Current Articles
PDF 6566KBThe accuracy of excavation response prediction models is generally influenced by various uncertainties, including those related to soil parameters, model uncertainties, measurement errors. Bayesian methods provide a novel way to reduce and/or quantify these uncertainties, and is a natural framework for improving model predictions by systematically integrating prior knowledge with observational data. However, existing Bayesian updating methods typically addressed the uncertainties with soil parameters or/and model biases, while the measurement errors are ignored. Besides, correlations between different excavation stages are also overlooked for mathematical convenience. These simplifications may lead to unreliable predictions in practice. In this study, a novel Bayesian updating method is proposed, which simultaneously incorporates uncertainties in soil parameters, model bias, observational errors, and stage correlations. Two case studies are used to illustrate and validate the method. The results demonstrate that the proposed approach significantly enhances the accuracy of semi-empirical models in predicting excavation responses across different soil types.
Comparative study of the water retention behavior, unsaturated compression behavior and collapse behavior of the sediment depth of 10m and 30m intact loess deposited in the Heifangtai area. To investigate the effect of stress and hydraulic history on the water retention behavior, unsaturated compression behavior and collapse behavior. Mercury intrusion porosimeter (MIP) and scanning electron microscopy (SEM) were used to evaluate the microstructure of two different sediment loess. The result shows that the sample of HFT30 m has a greater air entry value than the sample of HFT10 m, this is because HFT30 m loess has a smaller dominated size for large pores than HFT10 m loess and hence exhibits a larger (air entry value, AEV). For the compression behavior, the HFT30 m loess has larger compression index than the HFT10 m. The yield stress increased with the increase of suction, however, the increased rate of yield stress decreased obviously with the further increase of suction. When the suction was larger than 100 kPa, the yield stress of HFT10 m was smaller than that of HFT30 m; however, with the suction decreasing to smaller than 100 kPa, an unexpected result is observed that the yield stress of HFT30 m was smaller than that of HFT10 m. Overall, the experiment results from compression tests on the two unsaturated loess samples with different depths demonstrated that the effect of suction, saturation and overburden pressure was of great significance to the compression behaviors of unsaturated intact loess. For the collapse behavior, both the HFT10 m loess and HFT30 m loess exhibit that the collapse volumetric strain increases with the increase of net vertical stress, and then decreases slightly with the increase of net vertical stress. Furthermore, the HFT10 m behaved a larger collapse volumetric strain under the same net vertical stress than HFT30 m loess.
In order to predict the criticalmechanical parameters of loess accurately and quantify the uncertainty corresponding to the prediction results reasonably, anunified framework for probabilistic prediction of critical mechanical parameters of loess by machine learning methods is proposed. By fitting probability density function to the bias of the training dataset, a 95% confidence interval for the prediction results is constructed, and the size of the confidence interval reflects the rationality of the prediction results.(Result) Predicting cohesion of loess based on four machine learning methods, namely, random forest, decision tree, extreme gradient boosting and adaptive boosting, the corresponding coefficients of determination R2 reached 0.84, 0.75, 0.81 and 0.79, respectively. The proportion of measurement data included in the 95% confidence interval constructed by the four methods is around 95%. It is shown that the 95% confidence interval obtained from the bias based on the training dataset is relatively reliable and can quantify the uncertainty of the prediction results reasonably. In addition, the cohesion of loess can be predicted accurately using the four machine learning methods.
In the Loess Plateau region, excavation-induced unloading is prone to triggering slope instability disasters, and the water sensitivity and creep characteristics of loess are key factors governing the deformation of excavated slopes. To investigate the creep characteristics and long-term strength of loess, red clay, and composite soil under excavation conditions, this study relies on an excavated slope project in Zhongliang Town, Tianshui City. Through triaxial unloading creep tests, the unloading creep behaviors of loess, red clay, and interface specimens were studied under different moisture contents (12%, 18%, and 24%) and confining pressures (100 kPa, 200 kPa, and 300 kPa). The results show that: (1) All specimens exhibit three-stage creep characteristics, namely attenuated creep, steady-state creep, and accelerated creep. Higher confining pressure makes the soil more sensitive to unloading; an increase in moisture content exacerbates the creep of specimens and renders them more susceptible to failure. Under the same unloading conditions, the creep strain of red clay is slightly lower than that of loess, while the interface increases the creep strain. The interface specimens have the lowest critical unloading amount and are more likely to fracture along the interface. (2) The long-term strength parameters (cohesion c, internal friction angle φ) increase with the rise in confining pressure and decrease with the increase in moisture content. The φ value of the interface specimens is the smallest, reaching 12.54°. (3) The creep strain of the interface specimens increases acceleratively with the intensification of unloading amount, and the interface is a shear-weak zone. This study can provide experimental basis for the stability control and prevention of loess-red clay composite slopes under excavation conditions.
President Xi has emphasized that protecting the ecological environment of the Qinling Mountains holds great and far-reaching significance. The northern slope of the Qinling Mountains is characterized by its unique location, prominent functions, and complex human-land relationships. As a result, the increasing contradiction between the trend of small towns developing in a clustered and networked manner and the single ecological structure with fragile environment has become the weakest link restricting the construction of ecological civilization and high-quality development in the Qinling Mountains. This paper breaks through the barriers of single disciplines and examines the small town clusters and ecological communities on the northern slope of the Qinling Mountains as manifestations of human-land system interactions, based on the concept of "community" in ecology. By comprehensively applying theories and methodologies from geography, ecology, urban and rural planning, and other disciplines, it explores mechanisms and pathways to alleviate conflicts between small town development and ecological protection. The research indicates that: (1) The interaction between small town clusters and ecological communities on the northern slope of the Qinling Mountains involves diverse feedback elements. These elements can be categorized as follows: control elements centered on topography and landforms, which are fundamentally decisive and generally do not undergo significant changes; driving elements centered on small town clusters, which are traditional material factors that promote the development and transformation of small town clusters and determine their development level and scale; and potential elements centered on ecological communities, which are soft environmental factors that drive the development and transformation of small town clusters.(2) The interaction between small town clusters and ecological communities constitutes a complex system involving bidirectional feedback between human activities and natural processes It exhibits hierarchical and progressive characteristics, evolving from low-level and simple states to high-level and complex states. Furthermore, the basic states of mutual nesting, mutual restraint, and mutual promotion, along with nonlinear feedback cycles, enable the three major relationships of coexistence, common loss, and common prosperity in community feedback to evolve and shift.The proposal of the strategy of "mutual feedback and common prosperity, dynamic balance" between small town clusters and ecological communities on the northern slope of the Qinling Mountains is an effective response to the flexibility in the spatial development of small town clusters and the foundational nature of ecological community protection and governance. By identifying "imbalances, " it establishes self-healing and remedial mechanisms for stopping losses, filling gaps, and regulating the system.
To address the diverse characteristics of ecological degradation and the limited efficiency of ecological restoration in the Qinling Mountains, this study investigates a tunable hydrogel material physically crosslinked from polyvinyl alcohol (PVA) and gellan gum (GG). A series of laboratory experiments were conducted to comprehensively analyze the properties of the hydrogel and its effects on reconstructed soil. The results indicate that: (1) the PVA-GG hydrogel exhibits excellent water retention and biodegradability, with a degradation rate of 18.2% after one month in a natural environment; the mechanical properties of the soil improve with increasing PVA content. (2) Soil particles and the hydrogel form binary aggregates, effectively enhancing the soil's water-holding capacity, crack resistance, water stability, and ecological restoration capability by leveraging the hydrogel's advantages. (3) After low-temperature curing, the hydrogel forms a film-like matrix that closely integrates with soil particles through encapsulation, adhesion, and pore-filling, resulting in a compact aggregate structure. As a tunable material, the hydrogel shows promising potential for ecological restoration across various degraded units in the Qinling Mountains.
In order toelucidate the impact of human activities on ecosystem in Xi 'an section of the northern foot of Qinling Mountains, the spatio-temporal evolution of human activities and ecosystem services in the region in the past 20 years was analyzed in the present study.The human footprint index is adopted to quantify intensity of human activities. The InVEST model is used to evaluate the ecosystem services.Their spatial correlation characteristics are analyzed from the perspective of altitude gradient.The results indicate that the area of high and relatively high level human activity intensity increased by 5.99 times and 1.85 times, respectively. The areas where the intensity of human activities increased were mainly concentrated in the plain area of Zhouzhi County, the control zone in front of mountains and the four main valley roads. Urbanization, highway construction and tourism development are the main reasons for the increase of human activities in the northern foot of Qinling Mountains. In the past 20 years, the water conservation, soil conservation and carbon storage service functions of the ecosystem in the northern foot of the Qinling Mountains have been gradually weakened by human activities. Their stability has improved. However, habitat quality has become more sensitive to changes in human activities. The increasing intensity of human activities in the low mountainous areas weakened the stability of habitat quality. The coordination areas between human activities and ecological environment in the northern foot of Qinling Mountains can be divided into four categories: intense conflict area of human-land relationship (< 853 m), unbalance area of human-land coordination (853~1 105 m), sensitive area of human-land coordination (1 105~1 576 m), balance area of human-land relationship (> 1 576 m).
The countryside is the most prominent geographical unit of contradiction between the natural environment, human activities and industrial development, and clarifying the mechanism of human-land-industry system mutual feedback in the countryside is the scientific basis for the practice of the strategy of rural revitalization. The northern foothills of the Qinling Mountains are a typical area for the study of rural human-land-industry system inter-feedback due to its sensitive geographic and environmental characteristics and the rapid development of the neighbouring socio-economies, which pose a major challenge to the coordination of the local human-land. Based on the strength of human-land coupling coordination, geomorphology, land use and tourism resources, this paper divides the countryside into eight types, explores the mechanism of interaction between the elements of the human-land-industry system, summarises the typical patterns, and puts forward a path for system optimisation and prevention of system collapse regulation.It is found that the coupling and mutual feedback are divided into four stages, namely, the period of self-sufficient human-land symbiosis, the period of labor migration and land abandonment, the period of ecological awakening of industrial integration, and the period of reconfiguration of the policy-regulated system. The human-land-industry system in the villages in the northern foothills of the Qinling Mountains is categorised into two typical mutual feedback patterns, namely agricultural characteristics and tourism characteristics, according to the industry or livelihood mode. In order to prevent system collapse, agricultural villages need to build a multi-level resilience defence system, covering disaster early warning, industrial diversification, cooperative network maintenance and dynamic policy regulation; tourism villages need to maintain sustainable development by deeply developing cultural resources, constructing a composite industry of 'tourism+', and resilient development policies to resist homogeneous competition and ecological overload risks development. The results of this study can provide a scientific basis for promoting the coordinated development of human-land-industry systems in ecologically sensitive areas.
Alkaline-carbonatite complexes host more than half of the world's rare-earth element (REE) resources. However, the key factors controlling REE fertility remain uncertain. In this study, we present petrography, whole-rock geochemistry, feldspar and apatite compositions, and Sr-Nd isotopes for the coeval Weishan and Xuezhuang alkaline (-carbonatite) complexes along the southeastern margin of the North China Craton. The Weishan complex is REE-mineralized, whereas the Xuezhuang complex is barren. Syenites from the two complexes show similar isotopic compositions, with (87Sr/86Sr)t = 0.707 297~0.709 173 and εNd(t)~+8.4, suggesting derivation from a common enriched lithospheric mantle source. In contrast, the Weishan complex contains more volatile-bearing minerals (fluorite, barite, apatite, and calcite) and displays distinct geochemical features, including lower CaO and P2O5 contents but higher Sr, Ba, Th, and U relative to Xuezhuang. Feldspar from Weishan shows lower Ca contents (lower anorthite component), and apatite has higher F contents. These mineralogical and geochemical differences indicate a higher volatile budget in the Weishan magma system, which promoted more advanced magmatic differentiation and ultimately facilitated REE mineralization.
Garnets preserved in calc-alkaline volcanic rocks can not only constrain the temperature and pressure conditions of magma crystallization, but also can reveal the evolution history of the host magma, presenting significant genetic implications. However, garnet occurring in calc-alkaline volcanic rocks is extremely rare globally, and there remain substantial controversies regarding the genesis of such garnets. In the Jurassic garnet-bearing dacite from Xishan complex, three genetic types of garnet crystals coexist: magmatic garnet, metamorphic garnet and peritectic garnet. Magmatic garnets mostly occur as single crystals without reaction rims, characterized by low MgO (0.92%~2.37%), CaO (1.21%~2.85%), and MnO (0.82%~1.64%) contents, but high FeO (36.01%~39.82%) contents. Metamorphic garnets develop albite reaction rims. In terms of composition, they are rich in MgO (7.42%~8.46%) and FeO (27.80%~30.99%), and poor in CaO (1.32%~1.33%) and MnO (0.56%~0.60%). For peritectic garnets, the contents of MgO (2.89%~3.55%), FeO (34.57%~37.39%), CaO (2.08%~2.51%), and MnO (0.72%~1.17%) are all between those of the former two types. In terms of rare earth elements (REE), all three types of garnet exhibit strong depletion in light REE(LREE). Notably, magmatic garnets are enriched in heavy REE(HREE), with the most significant Eu negative anomaly (Eu/Eu*=0.004~0.005). Metamorphic garnets are depleted in HREE, and the total rare earth element content (∑REE=64×10-6~72×10-6) is significantly lower than those of the magmatic garnets (∑REE=681×10-6~906×10-6), with a weaker Eu negative anomaly (Eu/Eu*=0.24). The characteristics of rare earth element of peritectic garnets are generally between those of magmatic and metamorphic garnets (∑REE=673×10-6~2 731×10-6; Eu/Eu*=0.02~0.03), and the content variation range is relatively large. Petrographic and mineral chemical characteristics consistently indicate that the magmatic garnets in the garnet-bearing dacite from Xishan complex is a product of early crystallization during magma evolution under high-temperature (740~959 ℃), high-pressure (> 7 kbar), and low oxygen fugacity (logfO2: -23.67 to -12.32) conditions in the lower crust. In contrast, the metamorphic garnets are metamorphic crystal captured from the source rock by volcanic eruption after the former crystallized. The peritectic garnets were formed by partial melting of biotite dehydration during the decompression partial melting process of metapelitic rocks in the lower crust of the study area. Combining with the regional geological context and Hf-O isotopic characteristics of zircon and garnet, this study suggests that the Xishan garnet-bearing dacite may be derived from relatively mature metasedimentary rocks and formed in an extensional tectonic setting.
This study conducted time-dependent compression tests on asperities with different height-to-radius ratios using ultra-hard gypsum. According to Hertz contact theory, the attenuation laws of the elastic modulus of different asperities over time were fitted. Time-dependent closure tests were performed on fresh fracture surfaces of red sandstone and limestone under varying normal stresses. By integrating wavelet analysis, region growth algorithms, and the reference surface method, a novel approach was developed for identifying the mesoscale asperity morphology of rock fractures, and compared the differences in the number, height, and height-to-radius ratio of asperities before and after the experiment. Utilizing Boussinesq's solution, an influence matrix was constructed to account for interactions between asperities. Based on the law of the elastic modulus decaying over time, enabling time-dependent closure calculations for different rock fractures under variable stress conditions. This approach precisely analyzes the temporal evolution of strain, contact area, and contact stress for individual asperity, with simulation results matching experimental data in terms of damage area and creep deformation. The study reveals the pivotal role of asperities with distinct mesoscale morphological features in the time-dependent closure process of rock fractures under compression.
The Mesozoic composite plutons widely distributed in South China are closely associated with rare metal mineralization. However, their genetic mechanisms remain highly debated. To constrain the petrogenetic model of such intrusions, this study focuses on the Jiuyishan composite pluton in southern Hunan, employing integrated whole-rock geochemistry, zircon U-Pb geochronology, and in situ Hf isotopic analysis. Zircon U-Pb dating yields weighted mean ages of 153.0±1.0 Ma, 153.1±0.9 Ma, and 153.8±1.5 Ma for the Shaziling, Jinjiling, and Pangxiemu plutons, respectively, indicating their emplacement during the early Yanshanian period. Whole-rock Sr-Nd and zircon Hf isotopic compositions suggest that these rocks were derived primarily from partial melting of ancient lower crust with minor mantle input, and are classified as A2-type granites formed in an intraplate extensional setting. Based on Rayleigh fractionation modeling of the whole-rock Rb-Sr system, we propose a multi-stage crystallization differentiation model: initial crystal mush underwent melt extraction at 40%-50% crystallinity, with the residual cumulates forming the Shaziling pluton, while the extracted melt subsequently migrated and experienced further differentiation, eventually emplacing as the Jinjiling and Pangxiemu plutons. This model provides new constraints on the magmatic evolution of Mesozoic composite plutons in South China and their implications for rare metal enrichment.
Intense cratonic destruction occurred in the eastern part of the North China Craton(NCC) during the Mesozoic, accompanied by tectonic⁃magmatic⁃metallogenic activities. During the peak period of destruction, high Mg# diorites related to the coeval Fe⁃Cu⁃Au deposits were widely distributed in regions like Luxi⁃Xuhuai⁃Dabie. Although there have been many studies on the genesis of rocks, the connection between the spatial distributions of high Mg# diorites and the evolution of the lithosphere still lacks systematic understanding. This paper selects the eastern part of the NCC as the study area. It comprehensively sorts out and summarizes the petrography, temporal and spatial distribution patterns, and Sr⁃Nd⁃Pb isotopic characteristics of high Mg# diorites, establishes the vertical connection on a large regional scale, and explores the evolution process of the lithosphere in the crust⁃mantle interaction, providing an important basis for magmatic and mineralization processes as well as spatial differences. These high Mg# diorites have consistent island arc⁃like trace element characteristics and enriched Sr⁃Nd isotope compositions, indicating the existence of recycled crustal materials in the source region. Moreover, there are significant spatial variations in the Sr⁃Nd⁃Pb isotopes: from south to north, the (87Sr/86Sr)i of diorites gradually decreased (0.711 7~0.704 3), and the εNd(t) gradually increased (-24.90~-1.77), indicating a trend of decreasing influence of crustal materials on the mantle from south to north, which supports the influence of the subduction of the Yangtze continental Plate during the Triassic on the lithospheric mantle beneath the NCC. On the other hand, the low (87Sr/86Sr)i diorites have low (207Pb/204Pb)i and (208Pb/204Pb)i like those of the ancient metamorphic basement of the NCC, indicating that there are also ancient NCC crustal materials in the source, which may be related to the lower crust delamination. Therefore, the high Mg# diorites reflect that the degree of modification of the lithospheric mantle from south to north by recycled materials (Yangtze crust) gradually weakens, accompanied by the crustal delamination beneath the eastern NCC, and the relative contributions of the subducting Yangtze crust and the NCC ancient crust. This also promotes a better understanding of the formation and evolution of high Mg# diorites, as well as the types and distribution patterns of deposits.
The soil on the ablation zone is degraded in shear strength parameters due to repeated wet and dry cycles, and there is spatial variability in the parameters, both of which are key factors affecting slope stability, while most of the existing studies only consider one of them. For this reason, a new method for analyzing the stability and reliability of slopes that considers both factors is proposed. In this, a parameter random field is simulated using the Karhunen⁃Loève method, and dimensionality reduction is performed using sliced inverse regression, which in turn leads to the construction of an extreme gradient boosting surrogate model based on the augmented whale optimization algorithm. The Three Gorges Reservoir Area Baishuihe landslide is analyzed as an example to explore the effects of degradation of shear strength parameters and spatial variability of the ablation zone on the probability of landslide failure. The results show that: the proposed method can greatly improve the computational efficiency and accurately estimate the probability of landslide failure (Pf); landslide Pf increases with the number of degradation times of the parameters of the fallout zone and tends to stabilize after the fourth time; the spatial variability of saturated permeability coefficients has a small effect on the reliability results when water level changes are not taken into account, while the spatial variability of the effective internal friction angle has a higher effect on the distribution of the factor of safety than that of the effective cohesive force.
Reservoir impoundment-induced valley contraction deformation poses a significant threat to the overall safety of dams, making the prediction of valley width deformation crucial for dam safety management. This study focuses on the valley deformation characteristics of the Xiluodu Hydropower Station and proposes a physics-guided intelligent prediction model to conduct valley deformation forecasting. Based on the analysis of valley deformation patterns, a statistical regression physics-based model (STPM) is developed to quantify the displacement components contributed by individual triggering factors to valley deformation. Building upon this foundation, a machine learning prediction framework (STPM-GRA-LSTM-RF) integrating grey relational analysis (GRA), long short-term memory networks (LSTM), and random forest algorithms (RF) is constructed with displacement components as inputs, incorporating uncertainty analysis in valley deformation prediction. Key findings include: (1) The valley deformation at Xiluodu is primarily caused by viscoelastic deformation, viscoplastic deformation, and effective stress-induced deformation, contributing 67.71%, 29.75%, and 2.51%, respectively, to the total displacement; (2) Compared with the STPM, LSTM, SVM, and XGBoost models, the proposed STPM-GRA-LSTM-RF model delivers higher predictive accuracy, moreover, by incorporating the physico-mechanical mechanisms of valley deformation, it significantly enhances the reliability of the predictions. This research provides valuable insights for deformation prediction and safety control of high reservoir slopes in analogous projects.
Neogene loess-red bed landslides are the most typical and concerning disaster in the Loess Plateau. Their failure mechanisms have become a frontier scientific challenge in the field of engineering geology that requires urgent breakthroughs. This study adopts landslide-controlling slope structures as the analytical framework to systematically examine geological structure types and occurrence characteristics of loess-red bed slopes from two perspectives: stratigraphic structures and geological interfaces. It elaborates on the triple control effects of structural planes in landslide formation-boundary confinement, hydraulic channelization, and mechanical weakening and summarizes landslide types and corresponding failure modes based on structure control effects.The evolutionary mechanisms of landslides under the influence of slope-control structures are comprehensively analyzed. Based on the current state of research, four critical issues are identified: ①How to quantifythe spatiotemporal coupling effects between geodynamic forces and the evolution of structural planes?②How slope structures induce slope instability via cross-scale energy transfer and damage accumulation?③At what critical state of key physical-mechanical parameters do controlling structures trigger slope failures?④How to develop multi-field coupled numerical models integrating structural control mechanisms with landslide kinematic linkages? To address these challenges, this study proposes prioritized research directions, which include dynamic evolution processes and governing mechanisms of sliding-control structures under multi-field coupling effects, mechanical deterioration mechanisms of red bed structure planes driven by water-rock interactions, critical state thresholds and diagnostic criteria for structure control in loess-red beds landslides, kinematic evolution modeling of loess-red bed landslides incorporating structure control effects.
Karst depressions and gully convergence zones offer favorable topographic conditions for the construction of pumped storage power stations, reducing excavation volumes and construction costs. However, intense karstification in these regions introduces complex engineering geological risks. Accurate evaluation of rock mass integrity and karstification degree is therefore critical for safe and efficient project development.(Purpose/Significance) To resolve challenges in the assessment process, such as difficulty meeting index measurement requirements, tedious repetitive work, and insufficient differentiation precision this study takes a pumped storage project in the Zigui karst area as a case. It uses multi-source exploration data (including drilled rock cores, borehole sound wave, and borehole TV images) for mutual complementarity. This reveals the patterns of rock mass integrity and dissolution degree across different strata in the study area, identifies the geological characteristics of each stratum, and assigns stratum labels accordingly. With stratum labels and rock mass acoustic wave velocity data as inputs, the Weighted Random Forests (WRF) method is applied to propose a multi-source information fusion assessment method for rock mass integrity and dissolution degree in karst areas.(Method) The results show that the various types of information at the same location are interrelated, and the differences in the presentation forms of the information only arise from the different exploration methods and information sources; there are significant differences in the geological characteristics of rock masses in different strata of the study area, and the corresponding characteristic manifestations of information from various exploration methods also vary, which affects the evaluation of rock mass integrity and dissolution degree using multi-source exploration information. The proposed method was trained with 1 073 samples and tested with 118 samples. The consistency rates for integrity assessment in the training set/test set reached 95.67%/94.92%, and the consistency rates for dissolution degree assessment in the training set/test set reached 98.02%/97.46%. Moreover, the results are in good agreement with the field targeted re-exploration and verification results.(Result) Compared with traditional methods, the proposed approach achieves higher accuracy, finer resolution, and improved automation, while effectively accounting for stratigraphic geological features. It provides a practical and efficient tool for evaluating engineering rock mass quality in karst-area pumped storage projects.
As human activities and major infrastructure construction extend into high-altitude mountainous regions, the risk of snow avalanche disasters is increasingly severe. Generating avalanche hazard maps is a crucial foundational task. This study, focusing on the Galongla Section of the Zhamo Highway in Xizang, developed a framework for identifying potential avalanche release areas and conducting large-scale hazard assessments by integrating GIS and RAMMS numerical simulation technologies based on DEM data. This study defined a "general scenario" based on standard terrain parameter thresholds and an "extreme scenario" in order to assess the potential maximum hazard. The results indicate that under general and extreme scenarios, 539 and 526 potential release areas were identified, respectively. In the general scenario, the avalanche-affected area was 43.89 km2, accounting for 54.58% of the total study area; under the extreme scenario, the affected area expanded to 53.24 km2, representing 66.20%. Additionally, 16.7% and 25.8% of the highway section in the Galongla area were classified as high hazard level under the two scenarios, with maximum avalanche impact pressures exceeding 580 kPa. The avalanche hazard in the Galongla Section of the Zhamo Highway, Xizang, can be categorized into four hazard level levels: high, medium, low, and none. The high hazard zones are prioritized as target areas for mitigation engineering. This research establishes a portable and efficient avalanche hazard assessment framework by combining GIS-DEM analysis and RAMMS simulation, offering a practical solution for data-deficient high-mountain environments.
Amid intensifying global competition for mineral resources, raising domestic security of strategic minerals requires higher-precision, more explainableexploration. Although petrology, spectroscopy, and mineralogy have amassed large volumes of heterogeneous data, limited cross-source fusion, weak semantic linkage, and misaligned taxonomies hinder their use in exploration. This paper proposes a systematic "Rock-Mineral-Spectrum" Knowledge Graph (RMS-KG) to address these gaps.We integrate remote-sensing imagery, reflectance spectra, mineral characteristics, and geological literature using a hybrid ontology approach that combines top-down domain modeling with bottom-updata construction. The schema covers core concepts in rock taxonomy, mineral attributes, and spectral features. Deep learning and semantic parsing extract entities, attributes, and relations from structured databases, semi-structured reports, and unstructured texts; knowledge is then fused in a graph database to enable semantic linkage, visual querying, and dynamic reasoning.RMS-KG contains on the order of tens of thousands of nodes and edges and includes more than 1, 000 rock-mineral types. It unifies the "rock-mineral-spectrum" semantics, supports mapping spectral fingerprints to minerals and rocks, and enables metallogenic-type inference from mineral assemblages. Two application scenarios, "spectrum-guided mineral identification" and "metallogenic-type inference", demonstrate its effectiveness and interpretability.RMS-KG provides a reusable knowledge substrate and reasoning capability for rock-mineral recognition and prospecting, improving the retrievability, computability, and reusability of geological knowledge and offering a generalizable paradigm for knowledge-centric geological AI.
Regarding the estimation of yield of underground explosions, this study systematically investigates the influence of burial depth, source components(ISO, CLVD, DC, and their combinations), and site conditions on the relationship between the source seismic moment M0(source) and the Lg-wave seismic moment M0(Lg). Based on theoretical synthetic seismogram simulations and Lg-wave spectral inversion methods, we quantitatively calibrated the M0(source)/ M0(Lg) ratio for different test sites. The results reveal significant site dependence: the ratio generally exceeds 0.2 for the North Korean site but falls below 0.2 for the Nevada site. Secondary sources(e.g., CLVD and DC) reduce Lg-wave excitation efficiency in the North Korean site by up to 50%, while increased burial depth weakens P-S conversion effects, further decreasing the ratio. By integrating a seismic source model, we established a yield estimation method based on M0(Lg) and validated it using data from North Korea's sixth nuclear test.
Regional-scale landslides triggered by extreme environmental factors pose a significant threat to life and property safety. Consequently, advancing the automation of regional landslide identification and enhancing the information transparency of potential hazard zones in complex terrain are paramount for the construction of geological hazard databases and effective risk management.Deep learning methods provide an effective solution, overcoming the problem of insufficient automation in traditional methods. However, existing research primarily focuses on optimizing model architecture and improving training strategies, leaving challenges in the effective fusion of multi-source topographic data and the enhancement of cross-regional identification capability. To address these bottlenecks, this paper proposes ResU-CBNet, a deep learning model with robust cross-regional identification capability. The model integrates a hybrid spatial and channel attention mechanism into the neural network and utilizes a residual network to replace the conventional network structure. The model's performance under multi-scale feature fusion conditions significantly outperforms that of single remote sensing data, specifically showing improvements of 2.1% in PA, 2.6% in CPA, 6.9% in F1_Score, and 2.9% in MIoU.Furthermore, the model validates its cross-scene generalization capability across regions with different scenarios, spectral bands, and spatial distributions, achieving PA and F1_Score performances of 92.8%, 91.3% and 83.2%, 80.0%, respectively. The identification results demonstrate a high degree of consistency with the actual regions.The cross-scene identification method presented here offers a valuable reference for intelligent landslide recognition and risk assessment.
The growing disconnect between geological big data and metallogenic knowledge poses a significant challenge to modeling co-associated mineral relationships, underscoring the urgent need for a knowledge-based methodology capable of supporting intelligent analysis. To address this, we propose a data-knowledge synergy-driven approach for constructing knowledge graphs, which integrates domain ontology with the BERT-BiLSTM-CRF model. By leveraging a "knowledge-guided, data-informed" mechanism, the method enables dynamic collaboration between ontology evolution and information extraction, systematically identifying ore deposit features and co-associated relationships from multi-source geological texts and establishing semantic mappings between exploration data and metallogenic knowledge. Experimental results show that entity recognition achieves an F1 score of 83.2%, representing a 15.4 percentage-point improvement over the baseline; entity redundancy is reduced by 5.7 percentage points, markedly enhancing graph consistency. The resulting structured knowledge graph, which comprises 12, 000 nodes and 28, 000 relations, has been deployed in visualization analysis, intelligent question answering, mineralization prediction, and data platform services. This work realizes deep integration of data and knowledge, offering an interpretable and actionable technical pathway for transforming mineral exploration from an experience-driven paradigm to one driven by data-knowledge synergy.
With AI-based automatic cataloging techniques increasingly becoming the mainstream, the limited generalization ability of pre-trained models has also emerged as a widely recognized issue. Through a review of several recent studies and a set of simple tests, this paper seeks to highlight this technological bottleneck and to outline perspectives on future development. On the one hand, the evaluation framework for AI models is in urgent need of updating: the prevailing assessment approaches based on manual annotations exhibit inherent limitations and often lack practical relevance for specific user applications. On the other hand, research on the relationship between training data and model performance remains at an early stage. Although different strategies have been proposed to address generalization issues, systematic discussions of this complex problem are still lacking. This paper aims to provide directional recommendations on potential pathways to overcome these challenges, with the hope of offering useful insights to researchers engaged in AI-based earthquake cataloging.
In order to provide theoretical guidance for the selection and development of geothermal resources in the Nanchang Basin, methods such as rock thermal property parameter testing and borehole temperature measurement were used to calculate the heat flow values in the study area, and the characteristics and influencing factors of geothermal flow were analyzed. The results show that The distribution of geothermal gradient in Nanchang Basin is 1.43-4.55 ℃/hm, with an average of 3.19 ℃/hm; The thermal conductivity of the rock is 2.132-4.300 W/(m·K), with an average thermal conductivity of 2.863 W/(m·K); The current heat flow value in the Nanchang Basin is calculated to be 62.26-96.95 mW/m2.The distribution of heat flow has a significant characteristic of "high in the east and low in the west".Based on the study of the characteristics of geothermal flow in the earth and the occurrence conditions of thermal reservoirs in the area, it is believed that the exploration prospects of geothermal resources in the Changbei-Liantang and Huangxi-Houtian regions are promising.
The Cretaceous tectonic evolution of northern Tibet remains highly controversial, significantly constraining our understanding of plateau uplift and the metallogenic background of world-class Cu-Au resources in this region. To reconstruct the Cretaceous evolution of northern Tibet, we conducted an integrated study on magmatic rocks from Jipusandui, Songxi, and Rutog in western Northern Tibet. Results indicate that the Jipusandui(~120 Ma) and Songxi(~110 Ma) intrusions are Ⅰ-type high-K calc-alkaline granites that underwent complex processes of melting, assimilation, storage, and homogenization, representing products of Meso-Tethys Ocean subduction. The Rutog magmatic rocks(~90 Ma) is characterized by a bimodal volcanic association composed of Nb-enriched gabbro and A-type granite, reflecting post-orogenic extensional tectonics. From 120~110 Ma to ~90 Ma, western Northern Tibet experienced an ocean-continent transition from subduction to collision. Inversion of crustal thickness and crustal contributions based on crust-derived magmas reveals that the crust of western Northern Tibet maintained a normal thickness (~30 km) during 160~100 Ma, but significantly thickened after ~100 Ma, reaching ~60 km by ~90 Ma-exceeding the present-day Iranian Plateau. The peak contribution of crustal materials at ~110 Ma suggests the onset of initial collision. Synthesizing results with regional Late Cretaceous molasse and mélange records, we propose that the Meso-Tethys Ocean underwent a diachronous ocean-continent transition from east to west during the Cretaceous, with the transition in western Northern Tibet occurring between 110 and 96 Ma. Following the closure of the Meso-Tethys Ocean, the Lhasa-Qiangtang collision resulted in pronounced crustal thickening and surface uplift, with an uplift magnitude at least comparable to that of the modern Iranian Plateau. This diachronous ocean–continent transition and subsequent orogenesis elevated the oxygen fugacityof magmatic systems, thereby creating favorable conditions for the enrichment and metallogenesis of giant Cu-Au resources in northern Tibet. From the perspective of magmatic records, this study reconstructs the Cretaceousocean-continent transition and orogenic processes in northern Tibet, providing a representative case study for understanding the orogenesis and metallogenesis in collisional orogens.
Danxia landform is widely developed in southern China, and its red feature is one of the main distinguishing marks of landform. Previous research results have shown that the red color of Danxia landforms is related to iron in the strata, but no in-depth study has been conducted on the state of iron occurrence. In this paper, the mineral composition, geochemistry and iron occurrence of samples from Danxia Mountain in Guangdong Province and Longhu Mountain in Jiangxi Province are studied. The results show that: (1)The Danxia landform stratigraphic samples in Guangdong Province and Longhu Mountain in Jiangxi Province show that the main mineral composition is quartz and feldspar, containing a small amount of clay and iron oxides, with a high degree of consolidation, and some layers contain gravel. (2) Geochemical analysis shows that the content of SiO2 in both is the highest, followed by Al2O3. The content of TFe2O3 (0.81%~1.63%) and FeO (0.08%~0.16%) is lower, indicating that the content of iron minerals is not the main factor causing redness. (3)The iron in the Danxia Formation of Danxia Mountain in Guangdong and Longhu Mountain in Jiangxi mainly exists in the form of clay iron and oxide iron, among which trivalent iron ions (oxides-Fe3+: 41.3% and 44.3%, clay-Fe3+: 49.7% and 47.6%) are much higher than the content of divalent iron ions (clay-Fe2+: 9.0% and 8.1%). All indicate that the content of iron in the Danxia Formation is relatively low, but it may be in the form of an iron oxide coating that exists on the surface of minerals such as quartz and feldspar, causing the formation to appearred.
Portable X-ray fluorescence spectrometer (pXRF) enables rapid and non-destructive in-situ analysis of major and trace element compositions in common rocks. To improve the accuracy of pXRF in geological sample analysis, 39 geological reference materials were selected, including igneous rocks, carbonate rocks, clastic sedimentary rocks, and sediments. Olympus Vanta pXRF was used to analyze their powder pellets, and calibration curves were established based on the correlation between the average values of multiple measurements of the actual element contents and the recommended values of the reference materials. This study confirmed the good precision and accuracy of TiO2, Sr, Zr, Y, Nb, and Cu themselves. It was found that the contents of SiO2 and CaO were significantly affected by matrix effects between carbonate rocks and igneous rocks, as well as between clastic sedimentary rocks, stream sediments, and soils, requiring the establishment of different calibration equations for calibration. In addition, through regression analysis, this study significantly improved the measurement accuracy of elements such as Al2O3, Fe2O3T, MnO, K2O, Rb, Zr, Pb, Zn, Cr, Ni, and Nb. Subsequently, the first 150 m core of the Lingyuan drill hole (YSDP-4) from the Yanshan Scientific Drilling Project was selected as the research object, and the pXRF data before and after calibration were compared with the fusion method. The results confirmed that the calibrated data were more consistent with the whole-rock powder data. The results demonstrate that this method can effectively improve data accuracy and expand the wide application potential of pXRF instruments in rapid core scanning analysis.
During the Jurassic-Cretaceous (J-K) transition, global ecosystems underwent profound changes. Studies on biodiversity during this period indicate that approximately 20% of marine species went extinct in shallow-sea environments, while on land, crocodyliform diversity declined by 55%-75%, and tetrapods and pterosaurs experienced a 75%-80% reduction in diversity. Overall, larger-bodied taxa were disproportionately affected. However, the evolutionary trajectory of terrestrial ecological diversity during this pivotal transition remains poorly constrained.In Asia, stratigraphic successions spanning the Late Jurassic to Early Cretaceous are well developed and have yielded abundant vertebrate fossils, making the region a key area for reconstructing the evolutionary history of terrestrial ecological diversity across the J-K transition. In this study, we compiled occurrence data of vertebrate fossils from the Late Jurassic to Early Cretaceous of Asia and integrated species-level ecological traits for analysis. Ecological classification was established based on habitat, diet, and body size, with body size data measured to refine trait differentiation. Resampling methods were applied to correct for sampling bias and uneven sample sizes.The results show that large-bodied saurischian dinosaurs (by approximately 50%-80%), turtles (about 40%-50%), and mammals (about 60%-70%) experienced marked declines in diversity across the J-K boundary, whereas freshwater fishes and some other reptilian groups were less affected. Analyses of ecospace structure reveal substantial adjustments in both species diversity and functional structure within Asian terrestrial ecosystems during the J-K transition. Furthermore, responses to environmental factors varied markedly among clades: overall, vegetation changes appear to have promoted increases in aquatic diversity, whereas climatic warming and increasing aridity exerted strong suppressive effects on the diversity of terrestrial groups.
