to reduced applicability due to varying actual geological conditions. The exact Zoeppritz equation will lead to increased computational demands due to its high complexity. Therefore, this study developed an adaptive prestack inversion method based on the quadratic Encoder-Decoder network. This inversion method used the high feature and relationship extraction abilities of deep learning to replace traditional relationships, thereby reducing the angle errors and adapting to varying acreage types and geological conditions. The quadratic Encoder-Decoder network used a quadratic algorithm as the optimization method, maximizing the efficiency of the standard Encoder-Decoder structure. Additionally, the Xavier initialization method was incorporated to enhance the randomness of model initialization, thus improving the robustness of the network. The results indicate that the quadratic Encoder-Decoder network, selected through orthogonal experiments, outperforms the single-decoder network in prediction and exhibits greater consistency with actual log curves. The P-wave velocity, S-wave velocity, and density profiles obtained from inversion are consistent with the geological conditions of the study area, exhibit strong lateral continuity, and can effectively achieve high-precision prestack inversion.

Marine seismic data from sparker sources exhibit strong noise interference, complex wavelet morphologies, signal distortion due to cables' feathering and undulation, and severe multiple wave interference, all of which greatly affect the imaging quality of seismic data. Based on the characteristics of marine seismic data from sparker sources, this study established a data processing workflow and elaborated on the key technologies involved, including LIFT high-fidelity noise suppression, cable drift correction, surface-consistent simulated annealing static correction, wavelet processing, and free surface multiple suppression technologies. The application to actual data indicates that this processing workflow can effectively address the challenges associated with marine seismic data from sparker sources, improve the signal-to-noise ratio, restore the wavelet bandwidth, and produce seismic sections with superior broadband imaging and clearly defined geological features. This study provides a novel technical means for the application of marine wide-azimuth seismic data.

The dolomite reservoirs of the fourth member of the Majiagou Formation in the Ordos Basin are characterized by thin layers,tightness,complex pore shapes,strong heterogeneity,and relatively weak seismic responses.The mechanisms behind the seismic prediction of the reservoirs remains uncertain, making fluid identification challenging.Traditional geophysical methods based on a single attribute fail to accurately predict fluids.Therefore,by comprehensively accounting for the pore shapes and connectivity,this study proposed a new fluid identification method using petrophysical modeling and analysis,as well as wave theory-based prestack AVO inversion,frequency-dependent AVO inversion of fluid factors,and probabilistic neural network(PNN)-based prediction of pore structure parameters.This method,comprehensively considering the impacts of elastic parameters,physical parameters,and dispersion properties,achieve remarkable results.Compared to traditional single-attribute fluid identification method,the proposed method demonstrates higher accuracy,particularly for gas-bearing areas,fully verifying its effectiveness in fluid identification and highlighting its potential for widespread application.

The Tarim Basin,one of China's most significant oil exploration areas,exhibits thick sedimentary rock layers and frequent tectonic movements.These characteristics have led to the formation of abundant source,reservoir,and cap rocks,creating favorable conditions for the generation and storage of resources like petroleum.However,the high topographic relief in the area poses significant challenges to observation system arrangement and acquisition engineering design.Moreover,undulating surfaces and complex subsurface structures affect the propagation of seismic waves,impairing the quality of seismic exploration data and complicating data preprocessing,imaging,and reservoir prediction.Given the data loss of the Tarim block caused by suboptimal data collection,this study conducted high-precision reconstruction of seismic data using the compressed sensing technique,aiming to provide seismic records with high integrity,reliability,and precision for the preprocessing/superimposition phase.Compressed sensing,a novel sampling technique,plays a significant role in data reconstruction.The key to this technique is the adequate sparse representation of seismic data.However,conventional transform methods like Fourier transform and discrete cosine transform(DCT) are merely applicable to simple global structures.Considering the high complexity of the Tarim block data,this study employed the Shearlet transform as the sparse basis function for data reconstruction through compressed sensing.The technology of this study was finally applied to process the actual data of the Tarim Basin,demonstrating high accuracy and applicability for the area.

Survey grid layout is a preliminary task of geophysical and geochemical exploration. The widespread use of handheld GPS has greatly facilitated navigation and positioning in small-to medium-scale geophysical and geochemical field surveys. This study, based on coordinate transformation theory, presented a method for generating large numbers of GPS waypoints for both regular and irregular survey grids for geophysical and geochemical exploration. Furthermore, an associated software program was developed. The software enjoys simple interfaces and convenient operations, enabling the quick generation of survey grids suitable for applications such as gravity, magnetic, and resistivity surveys, soil surveys, and stream sediment surveys. Therefore, this software can be used for field navigation.

This study investigated the Panxi area as an example to delineate faults and determine their current activity levels using aeromagnetic data and precise relocation results of small earthquakes. A total of 42 faults were delineated in the study area. Among them, 21 faults were inferred to be active currently, including 10 newly inferred active faults. This study examined the three-dimensional structural characteristics of some local sections of major faults in the study area. The results indicate that precise relocation results of small earthquakes can identify highly active faults and characterize their deep structures. Additionally, based on the delineation of basement faults using aeromagnetic data, the locations of active faults were constrained, and concealed active faults within the sedimentary cover were also determined. The proposed method holds certain practical significance for promoting research on the spatial distribution and activity of fault structures using aeromagnetic data.

This study investigates the spatial distribution characteristics, temporal changes, and influencing factors of total carbon density, organic carbon density, and reserves in the farmland soils in Luoyang City based on repeated sampling data from multi-purpose regional geochemical survey points over different periods. The results indicate that in 2005, the topsoils in the multi-purpose survey area exhibited an average total carbon content of 1.57%, an average total carbon density of 44.74 t/hm², an average organic carbon content of 1.12%, and an average organic carbon density of 34.27 t/hm². With annual average increases in the total carbon density and organic carbon density of 0.709 t/hm² and 6.643 t/hm², respectively, this year witnessed increases in the total carbon and organic carbon of 41.73 kg/(hm²·a) and 390.75 kg/(hm²·a), respectively. The respective reserves of total carbon and organic carbon were 12.511 3 million tons and 8.879 59 million tons, with respective increases of 198.28 thousand tons and 1.857 8 million tons. In 2018, the topsoils in the multi-purpose survey area displayed an average total carbon content of 1.18%, an average total carbon density of 34.27 t/hm², an average organic carbon content of 1.07%, and an average organic carbon density of 30.94 t/hm². With annual average increases in the total carbon density and organic carbon density of 9.642 t/hm² and 4.727 t/hm², respectively, this year witnessed increases in the total carbon and organic carbon of -2,410.5 kg/(hm²·a) and 1,181.75 kg/(hm²·a), respectively. The respective reserves of total carbon and organic carbon were 1.963 5 million tons and 1.772 61 million tons. The total carbon reserves in topsoils in the whole study area were 14.474 81 million tons, including organic carbon reserves of 10.652 2 million tons. The total carbon reserves decreased by 552.41 thousand tons, while the organic carbon reserve increased by 270.82 thousand tons. Overall, the study area experienced a total carbon decrease of 354.13 thousand tons and an increase in the organic carbon reserves of 2.128 62 million tons. The total carbon showed significant positive correlations with the contents of organic carbon, CaO, MgO, N, and P, the organic carbon displayed significant positive correlations with the contents of total carbon, N, and P, and there was a significant positive correlation between pH and the CaO content. Fertilization led to an increase in the organic carbon and total carbon contents in farmland soils, with total carbon being significantly affected by the CaO content. In the carbonate areas, alkaline soil environments exhibited carbon sink characteristics. In the southern regions with acidified soil environments, the decomposition of carbonates in soils led to carbon loss and reduced calcium content. The results of this study provide important scientific evidence for research on peak carbon dioxide emissions and carbon neutrality of Luoyang City.

Soil carbon pools constitute a crucial part of global terrestrial carbon pools. Hence, investigating soil carbon pools is critical for understanding the global carbon cycle and changes. Based on the soil carbon data obtained from a multi-purpose regional geochemical survey, this study estimated the densities and stocks of organic and inorganic carbon of soil at depths ranging from 0 to 20 cm, 0 to 100 cm, and 0 to 180 cm in oases on the northern margin of the Tarim Basin. Moreover, it delved into the spatial distribution of carbon density. The results of this study are as follows: (1) The compositions of soil carbon pools varied with the soil depth in the study area. At depths ranging from 0 to 20 cm, the organic carbon stocks accounted for 20.66% of the total carbon stocks. With an increase in soil depth, the organic carbon stocks gradually decreased, while the inorganic carbon stocks gradually increased. At depths ranging from 0 to 180 cm, the inorganic carbon stocks represented 85.73% of the total, suggesting that inorganic carbon predominated in the compositions of soil carbon pools; (2) The soil in three depth ranges exhibited organic carbon densities of 1,956.45, 7,913.37, and 119,73.19 t/km², which were all below the national average level, and inorganic carbon densities of 71,722.84, 37,605.54, and 71,914.93 t/km²; (3) The compositions of soil carbon pools varied somewhat across statistical units. In terms of soil types and land use types, the densities of organic and inorganic carbon were higher in fluvo-aquic soil, brown calcic soil, irrigation-silting soil, and solonchak but lower in aeolian sandy soil and irrigated desert soil. Cultivated land exhibited the highest densities of organic and inorganic carbon in the soil, whereas unused and construction land manifested the lowest carbon densities; (4) In terms of topography, undulating mountains manifested the highest soil organic carbon density, whereas alluvial-proluvial plains displayed relatively high inorganic carbon density; (5) The spatial distribution of soil carbon density in the study area was characterized by high organic carbon densities in the Yanqi Basin, medium organic carbon densities in part of Kashgar Delta (western and southern localities and eastern margin), and high inorganic carbon densities in the Aksu area. Overall, under the background of extreme drought, the oases on the northern margin of the Tarim Basin show high potential for inorganic carbon sink, with soil carbon sequestration significantly influenced by soil types, land use types, and geomorphologic landscapes.

Soil carbon pools play a significant role in regulating global carbon balance and mitigating greenhouse gases. Hence, estimating soil carbon stocks is critical for assessing the carbon cycle in terrestrial ecosystems. Based on the soil carbon data obtained from the land quality geochemical survey in the study area, this study estimated the stocks of total, organic, and inorganic carbon of various soil layers in Northwest China using the unit soil carbon amount (USCA) method. It analyzed the content characteristics of organic and inorganic carbon in soil under different soil, land use, and topographic types. The results of this study are as follows: (1) All the soil layers at depths ranging from 0 to 2 m in the study area exhibited total carbon of 10 099.4 Mt, including 1 224.8 Mt in the topsoil layer (0~0.2 m), 5 345.9 Mt in the upper soil layer (0~1.0 m), and 4 753.5 Mt in the lower soil layer (1.0~2.0 m). Inorganic carbon predominated in all the soil layers, with its proportion gradually increasing from top to bottom, whereas organic carbon was principally concentrated in the topsoil layer; (2) The high-value areas of inorganic carbon content were primarily distributed in the Huangshui Valley of Qinghai Province, and the Loess Plateau region covering the Longzhong area of Gansu Province, northern Shaanxi Province, and southern Ningxia Province. In contrast, the high-value areas of organic carbon content were chiefly distributed in the Qilian Mountains; (3) The aeolian sandy soil exhibited the lowest organic, inorganic, and total carbon contents in the topsoil and deep soil layers. The dark loessial soil and the loessal soil showed the highest inorganic carbon content in the topsoil layer. The dark felty soil and the dark loessial soil displayed the highest organic carbon contents in the topsoil and deep soil layers, respectively. Additionally, the dark felty soil had the highest total carbon content in the topsoil and deep soil layers; (4) Forests exhibited the highest organic carbon content in the topsoil and deep soil layers, and the highest total carbon content in the topsoil layer. Grasslands showed the highest inorganic and total carbon contents in the topsoil layer. Cultivated land had the highest inorganic carbon content in the deep soil layer. Bare land manifested the lowest inorganic, organic, and total carbon contents; (5) Mountains displayed the highest organic and total carbon contents in the topsoil and deep soil layers. Loess had the highest inorganic carbon content in the topsoil and deep soil layers. Plains showed intermediate carbon contents generally between those of loess and mountains. Besides, high-altitude areas manifested extremely high organic carbon content.

Based on the water quality analysis of 53 sets of shallow groundwater samples from the eastern start-up area, this study examined the distribution characteristics and genetic mechanisms of fluoride concentration (F^-) using methods such as chemical composition diagrams, proportion coefficient analysis, correlation analysis, and hydrogeochemical modeling. The results indicate that the F^- concentration in shallow groundwater ranges from 0 to 2.85 mg/L, with a coefficient of variation (C_V) of 72.78%, and is unevenly distributed horizontally. Horizontally, shallow groundwater with a high F^- concentration (also referred to as high-fluoride groundwater) is concentrated in the northwest and northern parts of the study area. Such groundwater exhibits complex hydrochemical types, and environments with weak alkalinity, low calcium, and high sodium favor fluoride ion enrichment. The F^- enrichment in shallow groundwater is primarily influenced by climate, terrain, topography, and hydrogeological conditions. Primary mechanisms behind the formation of high-fluoride groundwater include the dissolution of fluorine-bearing minerals in shallow groundwater, as well as the evaporation and concentration of shallow groundwater itself. Additionally, the ion exchange and adsorption processes further contribute to F^- enrichment. The F^- concentration in shallow groundwater in the northern part of the study area exceeds the threshold of the human health risk index and thus should be dealt with. The results of this study provide a scientific basis for the management and utilization of high-fluoride groundwater resources.