This special issue titled “Artificial Intelligence and Big Data Geoscience” consists of 17 papers covering topics such as knowledge graphs, deep learning-based image recognition, machine-readable expression of unstructured geological information, big graph data and community detection, association rule algorithms, 3D geological simulation and mineral prospecting, and the Internet of Things and online monitoring systems. A progressive multi-granularity training deep learning method is proposed for mineral image identification; the model achieves 86.5% accuracy on a commonly used dataset comprising 36 mineral types, increasing the accuracy of mineral identification. Knowledge related to porphyry copper ore in the Qinzhou-Hangzhou mineralization belt, South China, is collected using both primary and literature data sources, and Natural Language Processing (NLP) techniques are used to semantically correlate and reason over the knowledge graph, enabling automated knowledge extraction and reasoning. The association rule algorithm is used to analyze the correlation between trace elements and gold mineralization in major Carlin-type gold deposits in the “Golden Triangle” region of Yunnan-Guizhou-Guangxi provinces, China, and combined with the migration and enrichment law of elements to analyze the genetic mechanism of deposits. By builing a quantitative prospecting indicator method based on association rule algorithm, this study provides new ideas for establishing quantitative prospecting indicators for other types of deposits. In study of machine-readable expression of unstructured geological information and intelligent prediction of mineralization associated anomaly areas in Pangxidong District, western Guangdong, China, unstructured geological information such as stratigraphy, lithology and faults is processed by machine-readable conversion, and two machine learning algorithms—namely, One-Class Support Vector Machine and Auto-Encoder network—are applied to mine the geochemical test data of the stream sediment as well as the comprehensive geological information such as faults and stratigraphy, to extract the features of the mineralizing anomalies, and ultimately realize the intelligent circling of mineralizing anomalous areas. In study of networked monitoring of urban soil pollutants and visualized system based on microservice architecture, a system capable of real-time online monitoring, processing, and analyzing urban soil pollution data to enhance the timeliness of predictions and warnings is developed, where the integrated monitoring and data visualization system is based on the microservices framework Spring Cloud Alibaba. The above mentioned studies provide highly valuable application scenarios and research cases, reflecting to some extent the latest research advances in the field of artificial intelligence and big data geoscience in China, and are worthy of peer attention.

The mineral exploration knowledge graph (MEKG) is a component of Earth system-mineralization system-mining system correlation graphs, which represents an intersection between Earth science and data science and provides a novel approach for the prediction and evaluation of mineral resources. The conventional mineral exploration methods suffer from information asymmetry/inefficiency/inaccuracy, thus have limitations in effective utilization of geological data. To address this issue, we collect knowledge data relating to porphyry copper ore in the Qin-Hang mineralization belt using both primary and literature data sources, and construct a MEKG with automated knowledge extraction and reasoning using natural language processing (NLP) techniques. Briefly, the MEKG model represents the entities and attributes of porphyry copper ore and their relationships in the Qin-Hang mineralization belt; based on this framework, NLP techniques are used to semantically correlate and reason over the knowledge graph, enabling automated knowledge extraction and reasoning. In addition, we develop a Q&A and visualization system that allows users to query entities/attributes and their relationships to obtain relevant information and visualize the data structure and data relationship. This study demonstrates the effectiveness and accuracy of knowledge-based intelligent applications in porphyry copper ore exploration in the Qin-Hang belt through experimentation and testing. Compared with traditional methods, this application provides more comprehensive and accurate mining suggestions in a short time to aid geological exploration decision-making. Also this study can serve as reference for other mineral exploration fields. In the future, we aim to further enhance the performance and functionality of this knowledge model by broadening the graph algorithm applications and recommendation systems, so as to meet the needs of mineral explorations under different scenarios and expand the model’s application potential to other related fields.

Knowledge graph construction is an effective means of acquiring and representing knowledge in data-driven research, however, existing knowledge graphs have many problems and limitations in mineral resource prediction. Firstly, relevant studies are few while existing knowledge graphs lack spatiotemporal semantics, which limits the effective representation and analysis of the spatiotemporal characteristics of mineral resources. Secondly, existing graph construction methods emphasize text extraction at the data level, but lack ontology construction involving complex logical relationships and lack effective association between ontology and data layers. As a result, existing knowledge graphs lack in-depth and sufficient semantic information to meet the requirement of mineral resource prediction in expressing complex geoscience concepts and relationships. To address this issue, this study takes an ontology-guided approach to construct a knowledge graph suitable for mineral prediction tasks. We first construct the initial domain ontology on the basis of in-depth understanding of mineral prediction theories and methods; we then integrate the domain ontology with selected mature geological time ontology and geographical space ontology to expand the initial ontology—by embedding spatiotemporal semantics we can effectively express the spatiotemporal characteristics of mineral resources. We also pay attention to the association between ontology and data layers—by establishing rich semantic relationships we can achieve effective inter-node connection and information sharing in the knowledge graph. Experimental results show that the knowledge graph outperformed other existing graphs in terms of knowledge richness and confidence. This study provides a methodology for multi-ontology based knowledge graph construction for mineral prediction, thereby promoting further development of this field.

Big data and artificial intelligence have greatly transformed mineral exploration practices with the development of innovative mineral forecasting models and improvement of forecasting efficiency for strategic minerals. In the field of quantitative mineral forecasting, comprehensive intelligent forecasting by combining knowledge and data has gradually become a common consensus, however, the challenge lies in how to combine knowledge and data. Knowledge graphs integrate multi-source, heterogeneous geoscience big data and drive knowledge discovery through rules and reasoning. Here, we discuss the feasibility and technical roadmap of knowledge graph-infused intelligent and automated mineral resource forecasting, particularly in consideration of the characteristics of knowledge graphs in the era of big data and artificial intelligence. We focus mainly on the construction of multi-temporal, all-element knowledge graphs for mineral deposit-mineral exploration systems and the methodology for establishing forecasting models from the perspectives of ore commonality and distinctiveness based on knowledge graphs. The opportunities and challenges of knowledge graph embedding for geological anomaly information extraction and quantitative resource forecasting are also discussed, in the hope that the infusion of knowledge representation and reasoning from knowledge graphs into the technical workflow of quantitative mineral resource forecasting can aid geologists in building ore forecasting models and enhancing automated and intelligent mineral forecasting.

Mineral knowledge is important for geosciences research. Some mineral databases are used for storing and retrieving mineral knowledge, and common search engines can also answer mineral questions. But the mineral databases cannot answer mineral questions in natural language and the answers returned from the common search engines need to be filtered. To solve the above problems knowledge graphs have been used; however, the current mineral question-answering based on knowledge graphs can only answer simple questions involving one triplet, but not complex questions involving multiple triplets and multi-hop reasoning. This paper presents a mineral question-answering system based on multi-hop reasoning in knowledge graphs. The mineral entities, relations and questions are first transformed into vectors of complex domain to obtain their semantic and reasoning relations by using the ComplEx model, and Bert-LSTM-CRF is applied to obtain the head of the question. Candidate entities of the head are then obtained by calculating the edit distance and word segmentation, and a fully connected network is constructed to obtain the most related entity of the head of the question from the candidate entities and the entity is the start of the reasoning. Next, the entity and question vectors are concatenated into an input vector into the fully connected network to get their most related relation; afterward another entity most related to the starting entity/relation can be obtained from the mineral knowledge graph to start the reasoning of the next hop; the question of the next hop is updated by the concatenated vector of this hop to bring the reasoning information of this hop to the next hop. This process continues until the most related relation obtained is the stop sign predefined. The last entity obtained in this process is the answer to the question and the reasoning path is also remembered. This method is implemented using Python under Tensorflow and compared with related methods, which show the effectiveness of the method. Using this method, a question-answering system capable of answering complex mineral questions is developed under the front and back end separation architecture employing RESTful API, React, Ajax, echarts and Flask, which provides a platform for acquiring mineral knowledge and performing geosciences research.

The application of big data mining and machine learning algorithms in mineralization prediction has become an important research trend, but unstructured geological data cannot be directly mined—first they need to be converted to machine-readable expressions. In this study of the Pangxidong ore district in western Guangdong Province, the unstructured geological information such as stratigraphy, lithology, faults are converted into machine-readable format, and two machine learning algorithms, namely, One-Class Support Vector Machine and Auto-Encoder Network, are applied to mine the geochemical test data of stream sediments as well as the comprehensive geological information on faults, stratigraphy, etc. to extract the features of mineralization anomalies and ultimately achieve intelligent delineation of the anomaly areas. Through combined application of One-Hot Encoder and the weighted variable method for spatially weighted principal component analysis, the structural transformation of the unstructured geological information is realized, and geological information is maximally preserved for data mining. It is demonstrated that the application of One-Class Support Vector Machine and Auto-Encoder Network can effectively solve the problem of data imbalance, as the numbers of ore and non-ore spots in the study area are seriously unbalanced. The prediction results generated using the integrated, synthesized multi-source geological data are relatively consistent with the observed spatial distribution of Pb-Zn deposits and the actual geological structure in the study area, indicating the two algorithms can effectively identify potential prospecting targets and ore deposits. Compared with traditional geochemical prospecting methods, the intelligent prediction method can process and integrate multi-source geological information about the ore-forming processes and identify mineralization anomaly areas. This method is applicable in prospecting areas without prior ore discovery, thereby improving the efficiency of ore prospecting and increasing the possibility of finding ore deposits.

Carlin-type gold deposits are an important reservoir of gold. Due to the gradual depletion of shallow surface gold resources, there is an urgent need for new prospecting methods to explore deep and hidden areas. The advent of the big data era has opened up new prospecting ideas. Association rule algorithm one of the most commonly used mining algorithms and can be used to effectively mine the inherent correlation between data items in large data sets. In this study, association rule mining is used to analyze the correlation between trace elements and gold mineralization in major Carlin-type gold deposits in the Yunnan-Guizhou-Guangxi “Golden Triangle” region. Combined with element migration and enrichment patterns, elemental anomaly combinations are extracted, and quantitative prospecting indicators are established. The elemental anomaly combinations are divided into elements with strong positive correlation and significantly enriched (As, Sb, Hg, Tl, Ag, W, Rb), indicating sulfidation and clayification (Rb); elements with strong positive correlation and slightly enriched (Zr, Th, Ta, Nb, Hf) or with strong negative correlation and strongly depleted (Li, Sr), indicating decarbonation; elements with strong positive correlation and slightly enriched (Sn, Zn, Ni, V, Co, Cu), likely reflecting their low contents in ore-forming fluids; and elements with weak correlation and not enriched (Cd, Pb, Ba, Bi, U, Mo)—these elements show no significant correlation with gold mineralization. The elemental anomaly combinations obtained by big data approach is consistent with previous understanding of the genesis of Au deposits, i.e., Au is mainly formed under decarbonation and sulfidation processes accompanied by significant clayification, in which sulfidation is the main genetic mechanism. Through association rule mining, quantitative prospecting indicators are established: For sulfidation related elements (As, Hg, Sb, Tl, W, Ag, Rb), when the number of medium-high content elements in samples ≥1, 2, 3, 4, or 5, the corresponding Au contents ≥4.5×10^-9, 47.0×10^-9, 150×10^-9, 500×10^-9, or 1000×10^-9; when the number of high-content elements ≥1, 2, or 3, the corresponding Au contents ≥150×10^-9, 500×10^-9, or 1000×10^-9; during prospecting, both indicators should be used to ensure efficient delineation of ore bodies, without outcrops. For decarbonation related elements (Zr, Th, Ta, Nb, Hf), decarbonation is indicated when elemental content anomaly occurs at any two of the elements in samples. The method developed in this study for establishing quantitative prospecting indicators based on association rule algorithms should provide new ideas for other types of mineral deposits.

To investigate the current status, hotspots, and frontiers of big data and artificial intelligence research in the field of geology, this study conducts literature screening of relevant research publications between 2000-2022 using China National Knowledge Infrastructure (CNKI) and Web of Science (WoS) core databases. A total of 3600 Chinese and 1803 English articles are collected, and community structure analysis software CiteSpace is used for visual analysis of cooperation authors, research countries/institutions, keyword clustering, and keyword spatiotemporal distribution maps. Furthermore, a stochastic frontier analysis correction is conducted on publications by international top-tier geoscience journals (comprehensive impact factor ≥10) between 2021-2022. The global cumulative publication volume in this research field had surged in the past decade, led by Asian countries represented by China and European/American countries represented by the United States, with China and the United States showing no significant differences, and the betweenness centrality measures generally higher for European/American countries than for Asian countries. In China, research collaborations were mostly among domestic institutions and relatively rare with foreign research institutions, whilst the opposite was true in foreign countries. The research hotspots in this field were geological disaster prevention and control, earthquake interpretation, petroleum and natural gas exploration, and solid mineral resource prediction using machine learning and knowledge graphs. Research frontiers included significant geological events during Earth’s evolution, global climate change, polar and marine geology, digital geological modeling and quantitative analysis, earthquake prediction, and accurate assessment of geological disaster susceptibility by means of deep learning, integrated learning, and intelligent platform.

Mineral identification as a crucial aspect of geosciences is of great importance to resource exploration, rock classification, and geological monitoring. However, traditional methods are inefficient as they often rely on human experience and subjective judgment. In recent years deep learning-based image classification has been used for accurate and rapid mineral identification. While these studies have achieved certain results, the number of identifiable mineral types are limited and the identification accuracy need to be further improved. This paper aims to address the issue of uneven distribution of mineral image samples in a mineral dataset on 36 common minerals. DCGAN is first used to generate images for data augmentation focusing on the 11 minerals with low sample counts, and the best set of images is selected, by comparison, to expand the dataset. Next, to obtain a more reliable and precise identification model, ResNet, RegNet, EfficientNet, and Vision Transformer models with better performance on ImageNet are transferred to the mineral dataset. Based on the permutations of the trained base models, 11 ensemble models are obtained, with which 24 identification results are obtained using two voting methods, average and weighted soft voting. These results are then compared to select the one with the highest accuracy. The experimental results demonstrated that data augmentation using DCGAN improved the model accuracy by 3.12% averaged over all models. Among the ensemble models, weighted soft voting performed better and achieved the highest accuracy of 87.47% on the augmented dataset.

In earth sciences the rock microscopic data collection process is both labor-intensive and inefficient, which have a negative impact on research cost/reliability and open data sharing. Additionally, rock heterogeneity and variation in data collection methods typically result in small-scale datasets—this poses a significant challenge to deep learning frameworks that rely on large-scale datasets for training. To address this issue, we investigate how transfer learning can facilitate information sharing across different rock types and enhance model performance in tasks such as mineral identification and intelligent interpretation. By compiling thin section image datasets, taking in diverse rock sampling regions, rock types, mineral compositions, viewed under varying viewing modes, we delve into the mechanisms of transfer learning across different observational targets and methods, focusing on the deep representation of geological information. Our findings not only highlight the pivotal role of transfer learning in promoting information sharing and improving model performance within the field of geosciences, but also lay a foundation for the automatic and intelligent integration of geological insights. According to experimental results, transfer learning led to significant accuracy improvement, from 53.3% to 98.73%, in intelligent interpretation task, and a nearly 10% improvement in mineral identification task. These results convincingly showcase the great potential of transfer learning in addressing practical problems in geology as well as enhancing model generalization, model performance, and model stability.

In recent years mineral image recognition has become increasingly important for mineral identification with the use of deep learning. While such application has gained some success, further improvement is needed to enhance the identification accuracy on large-scale mineral datasets. The fine differences in morphology, texture, and color between different minerals may align with the characteristics of fine-grained recognition algorithms, yet results of fine-grained recognition for mineral identification have rarely been reported. This paper proposes a fine-grained mineral identification model based on Next-ViT, which allows precise classification of mineral images by progressive model training across different granularity levels. In this approach, Next-ViT, which combines the advantages of transformer and convolutional neural network, is utilized to extract rich image features; a random jigsaw generator is then employed to create mineral puzzles at different granularity levels encompassing various information from detailed to general. The model training involves progressive learning across multiple granularity levels. In the early stages, the model primarily focuses on fine-grained features, learning detailed information from the puzzles to differentiate between different minerals; as training progresses, model learning gradually shifts to higher granularity levels, capturing more abstract and global information. Through this approach, the model can effectively utilize information across multiple granularity levels, thereby improving the accuracy of mineral identification. Experimental results demonstrated the effectiveness of this approach, with the proposed model achieving an accuracy of 86.5% in mineral identification on a dataset on 36 common minerals.

Three-dimensional predictive modeling of mineral prospectivity is an important approach to deep mineral exploration. Although significant advancements have been made in the methodology and application of this approach, several key scientific and technological issues remain unresolved concerning the insufficiencies of multi-scale 3D predictive modeling methodology, uncertainty analysis and optimization of prediction results, mining of key factors in 3D mineralization prediction, and dedicated 3D deep learning models and methodologies tailored for 3D predictive modeling of mineral prospectivity. Focusing on these key issues, this paper conducts a comprehensive review of current research progress in the field, and proposes potential solutions and research directions to address these issues. Future developments in this field include methods for deep mining of 3D predictive information; applicable 3D deep learning models and training methods for enhanced predictive modeling; uncertainty analysis and optimization methods for improving the reliability and accuracy of 3D mineralization prediction; and a methodological framework for multi-scale predictive modeling so as to effectively guide deep mineral exploration at the levels of orebodies, ore fields, and ore deposits. Resolving these key issues will further develop and refine the theoretical and methodological frameworks of 3D mineralization prediction, significantly improve the efficiency of deep mineral exploration, and ultimately facilitate breakthrough in mineral deposit discovery in the deep earth.

3D structural geological models are a digital representation of geological structures and geological body (object) boundaries in geological space. With the increasing demands for raster and vector integration and spatial query and analysis of geological data, the construction of integrated spatial data model for unified expression of geological structures has become one of the basic problems of 3D geological information science. To address the problem of expressing complex geological structures by regular grids, PillarGrid, Stack-Based Representation of Terrains (SBRT), etc., this study proposes the InterfaceGrid data model to fully consider that the distribution of geological structures/attributes underground exhibit strong non-uniformity, discontinuity, spatially multi-scaled, and show longitudinal stratification and multi-attribute field coupling. By design, this InterfaceGrid data model can uniformly describe 3D geological structures and realize the vector raster integration expression of complex geological structures. In this paper, the formal expression framework of InterfaceGrid is constructed based on GeoAtom theory; the construction process of the InterfaceGrid model is described; and the data update and spatial query algorithms are designed based on the InterfaceGrid model. The volume visualization and online browsing of geological grid are realized using GPU ray casting and adaptive sampling strategy. Compared with SBRT, InterfaceGrid can more truly describe the geological boundaries and improve the accuracy of 3D structural geological models. The application of InterfaceGrid in the 3D grid construction of the global lithosphere verifies the applicability of InterfaceGrid in the organization and management of large-scale geological data. Compared with PillarGrid, the data volume is reduced by about 1/3 in InterfaceGrid, making it more suitable for the data-intensive geoscience network applications.

Geological sequestration can be used to reduce CO₂ emission without much effect on economic growth. It has become an indispensable technical approach to achieving dual-carbon goals. However, geological sequestration carries significant environmental risks from CO₂ leakage at storage sites. To ensure the safety and efficacy of carbon sequestration it is critical that potential leaks can be identified through continuous monitoring. In this regard, the Internet of Things (IoT) is ideal due to its large-scale, continuous monitoring, and intelligent analysis capabilities, yet this technology has not been widely implemented. This paper outlines the basis for sensor selection and sensor node deployment, proposes the design idea for underlying sensor technology, and establishes an IoT CO₂ monitoring system for storage sites. Specifically, infrared CO₂ sensor is selected as the primary sensor and laser CO₂ sensor as the secondary sensor, along with FT-IR patrol monitoring; a combination of real-time optimization of mobile deployment, random deployment, and fixed deployment is used in sensor node deployment; a mix of cluster topology and mesh topology is used in high-risk areas, and star topology and tree topology are used in edge areas connected to the main area. As technology advances, sensor mass production and sensor miniaturization will lead to more efficient and scalable sensor networks, and IoT monitoring technology will play a crucial role in continuous monitoring of carbon storage sites.

Analyzing the spatio-temporal variations of atmospheric carbon concentrations in energy resources development areas and identifying influencing factors are crucial for exploring a high-quality development pathway in the context of “Dual Carbon”. Xinjiang Uygur Autonomous Region serves as a vital base for energy and strategic resources in China. This study oriented to the current status of energy resource development in Xinjiang Uygur Autonomous Region, we collected and preprocessed Orbiting Carbon Observatory-2 (OCO-2) carbon dioxide Level 3 data products from 2015 to 2021. We analyzed the temporal trends and spatial distribution patterns of atmospheric carbon concentration in the study area and structured a deep forest regression model to analyze the driving factors of the spatio-temporal variations in carbon concentration. The results indicate that: (1) Xinjiang Uygur Autonomous Region, Junggar Basin, Turpan-Hami Basin and Tarim Basin’s XCO₂ concentration exhibited a cyclic upward trend from 2015 to 2021, with a “decrease-first then increase” growth rate, showing a distinct “high in spring, low in winter” seasonal variation trend; (2) in spring, autumn, and winter, the spatial pattern of XCO₂ concentration in Xinjiang shows a “high in the north, low in the south” trend, with high XCO₂ concentrations accumulating in basin and energy resource development areas. Conversely, a trend of “low in the north, high in the south” observed in summer; (3) topographic relief, wind velocity, NDVI, land surface temperature, precipitation, 10-meter V wind, 10-meter U wind, and energy development intensity significantly influence the spatio-temporal distribution of regional XCO₂ concentration, showing notable spatial heterogeneity and significant differences. These findings contribute to understanding the mechanism of carbon concentration evolution in energy resource extraction areas and hold profound implications for achieving national carbon reduction targets, guiding carbon neutrality strategies, and monitoring carbon emission reduction effects.

The nature of soil pollution—cumulative, hidden, latent, irreversible—makes it essential that urban soil pollution should be closely monitored and prevented. However, traditional monitoring methods cannot perform real-time pollution monitoring and have limited data processing capabilities. To address this issue, we aim to develop a pollution prediction and early warning system capable of real-time online monitoring, processing, and analyzing urban soil pollution data. In this paper we report the development of a monitoring and data visualization system based on microservice framework Spring Cloud Alibaba, whereby integrating the EMQX platform soil data are successfully collected and stored. Additionally, we develop a WebGIS module that interfaces with Geoserver—this module utilizes OpenLayers to render maps and soil element concentration charts, enabling the monitoring and visual analysis of soil conditions. We believe with breakthroughs in sensor technologies relating to chemical monitoring, real-time online monitoring, processing, and analysis of soil pollution data through Internet of Things (IoT) can be achieved. The IoT monitoring and visualization system has been tested and its effectiveness in identifying pollution changes, predicting trends, and devising effective prevention and control measures are demonstrated. Most importantly, practical applications confirm the system's notable advantages in enhancing the timeliness of soil pollution prediction and early warning.

Hydrogen energy is a clean, efficient, and zero-carbon energy source. The formation, transportation, and accumulation of natural hydrogen are closely related to plate tectonics. As the only rocky planet in the solar system known to have plate tectonics and liquid water, Earth has unique geological hydrogen generation pathways such as degassing, serpentinization, and water radiolysis. The ocean-floor, which occupies two-thirds of the Earth’s surface, has great potential for natural hydrogen generation through serpentinization, due to the extensive exposure of oceanic crust or mantle along or around microplate boundaries and ocean-floor fissures. Microplate boundaries, submarine plateaus, ocean floor fracture zones, micro-mantle blocks, and non-volcanic passive continental margins are favorable targets for exploring ocean-floor natural hydrogen. The northeastern continental margin of the South China Sea is also worthy of attention. However, it is difficult to establish a unified ocean-floor hydrogen accumulation model due to the significant differences and diversity in the formation, migration, and storage conditions of natural hydrogen in different tectonic settings. The predicted hydrogen sites, whether they can form reservoirs, how they form reservoirs, and the related exploitation and utilization technologies need to be explored in the future.

In the current context of the dual-carbon policy, the national demand for clean energy, such as hydrogen, is growing significantly. Serpentinization of peridotite is one of the most widespread water-rock interactions on the seafloor, and hydrogen gas, a primary product of this process, serves as a crucial pathway for the formation of marine hydrogen energy. Therefore, the deep oceanic crust holds highly promising hydrogen energy reserves, representing a vital breakthrough for alleviating current dual-carbon pressures and driving the development of new productive capacities. However, global technologies for detecting and extracting marine hydrogen energy are still in their infancy, presenting a significant opportunity for future seafloor energy exploration and growth. This paper systematically reviews the formation principles and distribution characteristics of marine hydrogen energy, outlining potential detection technologies and extraction methods. We propose that comprehensive geophysical exploration methods, such as multibeam bathymetry, magnetic surveys, gravity measurements, and multi-component seismic exploration, hold promise for detecting potential hydrogen reservoirs on the seafloor. Additionally, methods like hydraulic fracturing and microwave heating could be utilized for extracting hydrogen from these reservoirs. However, due to the limited understanding of marine hydrogen energy and the unique challenges associated with hydrogen storage and transport, there is a pressing need to develop specialized detection and extraction technologies tailored to marine hydrogen energy. Advance layout in this direction will provide the necessary technical support for the exploitation and utilization of marine hydrogen energy and spur revolutionary breakthroughs in various technological fields.

The Duvernay shale in the Upper Devonian is a formation of shale rich in oil and gas, created during the peak transgression period in the Western Canada Basin. This study aims to elucidate the factors influencing hydrocarbon enrichment in the Duvernay shale through the analysis of sedimentary elements, fluid distribution, reservoir quality, and drivers of organic matter enrichment. Utilizing data from cores, well logging, thin sections, scanning electron microscopy, 3D pore reconstruction, and organic geochemistry, this research examines the geological context to determine that oil and gas accumulation in the Duvernay shale is governed by the sedimentary conditions of siliceous shale, organic matter thermal maturity, reservoir quality, and stable structural settings. The Duvernay Formation, situated in a deep-water shelf environment during the Late Devonian, comprises primarily marl, mudstone, and shale lithologies. it is observed that Type II and III marine organic matters are abundant in the Duvernay shale through the identification of ten lithofacies, with siliceous shale being predominant. These organic materials exhibit moderate thermal maturity, falling within the condensate to wet gas stage, leading to a high condensate to gas ratio. Oil and gas are predominantly found in siliceous shale with organic pores created by clay-grade minerals, showcasing organic and intra-granular pore types. The shale possesses high effective porosity, with well-connected and horizontally distributed pores that display vertical connectivity characteristics. Diagenesis enhances the physical properties of the shale reservoir, while natural fractures boost permeability. Ultimately, the preservation of Duvernay shale oil is heavily dependent on a stable structural setting.

The North Yellow Sea Eastern Sub-basin is a typical Meso-Cenozoic small superimposed faulted basin, currently in its early exploration phase. Comprehensive studies on the complex tectonic-thermal evolution in this region are lacking. By integrating vitrinite reflectance (R_o) and apatite fission track (AFT) analysis, the thermal history and geothermal gradient of the Meso-Cenozoic era have been reconstructed in this basin, alongside assessments of erosion thickness and processes. These investigations have enabled an analysis of the thermal evolution history of Middle-Upper Jurassic source rocks. Results indicate that paleo-heat flow peaked at 75-90 mW/m² during 120-100 Ma, decreased to 60 mW/m² at 40 Ma, and then rose to 70 mW/m² at present. Correspondingly, temperatures declined gradually from 100-70 Ma but rapidly dropped during 70-40 Ma, with the geothermal gradient shifting from 34-36 ℃/km to 23 ℃/km before rising to 28 ℃/km. Overall, paleo-geothermal gradient and heat flow were higher before the Late Cretaceous, aligning with the transition from a faulted basin to a depression basin. Intense uplift and erosion occurred during the Late Cretaceous-Eocene, resulting in an erosion thickness of approximately 1.0-1.5 km. Deposition slowed or ceased during 100-90 Ma, with significant uplift occurring during 85-40 Ma, especially rapid during 70-40 Ma. The tectonic-thermal history has influenced hydrocarbon generation, with both Middle and Upper Jurassic source rocks experiencing early hydrocarbon generation during the Late Jurassic-Early Cretaceous. In the central depression, erosion thickness was thinner in the Lower Cretaceous-Eocene, while deposition thickness was thicker in the Oligocene-Quaternary, leading to higher present-day strata temperatures and favoring late hydrocarbon generation. Exploration efforts for potential accumulation should focus on areas surrounding hydrocarbon generation depressions.

Investigating the correlation between multi-phase tectonic activity and deep reservoir formation is crucial for oil and gas exploration endeavors. Utilizing seismic data from the Halahatang, Jinyue, and Fuman oilfields, coupled with an analysis of field geological outcrop faults, various seismic fine interpretation methods were employed to delineate faults within the study area. Building upon an understanding of the Middle Cambrian Yuertusi source rock and the characteristics of the Caledonian, Hercynian, and Himalayan accumulation stages, faults controlling oil accumulation were classified into four stages: Early Caledonian, Middle and Late Caledonian, Late Hercynian, and Himalayan. Further analysis of the inheritance relationship, source characteristics, and adjustment effects of multi-stage fractures, along with an assessment of various types of karst fracture-cavity reservoir development, led to discussions on the variations in karst fracture-cavity reservoirs under the influence of strike-slip faults in the study area. Key findings include: (1) Identification of primary factors influencing oil and gas reservoirs, including intra-source faults from the early Caledonian normal fault system facilitating hydrocarbon expulsion from Cambrian source rocks, and outer source faults formed during the late Caledonian enabling communication with source rocks for oil and gas migration and accumulation. Four source rocks-linking models were established based on this understanding. (2) Recognition of three main hydrocarbon generation periods in the study area: late Caledonian, Hercynian, and Himalayan, with inherited development of northwest strike-slip fractures into the Permian during the Late Hercynian period, impacting Garridonian reservoirs, and destruction and adjustment of early oil and gas reservoirs by northeast strike-slip fault systems inherited to the Neogene during the Himalayan period. Three modes of oil and gas remigration were established. (3) Establishment of six types of strike-slip fault control grades based on fracture matching relationships, along with classification of Middle and Late Caledonian strike-slip fault zones in the study area. A mining status map revealed a high matching degree between differential reservoir-controlling faults and oil and gas production. (4) Joint control of reservoirs by strike-slip faults and karstification in the study area, with an established matching relationship between the multi-stage fault system and various types of karst fracture-cavity reservoirs. This understanding has been successfully applied to well location exploration in the study area, yielding favorable results and providing guidance for the exploration and development of karst fracture-cavity reservoirs controlled by strike-slip faults.

The Paleoproterozoic Eburnean orogenic granites are widely exposed in the western part of Angola, offering an ideal setting to study the magmatism and tectonic evolution of the Angola Block. This paper presents systematic studies of petrology, petrogeochemistry, and zircon U-Pb geochronology on the exposed granites in the Dondo area, northern Angola Block. The results indicate that the emplacement ages of porphyritic biotite monzonite granite and biotite monzonite granite in the Dondo area are 1983.3±7.7 Ma and 1956.6±7.5 Ma, respectively, both products of middle Paleoproterozoic magmatic activity. The whole-rock samples are characterized by high SiO₂, ALK, 10000Ga/Al, FeO^T/(FeO^T+MgO), and Zr+Y+Nb+Ce, and low MgO, TiO₂, CaO, and P₂O₅. Trace elements are enriched in Rb, K, Th, U, Zr, and Hf, and depleted in Sr, Nb, Ta, P, and Ti. All samples are enriched in LREE and depleted in HREE, with no significant negative Eu anomaly. The crystallization temperature, calculated using zircon saturation thermometry, ranges from 757 to 889 ℃. Based on these geochemical characteristics, the granites in the Dondo area are classified as A2-type granite. Mineralogical and geochemical features suggest that the porphyritic biotite monzonite granite and biotite monzonite granite were generated by the mixing of crust-derived melts and mantle-derived mafic magma. The similar formation ages within analytical error, identical mineral assemblages, and consistent variations in major and trace elemental compositions indicate that their parental magma originated from a common magma chamber, with lithological differences resulting from melt extraction processes. It is proposed that the magma producing the potassium feldspar porphyry resided in the deep crust for an extended period, leading to stable crystallization of potassium feldspar, increased viscosity and density, and a frozen state of the magma. Subsequent thermal disturbance and volatile enrichment from mantle-derived magma injection rapidly reactivated the frozen magma chamber, resulting in crystal-melt separation. The extracted melt formed biotite monzonite granite, while magma mixed with pre-existing crystals formed porphyritic biotite monzonite granite. Combining regional and global tectonic evolution, it is suggested that the granites from the Dondo area formed in a post-collision tectonic environment between the São Francisco Craton and the Congo Craton. The Paleoproterozoic magmatic events in the Angola Block are likely responses to the Columbia Supercontinent assembly.

The Permian Liangshan Formation in the Pudi area of northwest Guizhou Province directly overlays the Cambrian Loushanguan Formation, exhibiting abnormal lithium enrichment in its clay rocks. Studying the enrichment mechanism provides valuable insights for lithium resource development and evaluation in clay rocks, as well as understanding lithium accumulation mechanisms. Various analytical methods including LA-ICP-MS, mapping, AMICS, XRD, SEM, along with U-Pb chronology, are employed to elucidate the elemental geochemistry, provenances, and occurrence states of lithium enrichment in clay rocks. Results reveal lithium-rich clay rocks primarily in the middle and lower sections of the Permian Liangshan Formation, with their occurrence controlled by the karst unconformity surface of the underlying Loushanguan Formation dolomite. Enriched elements such as Li, Ga, V, Nb, Zr, and F are identified, while Ba and Sr are relatively depleted. Light rare earth elements (La, Ce, Nd) predominate, with Y as the main heavy rare earth element. Lithium-rich clay rocks comprise terrestrial deposits with characteristics of terrestrial, transitional, and marine phases, formed in an oxidized environment under a tropical-subtropical warm and humid climate. The zircon age spectrum exhibits five peaks at 2.5 Ga, 1.4 Ga, 980 Ma, 780 Ma, and 530 Ma, with the main peaks at 980 Ma, 780 Ma, and 530 Ma. Lithium in the clay rocks mainly occurs within kaolinite, indicating multiple sources, with impure dolomite of the Cambrian Loushanguan Formation likely being the primary source.

In recent years, significant progress has been achieved in uranium exploration within the Lower Cretaceous strata of the northern Ordos Basin. Notably, numerous large-scale industrial drill holes, revealing substantial ore bodies, have been unearthed within the lower segment of the Huanhe Formation in the Tela’aobao area and its vicinity. Organic matter serves as a pivotal factor in sandstone-type uranium (U) mineralization. Despite the absence of visible organic matter in U-mineralization sandstone, there remains a dearth of research on the specific organic matter types influencing the U-mineralization process, thus leaving the uranium mineralization process ambiguous. This study focuses on the sandstone-type U-deposit in Tela’aobao area, located in the northern region of the Ordos Basin. Through comprehensive observation of drill cores and subsequent laboratory analyses, we have delineated the organic matter types present in the U-mineralization sandstone and investigated the relationship between organic matter sources and U-mineralization. Our findings indicate that the organic matter in U-mineralization sandstones is primarily disseminated and exhibits flowing characteristics within sandstone pores. This macromolecular organic matter, akin to bitumen, possesses a complex structure and low evolutionary degree, suggesting a mixed origin from both higher plants and lower aquatic organisms. The formation of dark gray and gray-brown uranium ores may be linked to exuded organic fluids. Ore-forming elements are likely transported primarily in colloidal form via organic matter complexes. Moreover, changes in physical-chemical conditions may constitute the primary driver behind the differentiation and mineralization of ore-forming fluids.

Deformation characteristics and timing of the fold-and-thrust belt are the key in verifying the geodynamic end-member models of continental lithosphere of the Tibetan Plateau and its periphery. K-Ar dating of authigenic illite from fault gouge provides an effective means in determining the timing of deformation events in the fold-and-thrust belt. The Xianshuihe-Anninghe-Xiaojiang fault is a large-scale sinistral strike-slip fault system accompanying the collision of the Indo-Eurasian Plates and the lateral extrusion of the Tibetan Plateau. The deformation process of the Xianshuihe-Anninghe-Xiaojiang fault system can provide significant insight into how the far-field stress of the Indo-Eurasian collision is transferred eastward. In this study, we focus on the Mianning-Xichang segment of the Anninghe fault. Based on the structural analyses characterizing fault kinematics, we dated authigenic illite from fault gouge to constrain the timing of deformation events along the Anninghe fault. Structural analyses reveal that the Anninghe fault experienced thrusting deformation under the EW-oriented compression. Detailed study of illite clay mineralogy and K-Ar dating of different grain-size fractions of the fault gouge suggest that the components of high-temperature 2M₁ illite relative to those of low-temperature 1M/1M_d illite polytype gradually decrease with the decreasing K-Ar ages. These results show that different illite K-Ar ages for different grain sizes are due to the mixing of two illite polytypes, represented by detrital 2M₁ and authigenic 1M/1M_d end members. Illite Age Analysis reveals that the K-Ar age for authigenic 1M/1M_d illite is (42.6±9.4) Ma, suggesting that the thrusting deformation along the Anninghe fault occurred in the middle Eocene. Integrated with previously published tectonic, sedimentology, low-temperature thermochronology and paleomagnetism studies, it is suggested that the fold-and-thrust belt in the hinterland of the Tibetan Plateau and its periphery underwent quasi-contemporaneous tectonic compression deformation during the middle Eocene. The dynamic mechanism of this event may be related to the reactivation of the pre-existing tectonic belt caused by the combination of the hard collision of the Indo-Eurasian Plates and intra-continental subduction of the terranes within the Tibetan Plateau. The mid-Eocene thrusting event along the Anninghe fault suggests that the far-field stress from the early stage of the Indo-Eurasian collision has been transferred to the southeastern margin of the Tibetan Plateau.

The regional tectonic uplift history and landscape evolution can be simulated based on the spatial distribution characteristics of slope-break knickpoints in bedrock channels, which result from the combined action of tectonic activity and water erosion. This study identifies slope-break knickpoints and knickpoint belts in the upper reaches of the Minjiang River using slope-area analysis and integral analysis methods. It then simulates the uplift history, formation times of river mouth knickpoints, and headwater migration processes and rates of slope-break knickpoints, combined with longitudinal river profiles. The findings are as follows: (1) In the upper reaches of the Minjiang River, slope-break knickpoints exhibit distinct layered distribution characteristics, forming three knickpoint belts at 1300 m, 2500 m, and 3500 m in the main channel and tributaries. Data from these knickpoint belts suggest that the area has experienced three relatively intense tectonic movements since the Early Pleistocene. (2) The tectonic uplift history in this region can be divided into four periods: slow uplift (20-12 million years ago), accelerated uplift (12-8 million years ago), stable uplift (8-2 million years ago), and intense uplift (since 2 million years ago), with an uplift rate of 0.6 mm/a since 2 million years ago. (3) The average horizontal migration rate of the 3500 m slope-break knickpoint ranges from 38.0 to 127.9 km/Ma, likely forming around 1.3 million years ago with a migration rate of 79.1 km/Ma according to the minimum residual squares method and optimal fitting method.

Situated at the southeast margin of the Tibetan Plateau, the Yunnan region is pivotal for investigating Late Cenozoic climatic changes. While considerable research has focused on the paleoclimate and paleoenvironmental evolution of Yunnan, the understanding of climate change from the Late Miocene to Pliocene primarily relies on carbon isotope and pollen records. Consequently, there is a dearth of high-resolution, continuous paleoclimatic records documenting humidity changes during this period. This study utilizes sediment cores from the Late Miocene to Pliocene in the Zhaotong Basin, northeastern Yunnan Province. Through sediment grain size analysis, the sedimentary sequence, lithological characteristics, and sedimentary structures indicate that the Zhaotong Basin was predominantly characterized by swamp-subfacies sedimentary environments during 8.8-6.2 Ma, transitioning to shallow lake subfacies during 6.2-2.8 Ma, and lakeside subfacies during 2.8-2.6 Ma. Grain size parameters of sediments in the Zhaotong Basin exhibit a drying trend of the South Asian monsoon during the Late Miocene to Pliocene. Combined with clay mineral and chemical weathering results from borehole data in the early period, it is inferred that the South Asian monsoon gradually weakened from the Late Miocene to Pliocene, primarily influenced by global cooling and decreasing global CO₂ concentrations.

The Tian Shan orogenic belt experienced activation and uplift during the Cenozoic era, attributed to the remote effects of the India-Asia collision. Adjacent to the southern margin of the Tian Shan orogenic belt, the northern Tarim Basin underwent bending subsidence and accumulated extensive Cenozoic strata, providing a robust foundation for investigating the uplifting process of the southern Tian Shan Mountains. In this study, we employ finite elastic plate numerical simulation to model basement subsidence profiles across various Cenozoic periods. Our findings underscore the control of basin subsidence by sedimentary load and tectonic load, with sedimentary load exerting a significantly greater influence on basin subsidence than tectonic load from ~5.3 Ma to the present. The rate of load change in the southern Tian Shan structure exhibits gradual increase from ~66 Ma to ~26.3 Ma, followed by a rapid ascent since ~5.3 Ma. Our analysis indicates that the initial uplift phase of the Cenozoic in the southern Tian Shan was confined to the Paleogene, with its relative elevation escalating from 400 meters to 2500 meters. Although the relative elevation of the southern Tian Shan has remained stable since ~5.3 Ma, the height of tectonic load continues to rise. This phenomenon is attributed to the intensified basin subduction, which has constrained the average elevation of the orogenic belt, thereby establishing a relative equilibrium between erosion and uplift processes in the southern Tian Shan.

Soil methane emissions in permafrost regions are integral components of the global carbon cycle and terrestrial ecosystem, playing a pivotal role in the feedback mechanism of carbon sink on climate change, thus warranting focused research in the domain of global climate change. The origins of soil methane emissions in permafrost regions primarily stem from microbial methane production and gas release from frozen soil layers and natural gas hydrates. While research on microbial gas sources is relatively advanced, investigations into methane emissions from frozen soil layers and natural gas hydrate gas sources are still in the qualitative analysis stage. Influential factors such as soil temperature, moisture, water table conditions, organic matter content, surface vegetation conditions, and others can significantly influence various stages of methane production, transport, and oxidation. Modeling stands as the primary approach for evaluating and forecasting soil methane emissions in permafrost regions, encompassing both early statistical models and more recent process models based on the mechanisms of methane emission from soil. Although the synthesis of research on methane emissions from permafrost soils has yielded insights into gas sources and single-factor effects, there remain gaps in the study of multi-source methane emission, particularly concerning methane release from permafrost soil and gas hydrates. Furthermore, the analysis of causal mechanisms and driving forces under multiple factors is lacking in the investigation of influential factors. Comprehensive monitoring research employing diverse methods and factors, such as metagenomic analysis of methane-producing microorganisms and isotope tracing of multi-gas source soil methane emissions, can be integrated with satellite remote sensing and other large-scale observation results to refine process models of methane emissions from permafrost soils. Given that changes in the carbon cycling system of the Qinghai-Tibet Plateau, revered as the “Third Pole” of the world, will exert significant impacts on climate change in Asia and globally, further exploration of soil methane emissions on the Qinghai-Tibet Plateau is imperative to facilitate the quantitative assessment of regional carbon emissions and advance global climate change research.

The granitic regolith, prevalent across South China, plays a pivotal role in geomorphological evolution, ecological dynamics, and mineral resource management. Understanding the formation and evolution of regolith hinges upon fundamental parameters such as production rate. The U-series disequilibrium method serves as a crucial geochemical tool for determining regolith production rates, yet its application in China has been limited due to the unavailability of a spike. In this investigation, the U-series disequilibrium method was employed to ascertain the production rate of granitic regolith in Longnan, Jiangxi Province. Results indicate U and Th contents in the regolith profile ranging from (3.25-3.39)×10^-6 and (41.46-47.67)×10^-6, respectively. Activity ratios of (²³⁴U/²³⁸U)_a, (²³⁰Th/²³⁴U)_a, and (²³⁰Th/²³²Th)_a vary between 1.008-1.023, 1.063-1.112, and 0.239-0.271, respectively. Consequently, utilizing the U-series disequilibrium method to fit uranium isotopes, the evolution time of the regolith within the 20-120 cm stratum is estimated at ~841 ka, with a regolith production rate determined to be ~1.2 m/Ma. Surface cover emerges as the predominant control factor over regolith production rate, with minimal influence from climate and tectonic activity. Furthermore, the regolith’s evolution state is identified as non-steady state, evidenced by a significantly lower regolith production rate compared to denudation rates determined by cosmogenic nuclides, resulting in gradual thickness reduction.

The Geochemical Survey of Land Quality conducted by the China Geological Survey from 1999 to 2021 collected 670321 composite surface soil samples (0-20 cm) and 167746 composite deep soil samples (150-180 cm) across approximately 2.665 million km², covering most densely populated areas and farmland in China. Each composite sample underwent analysis for 54 chemical indicators using a standardized method (Ag, As, Au, B, Ba, Be, Bi, Br, Cd, Ce, Cl, Co, Cr, Cu, F, Ga, Ge, Hg, I, La, Li, Mn, Mo, N, Nb, Ni, P, Pb, Rb, S, Sb, Sc, Se, Sn, Sr, Th, Ti, Tl, U, V, W, Y, Zn, Zr, SiO₂, Al₂O₃, TFe₂O₃, MgO, CaO, Na₂O, K₂O, pH, total carbon and organic carbon). This study examines the methodologies and fundamental characteristics of establishing geochemical background values, baseline values, and their upper and lower threshold limits for the 54 soil indicators in China. Additionally, it compares various thresholds of the background of cultivated soil with existing environmental and nutrient-related standards. The findings reveal that, on a national scale, most surface soil indicators exhibit content and distribution patterns inherited from deep soil, with notable changes observed only in the background values of specific indicators such as organic carbon, total carbon, N, S, Se, Hg, Br, Cd, and P. Spearman correlation coefficients for each indicator in both surface and deep soil consistently exceeded 0.50. Across the entire cultivated land in the country, only the upper limit of Cd background exceeds the risk screening value for soil contamination of agricultural land, while the upper limits of Hg, As, Pb, and Cr backgrounds were all lower than their respective risk screening values. Due to significant spatial distribution differences in many indicators, a simplistic national-scale background delineation is insufficient for refined soil resource management. Hence, the establishment of geochemical background/baseline at different scales based on national data is crucial to inform government regulations and standards for soil environmental quality management.

Soil environmental capacity assessments often overlook the influence of heavy metal sources, yet analyzing these sources is essential for understanding regional environmental dynamics and mitigating heavy metal pollution effectively. In Guangdong Province, Shantou stands as a prominent economic hub with significant industrial and urban development, raising concerns about soil heavy metal pollution and the region’s environmental capacity. This study focused on surface soil in Shantou, utilizing 511 soil samples to investigate environmental capacity. Employing GIS technology and geostatistics, we applied the enrichment factor method to evaluate heavy metal enrichment. Principal component analysis and correlation analysis were used to discern heavy metal sources, while the comprehensive index technique was employed to analyze soil heavy metal environmental capacity features and spatial distribution. The surface soil in Shantou exhibited concentrations of As, Cd, Cr, Cu, Hg, Ni, Pb, and Zn were 4.84 (0.23-119.4), 0.14 (0.013-4.94), 36.05 (2.2-109.1), 19.08 (2.9-453.1), 0.18 (0.008-1.394), 13.29 (3.7-229.6), 51.73 (9.9-1000), and 79.83 (9.2-309.5) mg/kg. Except for moderate enrichment of soil Pb, other heavy metals showed mild enrichment. Principal component analysis revealed five principal component sources: mixed natural background and agricultural activities, mixed agricultural activities and industrial production, industrial production, mining activities, and stone mining sources. Soil heavy metals’ static environmental capacity ranking was As > Zn > Cr > Ni > Pb > Cu > As > Hg > Cd, with a moderate comprehensive environmental capacity level. However, extensive human activities have led to generally low environmental capacities for heavy metals in the towns of Yanhong, Guiyu, and Chendian, posing certain risks. The findings of this study can serve as a scientific reference for early warning and management of soil heavy metal pollution, as well as the remediation and governance of contaminated soil in Shantou City. Additionally, it provides valuable insights for enhancing environmental capacity and soil quality.

Adakitic rocks hold significant geodynamic and metallogenic implications, and accurately determining their tectonic setting is crucial for understanding regional tectonic-magmatic evolution. However, due to the diverse sources, heat regimes, and magma generation mechanisms of adakitic rocks, conventional low-dimensional geochemical methods face limitations in tectonic setting identification. With the exponential growth of geoscience data and advancements in artificial intelligence, machine learning offers a novel approach to address this challenge. In this study, we integrate machine learning with geological big data to develop a high-precision adakitic tectonic setting discrimination model and visual representation. We compiled major and trace elements geochemical data from 1075 adakitic rocks worldwide and employed unsupervised learning techniques such as principal component analysis and t-distributed stochastic neighbor embedding for high-dimensional data reduction. Various machine learning algorithms including random forest, support vector machine, artificial neural network, and K-nearest neighbor were trained. Consequently, we established a Gaussian kernel support vector machine adakitic rock tectonic setting discriminator with 98.5% accuracy and proposed a Ba versus Sr/Nd diagram to delineate three tectonic settings: convergent margin, intraplate volcanism, and Archean craton (comprising greenstone belts). This study broadens the application of machine learning in adakitic rock tectonic setting analysis, offering fresh insights into tectonic-magmatic processes investigation.

Volcanic activity, as a significant source of geological carbon emissions and contributor to deep carbon cycles, transports carbon from the Earth’s interior to the atmosphere. Greenhouse gases, particularly CO₂, emitted by volcanic regions, exert a profound influence on global climate dynamics. Against the backdrop of global warming and initiatives such as the “carbon neutral” program, accurately estimating the flux of greenhouse gases from volcanic regions and assessing their impact on global carbon budgets are imperative. This paper elucidates the primary characteristics and survey methodologies for quantifying greenhouse gas emissions in volcanic areas. It proposes employing geostatistical methods to simulate CO₂ sampling data from volcanoes, exemplified by Mount Etna, Italy. Additionally, the feasibility of incorporating covariates for cokriging interpolation simulations is analyzed, with comparisons drawn against ordinary kriging interpolation methods. Our findings reveal a correlation between CO₂ release flux and soil temperature in volcanic regions, indicating that integrating soil temperature into cokriging interpolation simulations can mitigate error indices in the results. This research offers critical insights for quantitatively assessing the impact of volcanic activity on climate change and enhancing early warning systems for volcanic hazards.

Against the backdrop of big data science emerging as a new scientific paradigm, traditional qualitative geological research methods are transitioning towards quantitative research grounded in the concept of geological big data. Leveraging data from over 5300 gold deposits, this study quantitatively analyzes the metallogenic density and intensity of gold deposits across provincial (autonomous region), city, county, and Grade III metallogenic zones in China, as well as the metallogenic intensity of gold deposits across different metallogenic ages. The findings reveal a pronounced trend of spatiotemporal concentration distribution of gold deposits in China. Spatially, gold deposits exhibit regional concentration, with areas such as Jiaodong and Qinling emerging as high-density and high-intensity gold deposit regions. Xinjiang boasts the highest number of gold deposits, whereas Shandong Province exhibits the highest metallogenic density and intensity. At the prefectural city level, Yantai City in Shandong Province and Chengde City in Hebei Province stand out as the only two prefecture-level cities in China with over 100 gold deposits each. Among these, Yantai City in Shandong Province ranks first in terms of mineral producing areas, metallogenic density, and metallogenic intensity. At the county level, Toli County in Tacheng region of Xinjiang houses the largest number of gold deposits (52), Tongling district in Anhui exhibits the highest metallogenic density, and Laizhou City in Yantai, Shandong, holds the largest gold reserves (2341 t) and the strongest metallogenic intensity (1.35 t/km²). Analysis of metallogenic belts reveals that the eastern section of the northern margin of the North China block, particularly metallogenic belt (III-57), boasts the largest number of gold deposits (345), while the Jiaodong metallogenic belt (III-65) exhibits the highest metallogenic density and intensity. Temporally, gold deposits in China display an unbalanced distribution characterized by old-weak and new-strong mineralization, significant differences between north and south, a broad metallogenic time span, and substantial Cenozoic gold resource potential. The Yanshanian period emerges as the most significant metallogenic period for gold deposits in China, characterized by high metallogenic intensity, with 10.5 deposits/Ma and 93 t/Ma. Looking ahead, the focal point of China’s gold geological efforts will be to reinforce gold prospecting activities to safeguard national financial security.