Rivers are one of the important geological forces that shape the land surface. Understanding the development history of big rivers is helpful for us to improve the efficiency of using big rivers, so as to promote the development of human society. The Yellow River is the mother river of China. Researchers have conducted extensive research on it in the past hundred years, however, there is no clear answer to the question of when it was formed and the specific evolution process. In view of this, on the basis of our previous research results, this paper extensively collected and collated the data published by domestic and foreign researchers to reconstruct the evolution process of the Yellow River during the Cenozoic era. The results show that: The proto-Yellow River, which flowed along the plateau boundary, developed in the northeastern part of the Tibetan Plateau in the Paleogene. In the Miocene, the eastern Qilian Mountains developed parallel rivers flowing into the depocenter of Longxi basin, and large rivers with different flow directions appeared in the north and south parts of the Jinshan Gorge. The Yellow River had not yet appeared in the Bohai Bay Basin and the South Yellow Sea Basin during the Miocene. The Yellow River entered the stage of segmental evolution in this time. In Pliocene, the Yellow River had already connected Xining basin, Lanzhou basin, Yinchuan Basin and Hetao basin, while Guide basin and Gonghe basin still had internal flow system. At this time, the upper reaches of the Yellow River had not penetrated into the deeper hinterland of the northeast section of the Tibetan Plateau. The Jinshan Gorge still has the Yellow River flowing south into Weihe basin and north into Hetao basin, respectively. The Sanmenxia basin is still occupied by large ancient lakes, and the Yellow River material does not appear in the Bohai Bay basin and the South Yellow Sea basin. Pliocene is an important turning stage for the Yellow River to complete its final connection. In the early Pleistocene, the Yellow River connected the upper, middle and lower reaches, similar to the current connection between the Tibetan Plateau, the Loess Plateau and the North China Plain. The Yellow River flowing east to the sea was formed at this time. In the Late Pleistocene, under the influence of climate change, all sections of the Yellow River entered a new stage of evolution, and some of the river channels were occupied by ancient lakes again. In the period of abundant precipitation, the river channels were connected again. The combination of structure and climate has an important influence on the development of the Yellow River. The maintenance of the high landform in the hinterland of river system is one of the preconditions for the development of river system. In particular, the uplift of the eastern Qilian Mountains in Miocene and Pliocene shaped the basic form of the Yellow River on the interior of the Tibetan Plateau. The uplifting of the Helan, Yin and Qinling Mountains around the Ordos Plateau have shaped the deep graben of Yinchuan, Hetao and Weihe River, which has become an important factor limiting the full transmission of the Yellow River in dry weather periods. The Quaternary is a period of frequent climate change in East Asia, and the final connection of the Yellow River is closely related to the process of climate change. Conversely, the study on the formation and evolution of the Yellow River can effectively reveal the tectonic evolution and climate change of the Cenozoic in North China.

地震引发的地表裂缝对于揭示断层活动、解析地震构造特征以及震后灾害评估具有重要意义.本研究结合现场采集的高精度无人机（UAV）数据和深度学习技术，对2025年西藏定日M_S6.8地震的地表裂缝特征进行了自动识别与分析，揭示了裂缝的走向规律，并与InSAR变形数据进行了对比验证.基于现场无人机（UAV）获取的高分辨率影像，构建ResPSP-CBAM模型进行智能识别，成功提取了震后区域的地表裂缝分布.该模型集成了ResUNet的残差结构、空间金字塔池化（PSP）模块和卷积块注意力机制（CBAM），显著提高了地表裂缝识别的精度与鲁棒性.分析结果表明，ResPSP-CBAM模型在准确率、精确度、召回率、F1分数上表现优越，其相应数值分别为0.927、0.829、0.779和0.802，识别出的地表裂缝走向与InSAR解译的地表变形方向高度一致，进一步验证了该方法的有效性.本研究构建的ResPSP-CBAM深度学习模型显著提高了地震地表裂缝智能识别的精度和效率.识别出的地表裂缝包含原生和次生裂缝，且主要以断层破裂引起的原生裂缝为主，总体呈南北走向分布，与登么错断裂带走向高度一致.表明识别区域地震地表裂缝与断层的活动性密切相关.本研究为地震地表裂缝的智能识别提供了新的技术手段，对深入理解震源断层的构造特征提供了有力支持，同时为地震预测、预警及震后灾害评估提供了重要的科学依据.

On January 7, 2025, a M _S6.8 earthquake struck Dingri County in the Shigatse region of Xizang. In response to the event, it conducted field investigations and high-precision UAV aerial surveys, which indicate that the rupture extends from the central-northern segment of the Dengme Co (or Dingmu Co) fault to the eastern shore of Yangmudingcuomu Lake and the northern bank of the Pengqu River, totaling approximately 36.5 km. The surface rupture predominantly follows existing fault structures, with the southern end exhibiting multiple parallel faults and surface scarps up to 4 km wide. The rupture patterns are complex, combining seismic scarps and extensional fissures of varying dimensions. Notably, the northern section displays larger scarps, with maximum displacements reaching approximately (265±27) cm. Scarp heights vary significantly along the rupture, with minimal vertical displacement at north and south end. Our preliminary findings suggest that the surface rupture length and maximum displacement align closely with global empirical relationships between earthquake magnitude and rupture characteristics. Additionally, the characteristics of the Dingri earthquake’s surface rupture offer valuable insights for studying the geomorphic evolution of single-event seismic scarps and their long-term cumulative effects.

The moisture content of the soil is the main factor affecting the quality of fine-grained soil. Rapid recognition of soil surface moisture content is an urgent need for developing intelligent monitoring and construction technology in agricultural and geotechnical engineering. In order to overcome the limitation that traditional water content measurement or monitoring methods cannot meet the real-time nondestructive monitoring of soil surface moisture content, an intelligent moisture content recognition algorithm based on the image is developed. Firstly, we collected surface photos of 4 different types of soils under different moisture contents in the laboratory and obtained a high-quality sample library of more than 1 400 pictures, which laid a data foundation for machine learning model construction. Then the classical convolutional neural network is used to learn the image dataset of soil moisture content, and the intelligent recognition model of soil moisture content is established. The model comparison results show that the model based on ResNet34 architecture has the best moisture content recognition effect, and the average error of moisture content prediction on the test set is about 2%. This model basically meets the requirement of real-time nondestructive monitoring of soil surface moisture content and can provide an essential means for the development of intelligent monitoring and construction technology in agricultural and geotechnical engineering.

The understanding of the formation and evolution of geothermal water is an important basis for the scientific and rational utilization of regional geothermal resources. In this paper, the hydrochemical characteristics, geothermal storage temperature and recharge source of geothermal water in Guantang area on the east coast of Hainan were systematically revealed by using hydrochemical and isotopic analysis methods combined with regional geological structure characteristics, and a conceptual model of geothermal water cycle evolution in Guantang area was constructed. The results show that the hydrochemistry of geothermal water was mainly HCO₃·SO₄-Na type, and its components mainly came from silicate mineral dissolution and deep gas components. Geothermal water was mainly originated from meteoric precipitation, and the recharge area was most likely located at the altitude of 1 122.2-1 569.4 m. There was a significant mixing effect between geothermal water and shallow groundwater. Under the condition of steam loss before mixing, the mass percentage of steam loss from geothermal water was 18.2%-25.2%, and the initial temperature of geothermal water was 190.4-217.8 ℃, and the mass percentage of cold water mixing to geothermal water was 66.8%-80.8%. The geothermal water level decreased greatly with the increasing geothermal water exploration in this area in recent decades, so the increase of shallow groundwater supply may be the crucial process resulting in the decrease of geothermal water temperature in this area.

Global warming will lead to the thaw of permafrost degradation. The permafrost degradation can cause serious damage to engineering and infrastructure in permafrost regions, it can also change water, energy and carbon cycle between the land and the atmosphere and further create a feedback to climate change. Most permafrost is the ground below the earth’s surface and thus is difficult to be detected. There are many literatures used the terms of permafrost region and permafrost area ambiguously, and the incorrectly use of the two terms can lead serious mistakes in permafrost change and carbon budget calculation. Recently, there are many advances in remote sensing and modelling of permafrost. Here it systematically reviews the literatures to clarify the area of permafrost regions and the actual permafrost area in the world as well as in China. The most recent studies show that the permafrost region in the northern hemisphere is about 21×10⁶ km², and the actual permafrost area is about 14×10⁶ km². For the Qinghai-Tibetan plateau in China, the permafrost region occupies an area of about 1.50×10⁶ km², and the actual permafrost area is about 1.06×10⁶ km². More data are required to quantify the actual permafrost area in other permafrost regions in China.

Qiongdongnan basin is a newly discovered gas hydrate exploration target area in China. There are two types of gas hydrate (i.e. pore hydrate and chimney hydrate), which are characterized by different seismic reflection characteristics. However, there are limitations in using bottom simulating reflections (BSR) in seismic profiles to identify gas hydrate in deep water area of Qiongdongnan basin. For the identification and saturation prediction of gas hydrate in Qiongdongnan basin, AVO forward modeling and broadband seismic inversion analysis are carried out in this study. The elastic parameter characteristics of natural gas hydrate were clarified, and geophysical identification methods for gas hydrate were established. Then, isotropic and anisotropy quantitative saturation evaluation methods were established for pore and chimney hydrates respectively. Based on broadband seismic inversion and saturation prediction using high quality seismic data, this study successfully identified pore hydrate and chimney hydrate in Qiongdongnan basin, and delineated gas hydrate enrichment areas. Three hydrate deposits with maximum saturation above 50% have been selected, the prediction effect of which is further confirmed by drilling exploration. Our results contribute to the selection of hydrate drilling stations and reduce the risk of hydrate drilling in Qiongdongnan basin, and the research method has guiding significance for hydrate exploration in other similar basins.

The V-shaped conjugate strike-slip fault system is defined as strike-slip faults with obtuse conjugate angles, whose opening side has an acute angle between the V-shaped faults, pointing to the direction of maximum extension. Previous studies on V-shaped conjugate strike-slip faults mostly focused on their development background and associated dynamic mechanisms. However, few literatures exist to comprehensively review the geometry and kinematics of V-shaped conjugate strike-slip faults. Here, we firstly summarize previous findings on the geometry, kinematic characteristics and formation mechanisms of existing V-shaped conjugate strike-slip faults, and then select the V-shaped conjugate strike-slip faults in the southeastern Tibetan Plateau for a case analysis. The characteristics of V-shaped conjugate strike-slip faults in the western United States, central and western Eurasian plate and central Tibet show a negative relation among conjugate angles and corresponding fault slip rates and fault lengths. The four formation mechanisms of the V-shaped conjugate strike-slip faults are 1) the fault planes experienced rotation after their formation, 2) the faults were reactivated along preexisting structurally weak zones, 3) the faults followed the paired general shear model and 4) the fault evolved according to the maximum-effective-moment criterion. Integrating analyses of geophysical data, elevation difference and geometric characteristics, we infer that the development of V-shaped conjugate strike-slip faults (Batang-Litang and Derong-Xiangcheng faults) in the Chuan-Dian block in the southeastern Tibetan Plateau, is consistent with gravitational spreading of the Tibetan lithosphere under the paired general shear model. This provides important insights for understanding the continuum crustal deformation in the southeastern Tibetan Plateau.

Due to the complex interaction between geological environment and human activities, the Chinese Loess Plateau (CLP) is prone to frequent landslides. It is urgent to carry out landslide vulnerability assessment, selecting suitable influencing factors and training models. In this study, the CLP was taken as the study area. Based on field landslide survey and data collection, an evaluation system including topography, basic geological environment, meteorology and hydrology, human activities, soil physical and chemical properties, and vegetation coverage was built. The information model (IV) was used to connect the random forest model (RF) and convolutional neural network model (CNN) to build coupling models IV-RF and IV-CNN, and landslide susceptibility evaluation research was carried out. The results show that the accuracy of the coupling model (IV-RF, IV-CNN) is higher than that of the independent model (RF, CNN), and the AUC values of the four models are 0.916, 0.938, 0.878, and 0.853, respectively. The IV-CNN has stronger prediction ability and accuracy. The areas of extremely high, high, medium, low, and extremely low vulnerability areas in the IV-CNN model account for 8.78%, 7.47%, 15.34%, 19.82%, and 47.87% respectively, which are mainly distributed in the mountainous and loess hilly areas with complex geological environment and strong human activities in the south and east of the loess plateau. Slope, erosion type, landform type, clay content and distance from the road rank in the top five in the contribution rate analysis, and are the main control factors affecting the landslide development. The purpose of this study is to provide reliable scientific basis for the prediction and prevention of landslide disasters in the CLP, deepen the modeling idea for landslide vulnerability evaluation research, and optimize the uncertainty of independent model evaluation results.

As the importance of hydrogen continues to grow, large-scale hydrogen storage is receiving increasing focus. In this paper it extensively examines the classification, advantages, and drawbacks of underground hydrogen storage facilities through comprehensive literature research, providing a theoretical foundation for the implementation of such storage systems. Furthermore, it elucidates the interactions between hydrogen and minerals, and highlights the hydrogen adsorption characteristics of clay minerals and coal seams, offering novel insights into addressing challenges related to large-scale hydrogen storage and low-cost adsorption-based storage. The study findings reveal that (1) hydrogen storage facilities are primarily categorized into salt cavern storage, depleted oil and gas reservoir storage, and aquifer storage, with salt cavern storage currently being the most favorable option; (2) variations in temperature, pressure, concentration of fatty acids, and organic acid carbon number affect the hydrogen wettability of minerals, thus impacting caprock sealing capacity; and (3) certain clay minerals, coal seams, and other materials can adsorb hydrogen, presenting potential avenues for new underground hydrogen storage materials. Based on the above research and analysis, the main problems existing in underground hydrogen storage are pointed out, and the future development prospect of underground hydrogen storage is prospected, in order to provide reference for the site selection and implementation of underground hydrogen storage. The feasibility of underground porous material as a new large-scale hydrogen storage material is briefly summarized, in order to contribute to the search for diversified and suitable hydrogen storage materials.

On the basis of reviewing the current situation and history of groundwater resources evaluation in the Pearl River basin, this paper discusses the basic principle and basic database of SWAT distributed hydrological model. We divided the Pearl River basin into 129 four-level groundwater systems, and fully considered the parameter sensitivity of karst, bedrock fissures and pore aquifer media. Based on the monthly measured runoff of 9 hydrological stations from 2008 to 2016, we carried out parameter calibration and model calibration, calculated the rainfall assurance rate in different years according to the rainfall from 1957 to 2017, analyzed and evaluated the multi-year groundwater resources in the Pearl River basin. Finally, we carried out the inversion of rainfall infiltration coefficient and groundwater recharge model parameters. Through this assessment, the average recharge of the Pearl River basin from 2010 to 2016 is 148.802 billion m³, and the total recharge of groundwater in extra dry years (2011), normal years (2010) and high water years (2016) are 71.949 billion m³, 144.682 billion m³ and 178.187 billion m³ respectively, of which the rainfall in high water years is about 1.7 times that in extra dry years, the groundwater recharge is 2.48 times, karst. The total annual recharge of bedrock fissure and pore water bearing medium is 52.991 billion m³, 44.513 billion m³ and 51.298 billion m³ respectively. Through parameter inversion, the average annual groundwater recharge modulus is 10.83 L/s·km² and the rainfall infiltration coefficient is 0.246, in order to provide data support and scientific services for groundwater development, utilization, treatment and protection in the basin.

The Mesozoic in the Bohai Bay basin lacks unified stratigraphic framework. Based on logging data, sporopollen assemblage and age analysis of clastic and volcanic rocks, the stratigraphic subdivision and correlation of the Mesozoic were studied. The Mesozoic strata can be divided into eight stratigraphic units. The Triassic continental red strata develops the Aratrisporites-Triassisporis-Chordasporites assemblage, and the maximum depositional ages range from (241.4±6.8) Ma to (248.2±1.9) Ma. The Cyathidites-Osmundacidites-Cycadopites-Classopollis assemblage forms in the Jurassic coal-bearing strata, and the maximum depositional ages of the strata are from (173.4±1.1) Ma to (187.2±4.8) Ma. For the Jurassic gray sandstone strata, the maximum depositional ages range from (155.3±1.3) Ma to (164.4±3.8) Ma and the Cyathidites-Osmundacidites-Deltoidospora-Cycadopites assemblage are developed. The Jurassic red sandstone and gravel strata develop the Classopollis-Abietineaepollenites assemblage. In early stage of early Cretaceous, the red clastic strata with basic volcanic rock and the Cicatricosisporites-Lygodiumsporites-Classopollis assemblage formed. The volcanic rocks ages are from 140 to 132 Ma. For the gray clastic strata with intermediate-basic volcanic rock in the middle part of the lower Cretaceous, the Lygodiumsporites-Cicatricosisporites-Osmundacidites-Cycadopites assemblage and the volcanic rocks of the ages from 130 to 116 Ma are developed. The clastic strata with intermediate-basic volcanic rock in the upper part of the lower Cretaceous develop the assemblage of Cicatricosisporites-Schizaeoisporites-classopollis-Piceites, and the volcanic rocks ages range from 119 to 103 Ma. The Schizaeoisporites-Ephedripites-Classopollis assemblage forms in the Upper Cretaceous red clastic strata with acidic volcanic rock ranging from 76 to 69 Ma in age.

为理解洛阳盆地龙山-二里头文化转变与水文气候的关系问题，整理了这两个文化时期聚落的位置和规模信息（数量、等级和最大聚落占地面积）、全新世古气候记录和4 000 a BP左右的古洪水事件.分析结果表明：（1）在龙山文化时期（4 900～4 000 a BP），聚落的发展壮大可能与中全新世晚期稳定适宜的气候条件有关.然而，在龙山-二里头文化的转变阶段（4 000～3 750 a BP）,4 200 a BP气候突变导致极端洪水发生，对众多聚落产生了严重威胁，迫使盆地内泛滥平原区的居民迁往周围高地或外地.（2）在二里头文化时期（3 750～3 500 a BP），极端洪水不再发生，盆地内优越的地理位置和地貌条件为大规模的农业生产和聚落建设提供了广阔的空间.本研究成果对于阐明洛阳盆地龙山-二里头文化演变与水文气候的关系，具有重要的科学意义.

Natural fracture is an important reservoir space and main seepage channel of organic rich shale oil and gas reservoir, which affects the enrichment, preservation, single well productivity and development effect of shale oil and gas. The research on the development law of natural fracture is of great significance to the exploration and development of organic rich shale oil and gas. Based on the research results of marine and continental organic rich shale fractures in recent years, this paper summarizes the latest progress in the genetic types, development characteristics, main control factors, evaluation and prediction methods of organic rich shale natural fractures, and finally discusses the key research directions of organic rich shale natural fractures in the future. The natural fractures of organic rich shale can be divided into three categories and six sub categories: tectonic fractures, diagenetic fractures and abnormally high-pressure-related fractures. The main fracture types are intraformational open fractures, transformational shear fractures, bed-parallel shear fractures and bed-parallel lamellated fractures. The development degree of shale tectonic fractures is mainly controlled by brittle mineral content, organic matter content, high brittle shale layer thickness, structure, formation dip angle and fluid pressure. The formation and development of bed-parallel lamellated fractures are mainly affected by organic matter content, lamina type, lamina number, lamina thickness and later tectonic uplift. Due to the differences in mineral composition, lithofacies changes and thermal evolution of organic matter caused by different sedimentary environments between continental shale and marine shale, the development characteristics of fractures between continental shale and marine shale are obviously different. Compared with marine shale fractures, the distribution pattern of continental shale fractures is more complex, the scale of tectonic fractures is smaller, and the development degree of cross layer shear fractures and bedding shear fractures is low. At present, the evaluation and prediction of shale fractures are mainly carried out with the help of the existing conventional research methods of tectonic fractures in low-permeability tight reservoirs. How to combine geology, geophysics and machine learning to form a classification evaluation and prediction method suitable for different scales and types of shale fractures according to the characteristics of small scale of shale fractures and development of bed-parallel lamellated fractures, It is very important to improve the evaluation and prediction accuracy of shale fractures and better guide oil and gas development. The development law of deep organic rich shale fractures, the influence of natural fractures on hydraulic fracturing fractures, and the three-dimensional geological modeling of complex fracture network system integrating multi-scale, multi occurrence and multi Genesis shale fractures will also be important problems to be solved in the research of shale fractures in the future.

China’s continental shale oil resources have great potential, but the shale reservoir has the characteristics of high clay mineral content, strong heterogeneity and low ground energy, which seriously restricts the efficient exploration of continental shale oil in China. Focusing on the key links such as sweet spot optimization, horizontal well drilling, stimulated reservoir volume and oil test for production, comprehensively utilizing the coring, logging, logging and analysis and test data of JYY1 Well system, innovative research has been carried out on geology-engineering integration. The results show that the shale of Qingyi Member in Songliao Basin can be divided into two lithofacies types: high TOC bedding clayey shale and medium TOC laminated felsic shale. The former is mainly geological sweet spot and the latter is mainly engineering sweet spot. In this study, the integrated “double sweet spot” evaluation standard of geology-engineering has been established. Based on our horizontal well design concept of “drilling sand and fracturing shale”, the 1.94 m thick double sweet spot target layer was selected, and the technical breakthrough of 1 252 m horizontal drilling of ultra-thin target layer and 100% penetration rate was achieved by adopting the three-dimensional geophysical geology-engineering integration ultra-thin target layer guidance technology. The four advantages of supercritical CO₂ are rock breaking, dissolution, oil displacement and energy increase. The supercritical CO₂ + large-scale hydraulic sand carrying composite fracturing process and pressure controlled energy storage flowback technology have been innovated, which has realized the large-scale volume transformation of continental shale formation of JYY1 Well, and achieved a breakthrough in 16.4 m³/d high and stable shale oil, resulting in a geology-engineering integration method, technology, process and various parameters. It has reference significance for the efficient exploration and development of continental shale oil in Songliao Basin and the same type.

Landslide susceptibility mapping (LSM) is the foundation and critical part of landslide risk assessment. The bibliometric analysis of LSM literature can be applied to quantitatively analyze the research progress and development trend. The result will provide references for geological hazard risk assessment in China. In this study, based on the Web of Science and CNKI databases, the CiteSpace visual knowledge graph analysis tool has been used to carry out bibliometric analysis of LSM literature from 1985 to 2022. Moreover, the LDA analysis has been conducted on the abstract to subdivide the research in this field. The results show that: (1) LSM is still a research hotspot at present. In China, there are a large number of studies and international cooperation about LSM. (2) Four of the top 10 authors in the number of published papers on LSM are from China. The institution that has published the most papers on LSM is the Chinese Academy of Sciences. The Chinese Journal of Geological Hazard and Control is the most popular Chinese journal and the Natural Hazardsis the most popular English journals to publish LSM papers. The research on the subject of LSM has been substantially funded by the National Natural Science Foundation of China and the National Land and Resources Survey Project. (3) In the past five years, machine learning models (including deep learning, etc.) have been widely used as the most popular LSM models. (4) In order to achieve the simplification and intelligence of landslide susceptibility modeling and to improve the accuracy and practicability of the LSM results, the following parts of LSM, including the landslide inventory, conditioning factors, assessment unit, assessment model, connection methods and accuracy verification, need to be deeply explored in further studies.

To investigate the depositional environment evolution and controlling factors of a mountainous river estuary during the Holocene, in this study it selected a drilling core (MLX-S) from the river mouth of Mulanxi, a tidal-controlled mountainous river in central Fujian Province, China. The core was subjected to radiocarbon dating, optically stimulated luminescence dating, grain size analysis, XRF core scanning, and ICP-OES elemental measurements. During the Holocene, the Mulanxi estuary experienced a transition from fluvial facies to riparian marsh facies, shallow marine facies, and finally to tide-controlled estuarine facies. The sedimentary environment evolution was primarily influenced by relative sea-level changes in the study area. The relative sea level along the coast of central Fujian is around 9.0 ka and about -20 m. Subsequently, the marine transgression reached the strongest and the maximum sea level occurred around 5.8 ka, followed by a slow decline till the present. High-resolution XRF core scanning data, after calibration of element relative intensities, can indicate sediment provenance and paleoenvironmental changes. The downcore variation of K/Ti ratio suggests that since 8.8 ka, the sediments in the Mulanxi Estuary were mainly derived from a mixture source of local basin erosion and fine offshore material from the Changjiang (Yangtze River). The core Fe/Ti ratio reached its peak in four strong East Asian winter monsoon periods of about 3.6 ka, 4.3 ka, 5.4 ka and 6.0 ka, corresponding to periods of strong East Asian winter monsoon, which is consistent with the sedimentary records on the East China Sea inner shelf and tightly related to the changes of Min-Zhe coastal current intensity. This study reveals the typical river-sea interaction in the tide-controlled mountainous estuaries along the southeastern coast of China, and suggests the complex source-sink transport processes of Changjiang-derived fine sediments on the continental shelf driven by the Holocene monsoon and sea level changes. This study also promotes the potential application of XRF core scanning in the continental margin where has the unique sedimentary dynamics and environmental changes.

Anatexis is significant for the evolution and rheological properties of the continental crust， linking metamorphism， tectonic deformation， and magmatic activities in deep crust. According to the presence or absence of free water， anatexis can be divided into water-fluxed melting and dehydration melting. Dehydration melting involves the breakdown of hydrous phases， such as muscovite， biotite or amphibole. Generally， dehydration melting reactions begin at ~650 ℃ and produce melt， peritectic garnet and K-feldspar. Melt resulting from dehydration melting is water unsaturated， characterized by high Rb， high Rb/Sr ratio， high ⁸⁷Sr/ ⁸⁶Sr ratio， and low Sr， Ba， Ca. Melt productivity of dehydration melting reaction is controlled by P- T condition and water content in rock， and it is only possible to produce a large amount of melt at granulite facies. Water-fluxed melting reactions involve the presence of free water. Its most significant feature is that the temperature required is relatively low， and can produce voluminous melting at amphibolite facies. The resulting melt is water-saturated or water unsaturated that can extract from the source. Melt derived from water-fluxed melting reactions has high Sr， Ba， Ca， low Rb/Sr ratio and low Rb. Anatexis has a profound impact on the thermodynamic and rheological properties of the rocks. At the same time， the movement of melt out of the lower continental crust promotes the chemical differentiation of the crust and forms a wide range of leucogranite， playing a crucial role in the origin， reworking， and stability of the continental crust.

本研究基于密集远场地震台阵对2025年西藏定日6.8级地震的破裂过程进行快速测定，并结合地震破裂过程对地震烈度分布及潜在人员伤亡进行了快速评估.首先采用反投影技术获取震源区能量释放的时空分布特征，揭示震源破裂的动态演化过程；然后结合破裂过程与基于断层最短距离的地震动参数衰减模型，快速计算地震烈度的空间分布，明确灾害影响范围和强度；在此基础上，利用人员伤亡评估模型对地震可能造成的人员伤亡进行初步估算.研究结果表明，基于密集台阵的地震烈度快速评估方法在震后灾情快速评估方面具有较高的可靠性和实际应用价值，可以为政府决策、应急指挥和救援部署提供重要参考.

In most of the landslide susceptibility prediction (LSP) models, the landslide-non landslide spatial datasets are divided into training/testing datasets according to the principle of spatial random, however, this spatial randomness division inevitably introduces uncertainties into LSP modelling. Theoretically, LSP modelling is based on past landslide inventories to predict the spatial probability of future landslides, which has significant time series characteristics rather than only spatial random characteristics. Therefore, we believe that it is necessary to divide spatial datasets into the model training/testing datasets based on the time series of landslide occurrence. Taking Wencheng County in China as an example, 11 types of environmental factors and 128 time-accurate landslides are obtained; Then, the landslide and non-landslide samples connected with environmental factors are divided into two different types of training/testing datasets according to the principles of landslide time series and spatial random, respectively. The division ratios of training/testing datasets are set as 9∶1, 8∶2, 7∶3, 6∶4 and 5∶5, respectively, to avoid the influences of different ratios on the LSP results. Thus, the training/testing datasets under 10 combined working conditions are obtained. Finally, several typical machine learning models, such as Support Vector Machine (SVM), Multi-Layer Perceptron (MLP) and Random Forest (RF), are then trained and tested to perform LSP and analyze their uncertainties. Results show that: (1) The LSP uncertainties performed by the time series-based SVM, MLP and RF models are slightly lower than those by spatial random-based models, which verifies the feasibility of dividing by time series; (2) The time series division of training/testing datasets is actually a “deterministic” case among the spatial random division, which is more consistent with the actual situation of landslides. Of course, it is also feasible to carry out spatial random division for training and testing datasets when lacking landslide occurrence time.

Accurate displacement prediction plays an important role in landslide early warning. However, the majority of the existing data-driven models focus on single-point modeling based on time series data which cannot consider the deformation correlation in the whole slope. To overcome this drawback, this study proposed a spatial-temporal attention (STA) mechanism-based deep learning model which combined the convolutional neural network (CNN) with the long short-term memory (LSTM) neural network. A convolutional block attention module (CBAM) combined with CNN was developed to extract the spatial deformation characteristics of the slope. A temporal attention module and LSTM model were used to learn the significant historical information from the input external conditions time series data. The model also allowed to output the tempo-spatial attention weights to reveal the tempo-spatial characteristics of landslide deformation. The Paotongwan landslide with step-like behavior displacement in the Three Gorges Reservoir Area (TGRA) of China was used to validate the model performance. The results show that, the root mean square error (RMSE) and the mean average percentage error (MAPE) of the STA-CNN-LSTM model decreased 9.28% and 13.88%, respectively, compared with grey wolf optimization optimized support vector machine (GWO-SVM). The attention weight results calculated by STA-CNN-LSTM demonstrate that rainfall had a larger impact on the deformation of the Paotongwan landslide at the beginning of the monitoring while the influence of reservoir water level decreased with ongoing of the monitoring.

<正>青藏高原起因于印度板块与欧亚板块的持续碰撞，不仅成为世界的第三级，也成为全球地壳变动最为强烈、活动断层最发育、大地震最频发的地区之一，更成为不同类型地震遗迹的天然博物馆.自2000年以来，青藏高原先后发生过2001年可可西里M_W7.8级地震、2008年西昆仑于田M_W7.1级地震、2008年汶川M_W7.9级地震、2015年尼泊尔M_W7.8级地震、2021年玛多M_W7.3级等大地震，以及多次矩震级M_W接近7级强震，例如2010年玉树M_W6.9级地震、2022年门源M_W6.6级地震.其中，发生在青藏高原周缘的地震人员伤亡和财产损失巨大，但地处高原腹地无人区的2001年可可西里地震和2008年于田地震则灾害损失很轻.

Geotechnical engineering named entity recognition is an important prerequisite and the work foundation for geotechnical information mining and knowledge Graph. Aiming at the recognition and classification of named entities in geotechnical texts, this article first designs and constructs a named entity corpus of geotechnical engineering according to Standard for Fundamental Terms of Geotechnical Engineering (GB/T 50279-2014) and other national industry standards; and based on deep learning technologies, a named entity recognition and classification deep learning model GENER is proposed for geotechnical engineering text. In GENER, the distributed representation learning of geotechnical engineering text features is realized based on the BERT pretrained language model; the geotechnical engineering text context feature encoding is achieved based on the BiGRU context coding layer; and based on the label decoding layer of CRF, the context features are decoded to generate the label sequence of geotechnical engineering named entity. Finally, based on the geotechnical engineering corpus, the GENER model is experimentally analyzed. comparing with other deep learning models for named entity recognition based on pretrained language models, the GENER model has better performance. The precision reaches 90.94%, the recall reaches 92.88%, the F1-score reaches 91.89%and model training speed increased by 4.735% respectively.Experiments show that compared with BiLSTM-CRF and CNN-BiLSTM-CRF models, this model is more effective in small-scale corpus geotechnical engineering entity recognition.

Many loess landslides were caused by the agricultural irrigation in Heifangtai terrace, Yongjing County, Gansu Province. Their stability analysis and critical slip surface identification are particularly important, as it can provide a good support to disaster prevention. The loess landslides located on the margin of Heifangtai terrace have the characteristics of progressive backward failure, and the occurred and potential landslides are highly similar. The results from back analysis can provide important data basis for future landslide stability prediction. In this paper, the finite difference strength reduction method was used to calibrate the cohesion and internal friction angle of loess based on the NSGA-II genetic algorithm by setting three objective optimization functions (i.e., mean value of soil strength parameters, slip surface and factor of safety). Taking Dangchuan 2^# landslide in Heifangtai terrace as a study case, based on the slip surface observed after the first slide and assuming its factor of safety was equal to 1, the back analysis results show that the cohesion and internal friction angle of natural loess were 28.20 kPa and 25.16°; and the effective cohesion and internal friction angle of saturated loess were 16.59 kPa and 16.11°. Based on the computed results, the factor of safety and critical slip surface of the three subsequent slides were predicted, with their comparison with in-site observation information. The results show that a more reasonable estimation of loess strength parameters can be obtained by using the multi-objective constraint optimization algorithm, which provides a new solution for the stability analysis and quantitative risk assessment of landslides in Heifangtai terrace.

Investigations and studies on syn- exhumation or post-collisional magmatic rocks can provide key clues to understanding the crust-mantle interactions during the continental subduction and exhumation, and the tectonic evolution of collisional orogenic belts. In this paper it presents a comprehensive research of zircon U-Pb geochronology, petrology, geochemistry and isotopic geology for the syn-exhumation intermediate rocks in the east part of the East Kunlun orogenic belt. And the data could offer new insights into the generation of the syn-exhumation magmatism and the proto-Tethyan tectonic evolution. The results reveal that the Balong and Jinshuikou intermediate rocks yield zircon U-Pb ages of 420 Ma and 405 Ma, respectively, which overlap the exhumation timing of eclogites. The Balong diorite-porphyrite samples have relatively low MgO and Mg^# with high K₂O, while the Jinshuikou diorite samples have relatively high MgO and Mg^# with high Na₂O. The Balong diorite-porphyrite samples exhibit decreasing ratios of Nb/La with decreasing Mg^#, similar to the features of assimilation and fractional crystallization (AFC). And all of them are plotted on the curve of fractional crystallization. However, the Jinshuikou diorite samples do not show positive correlation between Nb/La and Mg^#, and those samples are plotted on the curve of partial melting. In addition, the Balong diorite-porphyrite samples have relatively higher I _sr ratios and lower ε _Nd(t) values than those of the Jinshuikou diorite samples. Based on the new petrological, geochemical and isotopic compositions, it concludes that the Balong diorite-porphyrite and the Jinshuikou diorite were respectively derived from basaltic magma differentiation and contamination, and partial melting of the lower crustal basaltic rocks. Combining with new regional studies, it proposes that those intermediate rocks might be formed in a post-collisional extension environment, and slab break-off is the critical factor for triggering the Silurian to Devonian syn-exhumation magmatism. The continental collision and the continental subduction in the East Kunlun region should have initiated at ~440 Ma.

Land subsidence is the loss of land elevation formed by a combination of natural and human factors. To prevent the delayed progressive geohazards, it is essential to predict large-scale land subsidence with high efficiency. However, the current prediction methods usually neglect spatial characteristics of land subsidence, which are time-consuming due to the issue of single-point cycle. To address the problem, a new prediction method of large-scale land subsidence based on multi-temporal InSAR and machine learning is proposed. Firstly, the large-scale land subsidence time series information is obtained by the SBAS-InSAR technique. Secondly, the spatial modes and the consistent principal components (PCs) are extracted from the time series information with the empirical orthogonal function (EOF). Finally, the PCs are trained and predicted by predictive model based on the ridge polynomial neural network with error-output feedbacks (RPNN-EOF), and the outcomes are reconstructed back to the land subsidence time series. The 84-view Sentinel-1A data from August 2018 to May 2021 of Yan’an New District were adopted in the land subsidence time seriesacquisition. Simultaneously, the spatial modes extracted by EOF can clearly reveal the spatial variation characteristics of the whole new district. The prediction results show that the root mean square error and modeling time of the proposed method is reduced by at least 22.7% and 27.5% respectively, in comparison with that by the single-point cycle pattern and the prevailing time series methods. Thus it has good practicality and applicability.

The Laohushan monogenic volcano is located in Dakouqin area, southern Jilin Province. In order to explore the geochemistry characteristics and genesis of the Miocene tephrite in this volcanic area, together with the coupling between it and the evolution of regional fault structure, the whole rock geochemistry and Pb isotope study on it were carried out. Na₂O/K₂O values of the tephrites are range from 1.30 to 1.38, with low A/CNK (0.97-1.09) and high Mg^# (58.05-61.11) values. The tephrites are enriched in high field strength elements (HFSEs), such as Th, U and Nb, and relatively weakly enriched in large ion lithophile elements (LILEs), such as Rb, Ba and K, with LREE/HREE=10.66-11.32. The values of ²⁰⁶Pb/²⁰⁴Pb, ²⁰⁷Pb/²⁰⁴Pb and ²⁰⁸Pb/²⁰⁴Pb are range from 17.661 to 17.675, 15.451 to 15.457 and 37.652 to 37.692, respectively. The parental magma is mainly from the asthenosphere mantle, with a small amount of subducted oceanic crust components, which shows the attribute characteristics of OIB (oceanic island basalt) or intraplate metasomatic enriched mantle source. The depth range of magma formation is about 93-105 km, and the existence of the Yitong-Shulan fault provides conduit and space for magma upwelling and fissure eruption. In addition, the formation of basaltic magma and its associated volcanic activity occurred in an extensional tectonic environment.

Marine sediment is not only a critical burial area of organic carbon from various sources but also a very active biogeochemical reactor, which plays a vital role in the global marine carbon cycle. Compared with the geochemical testing and qualitative description methods, the numerical model can break through the limitations of time and space and quantitatively obtain the reaction rate and flux of each carbon cycle process in marine sediments. Therefore, it has been paid more and more attention by the academic community. The degradation of organic matter in marine sediments is the most critical biogeochemical process driving the carbon cycle. Part of the dissolved inorganic carbon released into the surrounding pore water can diffuse to the overlying water column. The other part can form authigenic carbonate minerals with calcium and magnesium plasma precipitation. In this paper it firstly reviews the modeling process of three main types of sediment organic matter degradation models (discrete model, reactive continuum model, and Power model) and their applications in the global marine sediment organic matter degradation process. Then, starting from the primary and secondary reactions related to the degradation of organic matter, the description method of the reaction rate model of geochemical processes related to the degradation of organic matter in sediments is introduced, and the influence of the degradation of organic matter on the formation of authigenic carbonates and their carbon isotopes is discussed from the perspective of carbonate equilibrium system and isotope mass balance model. Finally, the problems and shortcomings of the current mathematical model in describing the degradation process of organic matter and the formation of authigenic carbonate are analyzed, and on this basis, the research points that need to be strengthened in the future are prospected. It is hoped that in this paper it will provide useful scientific support for understanding the mutual feedback between the ocean carbon cycle and global climate change and establishing a reliable prediction system for ocean carbon cycle and biogeochemistry.

The deterioration of shear properties of schist under dry-wet cycles has an important effect on the long-term stability of schist slope. Taking the mica-quartz schist widely distributed in Northwest Hubei as the research object, a series of laboratory tests were carried out to reveal the law and mechanism of its shear deterioration. The results of water absorption tests and direct shear tests show that the water absorption of mica-quartz schist increases, while the shear strength and residual shear strength decrease gradually with the increase of the number of dry-wet cycles, and the shear properties show obvious deterioration effect. Based on the changes of mica-quartz schist microstructure obtained by scanning electron microscopy, the deterioration mechanism of mica-quartz schist is revealed. Under the action of dry-wet cycles, the schist plane gradually expands and cracks, the strength of internal mineral particles is softened, the cementation between particles is weakened, and the rock skeleton becomes loose. What’s more, the cohesion is mainly affected by the degree of cementation between mineral particles, so the deterioration rate is fast. While the internal friction angle is mainly affected by the degree of embeddedness and the strength of mineral particles, so the deterioration rate is relatively slow.

The crude oil in the reservoirs in the Tahe area is mainly derived from the Cambrian Yuertusi source rocks. The discovered oil reservoirs are the result of mixing of crude oils of different stages and different maturities in the reservoirs. Based on the analysis of reservoir oil maturity and fluorescence spectra, this paper uses fluorescence spectrum parameters to establish a quantitative evaluation template for crude oil maturity, and assesses the maturity of crude oil at different episodes in the Tahe area by obtaining oil inclusion fluorescence spectrum parameters. The analysis results of the maturity parameters of crude oil show that the methylphenanthrene parameters and dibenzothiophene parameters in aromatic hydrocarbons are effective indicators for quantitatively evaluating the maturity of crude oil in Tahe area. Using aromatic hydrocarbon maturity parameters to calculate, the crude oil maturity (R _o) in Tahe area ranges from 0.6% to 1.4%, which has a good correlation with the crude oil fluorescence spectrum parameters, and a quantitative evaluation template for crude oil maturity is established based on the relationship between the two. The fluid inclusion analysis results show that there are four periods of petroleum charging in the Ordovician carbonate reservoirs in the Tahe area, corresponding to oil inclusions with orange, bright yellow, light yellow, and blue fluorescence. The fluorescence spectrum parameters of oil inclusions in different periods are also different. Based on the fluorescence spectrum parameters of oil inclusions in different periods and the established fluorescence spectrum parameters-maturity evaluation template of crude oil, it is determined that the maturity (R _o) of oil in orange fluorescent oil inclusions is less than 0.6%, that of oil in bright yellow fluorescent oil inclusions is between 0.6% and 0.8%, that of oil in light yellow fluorescent oil inclusions is between 0.8% and 1.0%, and that of oil in blue fluorescent oil inclusions is between 1.0% and 1.2%.The determined maturity of petroleum in different periods has an important reference for the study of petroleum accumulation process in Tahe area.

Fine fault identification is of great significance to improve the efficiency of exploration and development. Traditional seismic attribute fault recognition technologies identify fractures based on data discontinuity, and there are many interfering factors, making it more and more difficult to meet the needs of fine exploration in deep area. In order to improve the fault identification accuracy, this paper proposes a high-resolution intelligent identification method of fault. Based on the deep learning method to predict fault attributes from seismic data, a high-resolution and low-resolution fault label library is established, and a deep neural network is trained. It is confirmed by the model and actual data that the method solves the problem of high-frequency loss caused by up sampling in the convolutional neural network in deep learning, which reduces the resolution of faults, and improves the resolution ability. The root mean square error of the simulated data decreased by 40.02%.Compared with traditional algorithms, the method not only detects fault features more accurately, but also has a higher resolution than common deep learning fault recognition.

The Handan-Xingtai area is one of the important iron ore concentration areas in China. With the development of prospecting work, the prospecting effect of the traditional exploration technology is not ideal, new exploration methods and prospecting concepts are urgently needed. Infrared spectroscopy (SWIR+TIR) can be used to rapidly determine the types and spatial distribution of altered minerals, the changing law of spectral characteristic parameters of altered minerals is revealed, and the information related to mineralization further determined. Through infrared spectroscopy tests, it deciphered more than 20 minerals and determined five mineral zonation, from rock mass to surrounding rock is sericite-chlorite-calcite-albite zone (alteration zone Ⅰ)，chlorite-serpentine-tremolite-actinite-diopside-chrysolite zone (alteration zone Ⅱ)，calcite-dolomite-siderite zone (alteration zone Ⅲ)，sericite-illite-calcite-dolomite-siderite zone (alteration zone Ⅳ)，sericite-illite-kaolinite-montmorillonite zone (alteration zone Ⅴ). The study shows that it is closer to the mineralization areas, the lower the chlorite Fe-OH2250Pos, the lower the dolomite 6.45 μmPos and 11.2 μmPos reflection peak. Reduction of Al-OH2200FWHM of sericite minerals and the increase of Al-OH2170-IC of kaolinite minerals reflects the hotter fluid environment. Chlorite of low Fe-OH2250pos value (<2 245 nm), dolomite of low 6.45μmPos value (<6 440 nm) and dolomite of low 11.2 μmPos value (<11 250 nm) are effective prospecting sign in Baijian. The decrease of sericite Al-OH2200FWHM value（30→22）and the increase of kaolinite Al-OH2170-IC value（0.2→2.8） both can determine whether there is a mineralization system in deep. These research results provide a scientific basis for mineral exploration in the Baijian mining area, and also provide new ideas and methods for prospecting for the same type of deposits in Han-Xing area.

According to the geological characteristics of deep shale gas, such as large horizontal stress difference, small vertical stress difference and strong rock plastic characteristics, evaluation methods were established for the logging curve, total organic carbon, porosity, all gas hydrocarbon, key logging elements, silicon, mineral brittleness, rock mechanics parameters, etc.. And the geology-engineering double sweet spot study along the horizontal wellbore was carried out based on the design concept of geology-engineering integration, so as to accurately identify optimal geological & engineering sweet spot, providing guidance for horizontal well fracturing in deep shale gas play. Then, based on three-dimensional fracture network with high fracture conductivity as the objective function, the volume fracturing pump mode and key parameter optimizations were studied for deep shale gas well under narrow pressure window, i.e., circuitous temporary plugging fracturing technology optimization, the dynamic movement of proppant in complex fracture network and test of fracture conductivity, and research and development for integrated variable viscosity slick water, etc.. Field applications demonstrate that the geology-engineering integration volumetric fracturing technology could improve post-frac gas production by more than 30%-50% compared with adjacent wells. The technology can greatly improve the economic development of deep shale gas wells, and it also has important guiding significance for shale gas wells with vertical depth more than 4 500 m as well.

In order to discuss thedynamic mechanism of Late Cretaceous magmatism in the Lhasa block, this paper carried out petrographic, chronological, geochemical and mineral chemistry studies on the Menba granodiorites in the southern margin of the central Lhasa block. LA-ICP-MS zircon U-Pb dating indicates that the Menba granodiorites emplaced in the Late Cretaceous (83.2 Ma±0.9 Ma). The geochemical characteristics show that the Menba granodiorites are adakitic rocks.Electron microprobe data reveal that plagioclase belongs to andesine and oligoclase (An=16.2-34.7). This paper believes that the parent magmas of the Menba granodiorites may be a result of magma mixing between mantle- and crust-derived magmas and the mantle-derived magma underplating may have led to partial melting of the thickened lower crust and then reactions between them.Combined with the diagenetic environment and spatial distribution characteristics of the Late Cretaceous magmatic rocks. This paper concludes that the Late Cretaceous magmatism in the southern Lhasa block was mainly controlled by the ridge subduction of the Neo-Tethys. The upwelling of asthenosphere mantle along the slab window of the mid-ocean ridge induced the Late Cretaceous large-scale magmatism in the southern margin of the south Lhasa block, while the upwelling of asthenosphere material along the tear slab window of transition fault that cuts the mid-ocean ridge induced the small-scale intraplate magmatic belt, which approximately perpendicular to the large-scale magmatism in the southern margin of the south Lhasa block.

Fusulinids are important standard fossils of the Carboniferous and Permian periods. The identification of fusulinids is significant for determining the geological age and delineating the Carboniferous-Permian stratigraphy. Considering the limitations of current fossil detection methods, a fusulinid detection method based on a deep learning single-stage algorithm is proposed. Taking fusulinids as the research object, the original model is improved by channel pruning by jointly optimizing the weight loss function and L1 regularization of the BN layer scale factor to compress the model size. Furthermore, the knowledge distillation is utilized to restore the detection performance of the pruned model. The experimental results show that the method can achieve the classification and localization of the fusulinids in the thin section images. The average accuracy reaches 98.1%, which meets the requirements of the real-time detection model. In addition, the number of model parameters is reduced by 74.1%, which solves the problems such as the lack of arithmetic power existing in real scenes. The method can effectively achieve the detection of fusulinids, while extending the applicability of the model to embedded devices and providing more possibilities for deep learning to perform intelligent recognition in paleontological fossil images.

The sedimentary filling evolution during the Paleogene rifting process in Zhu I depression remains hotly disputed. Tracing source to sink process of the Paleogene was carried out on the sandstones of Wenchang Formation and Enping Formation in Zhu I depression, with the method of zircon U-Pb dating. The geochronology, provenance and sedimentary filling evolution in different evolution stages were further explored. The results show that the sags in Zhu I depression not only received the intrabasinal provenances from the paleo-uplift area, but also were affected by the terrigenous supply transported from the northern peripheral uplifts in South China Block during rifting stage. The Wenchang Formation sediments were dominated by intrabasinal short-distance provenances, which were derived from paleo-uplifts nearby the sags, including Mesozoic magmatic rocks, Mesozoic sedimentary strata and Cenozoic volcanics. The U-Pb geochronology of all sags was characterized by the dominance of Yanshanian clusters except for multi-peaks in Yanshanian, Indosinian, Caledonian periods in Lufeng Sag. The sediments of the Enping Formation were firstly affected by the materials from the South China Block in the north of Enping Sag and the west of Huizhou Sag. During the deposition period of the upper Enping Formation, the age spectra of Zhu I depression sediments show a wide Proterozoic to Mesozoic range with the peak complexity in Yanshanian, Indosinian, Caledonian and Jinningian periods. Thus, the sags were filled by the materials from both the South China Block and peripheral uplifts with the gradual increase of material supply from the South China Block, indicating that the basin provenances had changed from near-source to far-source ones. Meanwhile, the Zhu I depression was characterized by large shallow braided deltas that were formed via relatively long-distance transportation.

Sedimentary basins are prominent in convergent margin studies. Located near the volcanic arc in the forearc region, the forearc basin is relatively well-preserved after intense orogeny, with records of intact information about orogenic processes. This paper illustrates the tectonic setting, formation mechanism, provenance, and sedimentary-structural characteristics of the forearc basins exemplified by well-studied Cenozoic examples. Forearc basins are located between the volcanic arc and the trench-slope break. It can be formed in an extensional environment where normal fault forms a half-graben basin, or in a compressional environment where thrusts of the accretionary wedge serve as a dam to pond sediments. All kinds of sedimentary facies can be developed in the forearc basin located in a continental arc, while hemipelagic-pelagic facies are dominant in that located in an intra-oceanic arc. Terrigenous debris is mainly derived from the proximal volcanic arc and accretionary wedge and enters the basin through rivers, submarine canyons, and collapse. Sediments in the forearc basin usually grow thicker from the basin margin to the center and from the bottom to the top. Subduction of mid-ocean ridge, seamount, oceanic plateau, and fracture zone could affect the structure of subduction zones to various degrees, leading to inversion, exhumation, sedimentary hiatus, sedimentary provenance changing, and depositional environment changing of the forearc basin. The development of the forearc basin is difficult to be summarized by a single model. Thus, when forearc basin is applied to orogeny analysis, multi-disciplinary data should be considered in a comprehensive analysis of basin geological records.

In the southern Tibetan rift zones, there are several approximately north-south trending rifts distributed from west to east. As important tectonic extensional zones within the blocks, these rifts have developed a series of normal faults and experienced multiple strong earthquakes. Since the Late Quaternary, this region has exhibited intense tectonic activity with frequent earthquakes causing serious disasters. For instance, the January 7, 2025 Mw7.1 (CENC: Ms6.9) Tingri earthquake demonstrated the characteristics of “small earthquake with major disaster consequences”. To assess the seismogenic potential of normal faults within the rift zones and understand their disaster-inducing competence, this study divides 92 normal fault zones based on geometric characteristics and statistically analyzed fault trace lengths. Under the assumption of full-length surface rupture along fault traces during earthquakes, combined with empirical relationships between normal fault rupture length and moment magnitude, we estimated the maximum potential magnitudes of normal faults in the rift zones. Results indicate that these normal faults have upper seismogenic limits ranging from Mw6.5 to Mw7.5, with numerous historical seismic gaps. While generally demonstrating strong seismogenic competence, they exhibit an eastward-increasing strength pattern. Bounded by major fault zones to the north and south, and considering multiple historical seismic gaps along the southern magethrust (particularly in the context of accelerated Coulomb stress loading following the 2015 Nepal Mw7.8 earthquake and potential interaction/triggering effects between major boundary faults), the normal faults south of Yarlung Tsangpo River, especially those in the Tingri-Nyalam and Xiongqu fault, show high potential for future strong earthquakes.

To address the problems of landslide susceptibility prediction (LSP) modeling including possible errors in landslide and non-landslide samples, complex non-linear relationships between environmental factors and unaddressed machine learning interpretability, a deep learning-based Self-screening Bi-directional Long Short-Term Memory and Conditional Random Fields (SBiLSTM-CRF) model is proposed to reduce the impact of these problems on LSP and improve its confidence. The SBiLSTM-CRF model has the advantages of deep learning network with deep layers, wide width and iterative modeling, which can predict the non-linear relationship between environmental factors and automatically screen out the wrong landslide samples; it can select non-landslide samples from the initial low/very low landslide susceptibility zone through iterative modeling, and finally reveal the internal mechanism of the coupling of environmental factors to predict landslide susceptibility. The SBiLSTM-CRF model is used to predict landslide susceptibility in Yanchang County of China, and compared with cpLSTM-CRF, random forest (RF), support vector machine (SVM), stochastic gradient descent (SGD) and logistic regression (LR) models. The results show that SBiLSTM-CRF overcomes the problems of sample error and complex nonlinear relationship between factors in traditional machine learning, has superior performance in modeling susceptibility than conventional machine learning, and the interpretability of the model reveals that factors such as slope, elevation and lithology control the development of mounded landslides in Yanchang County.

Establishing a 3D orebody model is the foundation of digital mine and smart mine. In response to the phenomenon that the classical radial basis function surface reconstruction algorithm leads to surface boundary self-fitting and model discontinuity when the original data is sparse, this paper proposes a method of implicit automatic modeling of complex orebody with radial basis function incorporating multiple machine learning. This method uses the Stacking model to learn the distribution characteristics of the discrete point cloud data of the orebody contour lines to build a directed point set characterizing the geometric information of the orebody model. On this basis, the boundary points and normal vectors are extracted, the implicit field is established by the Hermite radial basis function, and finally the 3D orebody model is visualized based on the marching tetrahedron algorithm. The analysis was compared with the contour line splicing method, the classical radial basis function surface reconstruction algorithm, and the simple kriging interpolation method. The method can effectively reduce the phenomenon of self-fitting of surface boundaries, reduce redundant holes in the model, and improve the continuity of the model; the similarity between the contour lines cut by the established model and the original contour lines reaches 75.14%, which is comparable to the explicit model with a high degree of manual intervention; the gap between the model and the explicit model in volume characterization reaches the lowest.