This session targets innovative tools and methodologies that are redefining raw material exploration and characterization. We emphasize multi-scale, multi-source and multi-disciplinary approaches that integrate advanced sensing, modelling, automation and data-driven solutions. The session focuses in particular on method innovations in the field of remote sensing, geophysics, geochemistry, raw material processing, as well as on recycling processes.
We encourage interdisciplinary studies which use a combination of methods to solve challenges as diverse as, but not limited to:
• Next-generation sensing and imaging: non-destructive techniques, core scanners, and airborne/ground-based sensors for high-resolution, accurate, precise, and efficient resource identification.
• Smart field and analytical approaches: geophysical and geochemical mapping, isotope dating, and novel sampling workflows for multi-scale ore body understanding.
• Digital modelling and simulation: advanced conceptual models and quantification methods for deposits and mineral systems.
• Automation and real-time decision-making: AI-driven, automated data processing that enhances resource management, mining selectivity, and recycling efficiency.
• Information integration and visualization: innovative platforms for merging data streams from diverse sensors to improve accuracy and reduce uncertainty.
• Data-driven discovery: machine learning, geostatistics, data fusion, and computational advances unlocking new insights in mineralogy and geochemistry.
Orals: Thu, 7 May, 08:30–12:30 | Room -2.43
Modern mineral exploration and production increasingly rely on advanced spatial modeling techniques capable of handling complex geological settings characterized by structural discontinuities, irregular sampling, and physical barriers. Conventional covariance models based on Euclidean distance measures often fail to adequately represent such environments, limiting their effectiveness in resource estimation and uncertainty quantification. The adoption of non-Euclidean distance metrics offers a promising pathway toward more realistic geological modeling and improved decision-making in mining operations.
This contribution presents recent advances in geostatistical covariance modeling based on the Linearly Damped Harmonic Oscillator, implemented through the Harmonic Covariance Estimator (HCE) and the Advanced Harmonic Covariance Estimator (AHCE). Nine case studies are used to demonstrate the applicability and robustness of these models across a broad range of mining-related scenarios, including univariate and multivariate mineral datasets, anisotropic orebody structures, unevenly distributed sampling, conditional simulations for uncertainty assessment and Gaussian anamorphosis models. Comparisons are made against established covariance models commonly used in mining geostatistics under both Euclidean and non-Euclidean distance frameworks.
Model performance is evaluated using leave-one-out cross-validation and eigenvalue-based validity testing. Results show that harmonic covariance models remain mathematically valid and predictive in complex geological environments where traditional approaches often fail. These advances provide a flexible and reliable framework for next-generation mineral resource modeling, supporting more accurate exploration targeting, improved production planning, and sustainable resource management in the mining industry of tomorrow.
The research project is implemented in the framework of H.F.R.I call “Basic research Financing (Horizontal support of all Sciences)” under the National Recovery and Resilience Plan “Greece 2.0” funded by the European Union – NextGenerationEU (H.F.R.I. Project Number: 16537)
M. D. Koltsidopoulou, A. Pavlides, D. T. Hristopulos, E. Α. Varouchakis, 2025, Enhancing Geostatistical Analysis of Natural Resources Data with Complex Spatial Formations through non-Euclidean Distances, Mathematical Geosciences, in print.
A. Pavlides, M. D. Koltsidopoulou, M. Chrysanthi, E. A. Varouchakis, 2025. A Kernel-Based Nonparametric Approach for Data Gaussian Anamorphosis, Mathematical Geosciences, https://doi.org/10.1007/s11004-025-10251-z
E.A. Varouchakis, M. D. Koltsidopoulou and A. Pavlides, 2025, Designing Robust Covariance Models for Geostatistical Applications, Stochastic Environmental Research and Risk Assessment, https://doi.org/10.1007/s00477-025-02982-6
How to cite: Varouchakis, E., Chrysanthi, M., Koltsidopoulou, M., and Pavlides, A.: Advanced Geostatistical Models for Robust Mineral Resources Estimation in Complex Geological Settings, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17036, https://doi.org/10.5194/egusphere-egu26-17036, 2026.
The northeastern Estonian Precambrian basement, encompassing the Tallinn, Alutaguse, and Jõhvi domains, forms part of the eastern sector of the Fennoscandian Shield. This crustal segment comprises Paleoproterozoic back-arc volcanic–sedimentary successions intruded by Svecofennian granitoids and metamorphosed to amphibolite–granulite facies. Its lithological architecture and metallogenic characteristics show strong affinities with established mineralised provinces of southern Finland and central Sweden, including the Orijärvi and Bergslagen districts.
In this study, more than 500 historical drill cores, together with associated legacy geophysical datasets, were reanalysed to re-evaluate the mineral and critical-metal potential of the NE Estonian basement. Base- and precious-metal anomalies (Cu–Zn–Pb; Au–Ag–As–Sb) are spatially associated with magnetite-bearing and sulphide–graphite gneisses. High-resolution MSCL-XYZ scanning of archived drill cores further reveals a range of multi-element associations indicative of diverse mineral systems, including Ni–Co–Cr, Mo–W–Bi, Sn–Zn–Cd, Cu–Ni, Nb–Y–P, and Au–Ag–As–Sb–Bi–W–Se–Sn. These signatures delineate previously unrecognised prospective intervals across all three basement domains.
A compositional geostatistical workflow was applied to historical whole-rock geochemical data to mitigate biases arising from heterogeneous sampling density and analytical variability. Exploratory analyses conducted on raw datasets were complemented by centred log-ratio (clr) transformation, which enhanced coherence in multivariate patterns. Clr-based spatial maps, principal component analysis, and heat-map visualisations significantly improved the reliability of regional-scale interpretations and reduced artefacts related to mismatched neighbouring datasets.
Lithological descriptions from historical drilling, often incomplete or inconsistent, were reinterpreted using major-element geochemistry, while trace-element data were reassigned within a refined Tallinn–Alutaguse–Jõhvi basement framework. Integration of these geochemical reclassifications with gravity and magnetic data constrains subsurface architecture and strengthens correlations with mineral systems recognised in the southern Svecofennian domain and the Bergslagen province.
Overall, the integrated geochemical, geostatistical, and geophysical approach provides an updated metallogenic framework for the NE Estonian basement and identifies new exploration targets for critical raw materials, supporting ongoing research within the Horizon Europe DEXPLORE programme.
How to cite: Solano Acosta, J. D., Graul, S., Soesoo, A., All, T., and Vind, J.: Critical Raw Material Potential and Mineral System Structure of the Northeastern Estonian Basement: A Geochemical, Geostatistical, and Geophysical Review, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20287, https://doi.org/10.5194/egusphere-egu26-20287, 2026.
The composition contents of various minerals in the rock are a key concern in geophysical exploration and development. It is essential for lithology classification, the quantitative assessment of mineral resource potential, and reserves prediction. However, accurately calculating these mineral components is often highly challenging for formations with complex lithology, particularly when core samples and formation elemental logging data are scarce. In recent years, with the rapid development of artificial intelligence, utilizing big data and deep learning technologies to improve the accuracy and efficiency of well logging interpretation has become a research hotspot. Nevertheless, traditional data-driven models suffer from a lack of interpretability, which imposes certain limitations on their practical application. As a novel model integrating physical laws, Physics-Informed Neural Networks (PINNs) can constrain prediction results, rendering them more physically meaningful.
In this study, we propose a mineral content prediction model specifically designed for formations with complex mineral types. The model is capable of accurately calculating mineral contents using conventional logging data. First, based on the mineral types present in the formation, forward modeling is used to generate data and construct the training dataset. Subsequently, a CNN (Convolutional Neural Network) model is employed to predict the mineral content. By simultaneously constructing data loss and physical loss functions, the interpretability of the prediction results is ensured. The physical loss is mainly constructed by the volume model. The validity of the model is verified using forward modeling data. Finally, the model is applied to the processing of real logging data. The prediction results demonstrate good consistency with the mineral content obtained from X-ray Diffraction (XRD) analysis of core samples indicating that the model can accurately reflect the variations of complex mineral contents. This study provides a new method for the evaluation of mineral content, which is expected to offer a potential technological pathway for the identification of deep-seated ore bodies and the estimation of resource reserves.
This work is supported by National Science and Technology Major Project for Deep Earth Probe and Mineral Resources Exploration under Grant 2025ZD1008500.
How to cite: Shu, J., Zou, C., and Peng, C.: A method for estimating the mineral contents from well logs using physics-informed neural networks, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11149, https://doi.org/10.5194/egusphere-egu26-11149, 2026.
Electrical Resistivity Tomography (ERT) provides an effective means for probing the internal electrical structure of rock cores and plays an important role in understanding the electrical properties of ore-related geological bodies. Recovering informative structural representations from limited and highly coupled measurement data, however, remains challenging, particularly for drill cores, where complex resistivity distributions are commonly observed. Restricted electrode configurations and scale effects further hinder the ability of conventional inversion schemes and existing convolutional neural network (CNN)–based approaches to preserve structural continuity and spatial correlations in core-scale ERT imaging.
In this study, we investigate a dual-branch CNN–Transformer architecture designed for learning electrical structure representations from core-scale ERT data. The proposed approach adopts an end-to-end image-to-image learning paradigm to explore how complementary data organizations can be leveraged for representation learning. Two dedicated Transformer branches are incorporated: the first branch exploits potential difference data acquired from multiple sets of sequentially excited adjacent electrode pairs with consistent relative spatial configurations, while the second branch utilizes potential difference measurements collected at multiple spatial locations under a single electrode excitation.
By integrating the local feature extraction capability of CNNs with the global dependency modeling strength of Transformers, the proposed architecture aims to construct more expressive representations of complex electrical structures, thereby supporting improved structural coherence and spatial resolution in ERT imaging. Preliminary results, evaluated using quantitative imaging metrics including correlation coefficient and structural similarity index, suggest that the learned representations capture coherent electrical features under varying anomaly geometries, resistivity contrasts, and spatial distributions. These early findings demonstrate the feasibility of combining CNNs and Transformers for electrical structure representation learning in core-scale ERT and provide a methodological foundation for subsequent development of effective deep learning–based inversion strategies oriented toward deep mineral exploration applications.
This work is supported by National Science and Technology Major Project for Deep Earth Probe and Mineral Resources Exploration under Grant 2025ZD1008500.
How to cite: Shen, W., Zou, C., and Peng, C.: Learning Electrical Structure Representations from Ore-Bearing Cores ERT Data Using a Dual Branch CNN Transformer Architecture, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7353, https://doi.org/10.5194/egusphere-egu26-7353, 2026.
Modern society depends on raw materials for construction and infrastructure, but also increasingly for batteries, renewable energy, electronics, and the broader green transition. At the same time, mining faces tightening environmental expectations, safety requirements, and rising operational costs. The challenge is clear: how can we produce the minerals Europe needs while improving safety, lowering environmental impacts, and strengthening public trust? This book addresses that question by presenting practical, tested solutions based on a new generation of sensing and data technologies spanning Earth observation (EO) satellites, drone-based measurements, GNSS positioning, and proximity (in situ) sensing.
The volume was initiated and is primarily built on results from the EU Horizon 2020 project GoldenEye, which advanced the use of innovative monitoring and characterisation technologies to support safer and more sustainable mineral operations. GoldenEye’s central idea is simple but powerful: mining can be measured, understood, and managed more intelligently when we integrate information across scales from satellites that view entire mining districts, to drones that deliver site-scale detail, to local sensors and positioning systems supporting real-time operations underground and in active pits. Together, these technologies create objective, repeatable evidence of change. They can detect subtle ground movements, monitor tailings stability, map mining activity, characterise rock and ore properties, track vegetation and land-use evolution, and support early warning for environmental risks.
Crucially, the book treats mining as a complete life-cycle system, not only as “exploration and extraction”. The approaches discussed apply from early mineral exploration and resource evaluation, through mine development and active production, and onwards to closure, post-closure monitoring, and even mine reuse. For exploration, EO and hyperspectral methods can improve mineral targeting and reduce the need for costly field campaigns in remote areas. During operations, high-resolution sensing and precise positioning enable more efficient workflows and better safety management. For closure and post-closure, satellite and drone-based monitoring support objective tracking of ground stability and ecosystem recovery, strengthening compliance, transparency, and community confidence.
The volume is grounded in real-world deployment and realistic constraints. It discusses not only what technologies can do, but also their strengths, limitations, and readiness for adoption. The latter includes the skills needed, regulatory integration, and how multi-source data can be translated into reliable decisions. Overall, the book serves as both an accessible introduction and a scientific reference: responsible mining is inseparable from better measurement, and the GoldenEye legacy shows how modern sensing can enable safer, more sustainable, and more transparent mineral production.
How to cite: Joutsenvaara, J., Kotavaara, O., and Paavola, M.: Earth Observations and Proximity Sensing Technologies: Safer, More Sustainable, More Efficient Mining, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19063, https://doi.org/10.5194/egusphere-egu26-19063, 2026.
The transition toward sustainable and resilient supply chains of critical minerals necessitates exploration workflows that are both data-intensive and methodologically transparent. Southeast Asia, which is located at the junction of three main metallogenetic domains, has huge potential for mineral exploration. However, in many ASEAN Member States, mineral exploration remains constrained by heterogeneous data quality, limited interoperability across survey systems, and insufficient integration of multi-scale observation modalities. To address these challenges, the Coordinating Committee for Geoscience Programmes in East and Southeast Asia (CCOP) and the Korea Institute of Geoscience and Mineral Resources (KIGAM) are jointly implementing the ASEAN-Korea Cooperation Fund project (2024-2026), aiming to advance capacity and infrastructure for technology-enabled, database-driven critical mineral exploration.
This contribution presents an integrated framework that couples field-scale acquisition systems with a data platform and a digital-twin-based 3D modeling exploration technology. The proposed workflow assimilates multi-source exploration datasets, including geological mapping, geochemical mapping, geophysical measurements, especially drone-based magnetic surveys, and in-situ terminals, into a unified digital representation of the subsurface. Within this digital twin paradigm, structural elements, geophysical inversion outputs, and associated attribute metadata are harmonized to support iterative model updating, uncertainty reduction, and reproducible interpretation of mineralization processes.
The platform implementation further emphasizes scalable database architecture, secure transmission and governance mechanisms, and interoperable interfaces to facilitate standardized data exchange and analysis. By extending conventional 2D GIS-based repositories toward a 3D exploration database with visualization and model-based analytics, the framework contributes to improved decision support for critical mineral exploration and underpins more robust mineral distribution databases aligned with principles of transparency and materiality commonly required for public reporting.
The CCOP–KIGAM-ASEAN regional collaboration demonstrates how digital-twin-based 3D modeling and integrated exploration data platforms can enhance analytical rigor, operational efficiency, and regional knowledge infrastructure for potential mineral exploration in ASEAN.
Keywords: Critical minerals; Mineral Exploration; Digital Twin; 3D Geological Modeling; Data Platform; ASEAN
How to cite: Wu, S. and Park, G.: Digital-Twin-Based 3D Geological Modeling and Integrated Exploration Data Platforms for Critical Mineral Exploration in ASEAN, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8578, https://doi.org/10.5194/egusphere-egu26-8578, 2026.
The retrieval of drill cores is a costly component of mineral exploration. Improving the spatial overview of mineral abundances within a deposit can substantially reduce the need for drilling. We present an unsupervised, automated annotation strategy for pixel-wise mineral labeling in hyperspectral imagery of simple deposit styles. In this context, a simple deposit style refers to deposits with very low or no mineral transitions and predominantly homogeneous, dominant mineral occurrences.
The automated annotation is based on handcrafted, mineral- and deposit-specific normalized difference indices (NDI). The objective is to extract a large number of representative mineral spectra for each occurring mineral. These spectra are subsequently used as training data for a targeted hyperspectral neural network with positional encoding, which is expected to generalize better to more complex deposit styles.
As a first step, the normalized mineral indices were successfully learned by the network, achieving an F-score of 0.98. This result represents a promising step toward physics-informed, neural-network-based mineral classification in hyperspectral imagery.
How to cite: Kahl, M. and Schodlock, M.: Physics-Informed Annotation for Learning-Based Hyperspectral Mineral Mapping, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-21304, https://doi.org/10.5194/egusphere-egu26-21304, 2026.
Three-dimensional mineral prospectivity mapping (3D MPM) plays a key role in predicting deeply concealed mineral deposits; however, integrating heterogeneous datasets within machine learning frameworks remains a major source of uncertainty. In this study, we develop a gradient boosting ensemble method that explicitly adapts to different data representations and apply it to the Haopinggou gold polymetallic deposit in the western Henan metallogenic belt. Guided by mineral system theory and a 3D geological model, model performance and feature contributions are quantitatively evaluated using the SHAP framework. The results demonstrate that the binary-data-based gradient boosting model achieves higher AUC values and prediction accuracy than alternative approaches, and more effectively delineates deep exploration targets. These findings highlight the practical value of representation-aware ensemble learning for deep mineral exploration and target delineation.
How to cite: Fan, M., Xiao, K., Sun, L., and Xu, Y.: Three-Dimensional Mineral Prospectivity Mapping by a Gradient Boosting-Based Integrated Learning Method with Data Representation Adaptability, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8798, https://doi.org/10.5194/egusphere-egu26-8798, 2026.
With a focus on geo-modeling applications for sustainable deep mineral exploration, we propose affordable, but still accurate subsurface modeling technique that can generates realistic 3-D geological models. Conventional geostatistical methods based on two-point statistics, often compromise their performance in deep and structurally complex geological settings mostly due to limitation in modelling complex spatial continuity patterns. On the other hand, deep generative modelling techniques, such as generative adversarial networks (GAN), allow to predict complex spatial patterns but have difficulties to create large-scales models in three-dimensions and be locally conditioned by observations.
We introduce a deep generative framework that adapts conditional GANs with spatially adaptive normalization (cGAN–SPADE) for 3-D geological modeling under sparse and evolving data conditions to predict high resolution subsurface models with real-time data assimilation capabilities. The goal is to generate geo-models based on a priori geological information (i.e., expected geometries and probability maps) with real-time model update as new data are acquired during drilling.
The cGAN-SAPDE is trained with samples based on prior geological knowledge and existing borehole experimental data. Training proceeds through a generator and discriminator scheme in which generator produces new models based on input training data while the discriminator output is the probability of input image being real based on the corresponding conditioning map.
A conditioning map is introduced at each generator’s layer, where it modulates the intermediate activations using SPADE normalization. This mechanism injects spatially varying conditioning information into the network, enabling the generator to preserve structural coherence and fine-grained spatial details in the synthesized outputs.
Experimental results on industry-standard challenging 3-D synthetic data sets show the ability of the network to predict high-resolution 3-D geological models that simultaneously match a priori information and direct measurements acquired in real-time scenario.
This project has received funding from the European Union’s Horizon Europe Research and Innovation Program under the Grant Agreement No.101178775
How to cite: Akram, N. and Azevedo, L.: AI-driven framework to reconstruct real-time 3-D geological models for In-Situ Exploration of Critical Raw Materials, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7960, https://doi.org/10.5194/egusphere-egu26-7960, 2026.
Introduction: This work deals with the regionalized classification of hydrothermal alteration types from data of continuous features (assays of trace elements and sulfide minerals) in a porphyry copper-gold deposit in Mongolia, using supervised learning algorithms. Traditional machine learning methods ignore the spatial correlations of regionalized data, whereas geostatistics can take advantage of these correlations and enhance classification scores. The novelty of our proposal lies in the deployment of a complementary set of features (‘proxies’) at the sampled data points, calculated ingeniously through geostatistical simulation with nugget effect filtering.
Methodology: We perform the cleaning and preparation of a vast set of exploratory drill hole samples, including the splitting of this dataset into training and testing subsets in the ratio 70:30. The dataset is used for the geostatistical modeling of the feature variables to simulate (by spectral simulation with filtering) the same feature variables at the training and testing data points. Because of the nugget effect filtering, the simulated values ('proxies') do not coincide with the measured (noisy) values and exhibit a stronger spatial continuity. The proxies are then taken as the input for a supervised classification of the hydrothermal alteration type on the training data, which incorporates misclassification cost matrices that account for geological criteria. The performance of the classifier is finally assessed on the testing data on the basis of standard metrics.
Results and Conclusions: Compared to the traditional approach, where hydrothermal alteration types are predicted directly from the measured features, the classification that uses the geostatistical proxies systematically provides better scores (accuracy rate and Cohen’s kappa statistic increased by 5 to 10 percentual points), showing the importance of incorporating proxy variables obtained by a spatial processing of the input information. Another advantage of using geostatistical proxies in the classification is the handling of missing data, insofar as these proxies provide a ‘clever’ alternative to the imputation of missing values, based on the spatial correlation structure of the feature variables and neighboring information, instead of a simple median value by alteration class. The use of geostatistical proxies can therefore be decisive in the presence of highly heterotopic datasets, for which discarding missing data implies a considerable loss of information. In a nutshell, our study demonstrates two things: the first is how geostatistics enriches machine learning to achieve higher predictive performance and to handle incomplete and noisy datasets in a spatial setting. Secondly, it establishes that better prediction accuracy can be achieved than in previous studies, where alteration types were predicted solely from geochemical data.
The proposed approach has far-reaching consequences for decision-making in mining exploration, geological modeling, and geometallurgical planning. We expect it to be used in supervised classification problems that arise in varied disciplines of natural sciences and engineering and involve regionalized data.
How to cite: Borah, A. and Emery, X.: Integration of Machine Learning and Geostatistics for Hydrothermal Alteration Classification in Smart Mining, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2073, https://doi.org/10.5194/egusphere-egu26-2073, 2026.
The PERSEPHONE Project supports the EU’s strategy to access deeper, previously abandoned or otherwise challenging underground mineral deposits in a more sustainable, safe and digitalised manner. In this context, the field deployment reported here at the Koutzi Mine (Evia, Greece) in September 2025 represents one of the demonstration missions of PERSEPHONE, during which a robotic platform performed mapping, relocalisation and multispectral mineral imaging without reliance on external infrastructure.
Robotic exploration of underground environments can serve not only as a means of new discovery, but also as a valuable tool for the remapping of historic galleries and more broadly for subterranean exploration (including caves and other naturally occurring voids). For instance, the UNEXMIN/UNEXUP projects have employed robotic systems to re-survey Europe’s abandoned flooded mines, as well natural flooded cavities like the Molnár János cave (Hungary).
The geological setting of the Koutzi Mine is characterised by a narrow-vein magnesite deposit hosted in ophiolitic ultramafic lithologies on the island of Evia. This historic mine was reopened in 2021 and employs sub-level stoping with battery-operated excavators, reflecting a precision extraction philosophy designed to minimise environmental footprint. However, some of the older, smaller galleries remain unsafe for human exploration. The occurrence of magnesite (MgCO₃), frequently resulting from carbonation of ultramafic rocks, together with accessory white minerals such as sepiolite or opal in fault or alteration zones, provides a good target for multispectral imaging: determining vein type, thickness and mineral differentiation in this environment improves both exploration efficiency and robotics mission planning.
The exploration campaign comprised two phases. In the first phase, a agile mobile robot equipped with LiDAR and IMU sensors operated autonomously within the gallery, constructing a detailed volumetric map of several sections of the mine without use of GPS or pre-deployed reference beacons. Zones of interest were identified using the onboard visible-light camera to locate white-mineral zones. In the second phase, a second robot was introduced, successfully relocalized itself within the map created by the first robot and deployed to capture high-quality multispectral imaging of the identified white-mineral vein zones. The multispectral imaging subsystem comprised a near-infrared (NIR) camera and a UV-fluorescence camera mounted on the robot’s sensor suite. The objective was to acquire precise spectral–spatial data on vein geometries and white-mineral occurrences (distinguishing magnesite, sepiolite and opal) and to characterize thickness and orientation of the mineralized zones. By planning reference viewpoints with high overlap (80 %), the system links multispectral data with the 3D map context and supports subsequent data-driven analytics. Together with autonomous mapping and relocalization in absence of external infrastructure, this experiment provides a proof-of-concept of integrated robotic exploration, targeted mineral sensing and operational autonomy in an underground mining environment.
How to cite: Burlet, C., Stathoulopoulos, N., Viswanathan, V. K., Satpute, S. G., Stasi, G., and Nikolakopoulos, G.: Towards robotic exploration without external infrastructure in underground mining environments: a case study from the PERSEPHONE project, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9939, https://doi.org/10.5194/egusphere-egu26-9939, 2026.
The study area is located in the Eastern Taurus Belt, north of the Keban Reservoir Lake, between the districts of Pertek and Çemişgezek in the province of Tunceli. In the Eastern Taurus Belt, mineralizations are widespread associated with the intrusion of magmatic intrusions into carbonate-rich rocks. The studied zone reflects mineral associations that developed primarily due to iron-bearing minerals associated with skarn formations. The skarn formations developed between the Keban Metamorphics (Permo-Triassic) and the Pertek Granitoid (late Cretaceous) are approximately E-W trending and observed in a narrow line in the region. The most common iron-bearing mineral groups in the area are mainly found as magnetite or ilmenite, as alteration minerals are limonite, hematite ± actinolite. Remote sensing methods were tested to support classical methods in tracking the distribution and traces of these mineralizations. In this context, work was carried out to detect iron-rich zones (FeOx) along the Pertek-Çemişgezek (Tunceli) line. The composite images were used for this region, referencing known iron zones, by the ASTER satellite and image enhancement methods. Accordingly, the main target areas in the southern part of Tunceli province were determined as Köçek Village, Çemişgezek Ferry Terminal in the southwest, the area between Kolankaya and Çataksu in the southeast, and the area bounded by Tozkoparan in the northeast. The image from the ASTER satellite (AST_L1T) was cropped according to the study area, and all work was performed on this dataset. The cropped image set has been limited to fit the workspace. All work was performed using the VNIR and SWIR bands of the ASTER images. Radiometric corrections were made on the relevant dataset, and spectral anomalies were minimized. The VNIR spectral bands, which have a 15-meter ground resolution, were downsampled to a 30-meter ground resolution and balanced with the SWIR spectral bands. By comparing with known ground control points, RGB composite images showing the iron-rich zones in the region were created using different band combinations. As a result, it was determined that VNIR Band 2 / VNIR Band 1, SWIR Band 6, and VNIR Band 3 had the best combinations. In the controls performed, a 94% correlation was tested over the observation points and known iron occurrences. Ultimately, known mineralized zones were found to contain both iron-bearing and iron-rich zones. They were observed primarily Ayazpinari iron (Fe) occurrences, Ballıdut FeOx Alterations, and Çemişgezek Elazığ Road Cut FeOx alterations by both satellite observations and field verification studies.
Note: This study was supported by Fırat University project MF-25.09.
How to cite: Tutlu, R., Ural, M., and Eğri, M.: Detection of Iron-Rich Zones Developed By Skarnification In The Cemisgezek-Pertek (Tunceli) Region Using Remote Sensing Methods, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8118, https://doi.org/10.5194/egusphere-egu26-8118, 2026.
Porphyry Cu deposits host the majority of global Cu resources and high-grade hypogene porphyry Cu deposits are of particular interest to industry because of the reduced waste and energy consumption required in exploitation, leading to favorable economics and reduced environmental impact. Detailed core logging, combined with TESCAN TIMA mineral quantification at two globally significant, high hypogene Cu grade, supergiant porphyry deposits – Resolution, USA and Hugo Dummett North, Mongolia, indicate that the majority of the chalcopyrite±bornite-pyrite are intergrown with muscovite that overprints earlier potassic alteration assemblages containing biotite and/or K-feldspar. Copper grades increase with the intensity of muscovite overprinting on primary potassic assemblages supporting the link between high-grade Cu mineralization and phyllic alteration. Another zone of high-grade Cu mineralization occurs in the upper parts of the phyllic alteration zone and/or within later advanced argillic alteration, associated with high-sulfidation bornite±digenite±covellite±chalcocite-pyrite assemblages, that partly replace earlier chalcopyrite. These two high grade domains have comparable features in many other significant HGHP deposits (Chuquicamata, Rosario, MMH, Onto, Butte) – all strongly telescoped systems that host significant amounts of high-grade Cu mineralization in phyllic and/or advanced argillic alteration that overprint potassic alteration.
We suggest there are at least three reasons for the development of high-grade hypogene ore in telescoped porphyry systems: 1) rapid unroofing and exhumation can generate steep thermal gradients, promoting a rapid decrease in Cu solubility and efficient precipitation of sulfides; 2) the most significant permeability creation in porphyry systems often develops late – during rapid, syn-mineralization exhumation and magma doming stages – when the rock mass behaves in an increasingly brittle fashion; 3) telescoping during syn-mineralization exhumation leads to overprinting of early sulfide assemblages by late-stage acidic and oxidized hydrothermal fluids that remobilize and concentrate early Cu, leading to the precipitation of sulfides with high Cu/S ratios. We conclude that the coincidence of rapid exhumation and long-lived hydrothermal activity exerts a first order control on the formation of high-grade hypogene porphyry Cu mineralization, meanwhile some other factors (such as favorable host rocks, high density of veins and breccias) are potential to form an individual high-grade porphyry Cu deposit.
How to cite: Yang, C. and Wilkinson, J. J.: Formation of giant high-grade hypogene porphyry copper deposits during phyllic to advanced argillic alteration: textural evidence from automated SEM mapping, Resolution and Hugo Dummett North deposits, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-23164, https://doi.org/10.5194/egusphere-egu26-23164, 2026.
The Anti-Atlas, one of the oldest mountain chains in Morocco, has undergone multiple orogenic events that shaped its complex geology, making it a major province of sediment-hosted copper deposits, particularly in its western part. This study adopts a multi-scale, interdisciplinary workflow combining high-resolution hyperspectral remote sensing (up to 5 m spatial resolution), field-based spectral validation, geochemical analyses, and airborne geophysical data to achieve a comprehensive characterization of mineralization processes. Regional mapping of structural lineaments and copper-related alteration zones guided field investigations and the sampling of both mineralized and non-mineralized facies, allowing constraints to be placed on the origin of mineralization. These surface observations were subsequently linked to subsurface architecture through airborne geophysical modelling of regional geological cross-sections derived from field data. The integrated interpretation of all datasets enabled the development of a coherent geodynamic model adapted to the Alma Inlier. Overall, the proposed approach enhances exploration efficiency, reduces uncertainty, and supports more sustainable mineral exploration strategies.
Keywords: Western Anti-Atlas, copper deposits, Hyperspectral remote sensing, Geochemical and geophysical integration, Exploration
How to cite: M'hamdi Alaoui, I., Akhssas, A., Bahi, A., Gautier, S., Ibouh, H., Berkat, N. E., Boumehdi, M., Khebbi, H., and Abouabila, Y.: From Space to Field: Multi-Scale Characterization of Sediment-Hosted Copper Deposits in the Alma Inlier, Western Anti-Atlas, Morocco, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20552, https://doi.org/10.5194/egusphere-egu26-20552, 2026.
Europe is highly dependent on foreign suppliers for several critical raw materials (CRMs), owing to limited domestic mining production. Antimony (Sb) has been included among Europe’s CRMs since the first list published in 2011, due to its extensive use in strategic industrial sectors. To meet the steadily increasing demand, new Sb orebodies must be identified, explored, and exploited within the European Union to diversify supply chains and reduce geopolitical risks. In parallel, the recovery of Sb from secondary sources, such as historical mining wastes, represents an additional opportunity.
Within this framework, Italy has adopted the EU Critical Raw Materials Act, promoting the development of a national exploration plan. Antimony was historically mined in two Italian regions, Tuscany and Sardinia, leaving a substantial legacy of geological data (e.g., mining reports and drill logs) as well as significant volumes of mineral wastes. These Sb districts, where stibnite (Sb₂S₃) is the main economic mineral, represent an exceptional case study for assessing the potential Sb resources and associated CRMs in Italy. This study focuses on the Tuscan Sb district (e.g., the Mancianese area, southern Tuscany), where most of the available geological information is outdated and where robust constraints on orebody geometries, volumes, and associated CRM contents are still lacking (e.g., Lattanzi 1999). Here, we present the first results of an ongoing research project aimed at:
- Geological, mineralogical and geochimical data of Sb resources in Tuscany unravel ore genesis ;
- a 3D geological model of the selected orebodies, and potentially unexploited bodies, with probabilistic functions to conduct uncertainty analysis.
Field surveys and sample collection were carried out in the Mancianese area and were integrated with textural analyses (reflected-light microscopy and SEM), mineral chemistry investigations (EPMA and LA-ICP-MS), stable and radiogenic isotope analyses and fluid inclusion studies. The collected dataset was used to reconstruct a 3D model of selected orebodies using GemPy, an open-source, Python-based geological modeling software. The results highlight the subsurface extent and continuity of mineralization, allowing a first-order estimate of the potentially available Sb resources. The resulting geological model not only contributes to the evaluation of the Italian Sb mining potential, which remains poorly constrained to date (SCRREEN, 2023), but also provides a robust framework for reconstructing the processes responsible for stibnite mineralization. This represents a valuable basis for future exploration and prospection campaigns in Southern Tuscany, offering essential knowledge for characterizing the mineral resource and developing genetic models that can also be applied to similar geological settings across Europe.
How to cite: Galione, M. R., Costagliola, P., Lattanzi, P., Morelli, G., Nannoni, A., Rimondi, V., Ruggieri, G., Trumpy, E., and Vezzoni, S.: The antimony (Sb) resource in southern Tuscany (Italy): A multi-scale approach from textural and geochemical characterization to 3D geological modeling (Montauto mining area) , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-12028, https://doi.org/10.5194/egusphere-egu26-12028, 2026.
Increasing urbanisation and stricter environmental regulations have significantly restricted the exploitation of new gravel quarries as well as the local extraction of natural hard rocks and cement raw materials (lime, marl, clay), posing major challenges for the resource-intensive construction sector. In response, urban mining is gaining importance as a key strategy for circular construction. While natural aggregates from primary quarries provide well-established and consistent quality for concrete production, recycled aggregates (RA) and alternative cement raw materials derived from construction and demolition waste exhibit highly variable performance, strongly governed by source material characteristics and processing routes.
Luxembourg offers a particularly relevant case study due to its pronounced geological diversity and building heritage. The country is divided into the Palaeozoic Eisleck in the north, dominated by schistose rocks affected by Variscan deformation, and the Mesozoic Guttland in the south, characterised by an alternation of sandstones, limestones, dolomites, and marls with limited tectonic overprint. Most of these lithologies were historically used as local building stones, particularly in rubble stone masonry, which was constructed up to the early 20th century. As limestone and marl quarries supplying the cement industry become increasingly depleted or impossible to expand, construction and demolition waste from decommissioned buildings is becoming a significant secondary raw material source.
RA obtained through urban mining originates from highly heterogeneous feedstocks, including demolished concrete, manufactured masonry units, and natural rubble stone masonry. The suitability of rubble stone masonry for structural recycled aggregate concrete (RAC) depends on geological origin, mineralogical composition, the amount and properties of adhering mortar, and potential chemical pre-contamination, particularly by sulphates and chlorides. Porosity and pore-size distribution govern water absorption, workability, and strength development, while mineralogical factors such as alkali–silica reactivity critically affect durability. In addition, the presence of potentially toxic constituents may further limit reuse options.
This contribution presents an integrated geological–engineering approach for the evaluation of locally sourced RA. A material matrix for systematic lithological classification is proposed, linking geological characteristics with processing requirements and concrete performance. Adapted treatment chains - including selective demolition, targeted pre-sorting, and controlled crushing and screening - are identified as essential to ensure consistent RA quality.
Within the regulatory framework of EN 206, EN 206/DNA-LU, and EN 12620, the study demonstrates that properly processed rubble stone masonry can serve as a technically robust and normatively compliant raw material for RAC, supporting sustainable resource management through urban mining.
How to cite: Schäfer, M., Kuhlmann, N., Berna, T., Bender, M., Colbach, R., Thein, J., Schosseler, P., and Maas, S.: Urban Mining in Luxembourg: Integrating Geology and Engineering for Reliable Recycled Aggregate Concrete, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19088, https://doi.org/10.5194/egusphere-egu26-19088, 2026.
Metal scraps pose an economic and ecologically viable source for secondary resource supply to our industries, which call for more independence from global crises and strategic uncertainties. Well advanced technologies exist for steel and aluminum based on mechanical sorting using basic physical properties in order to split the major Fe- and Al rich fractions. However, many high-tech products require a precise composition specified by narrow acceptable ranges of alloy elements to achieve distint performances of a given alloy type. Here, traditional recycling stream processing bears limitations due to the generation of sorting fractions that contain mixes of variable alloy types, both, in steel as well as aluminum sorting products. Metallurgical processing of such mixed alloys, especially mixed aluminum alloys, leads to lower quality metals with less defined performance specifications and hence, the material is then lost for high-tech industries as a secondary resource. A more detailed, quantitative identification of specific alloy elements provides a solution, which allows for the differentiation between and consequent separation of alloy types. Here, laser-induced breakdown spectroscopy (LIBS) has shown enormous potential for trace (alloy) element detection. The remaining challenge or limitation lies in the strong matrix dependence of LIBS. This means, that a well pre-defined and homogeneous material stream is required for the accurate application of LIBS for element quantification and associated alloy identification.
We propose a hierarchical system to adapt LIBS analysis in a flexible way to the requirements of heterogeneous scrap recycling streams. We developed a clustering method to first identify the metal type, steel or aluminum, in mixed recycling products. The identified metal type provides the information on matrix conditions. Using then the respective calibration model for this matrix condition allows estimating precise alloy element concentrations in order to identify the alloy type. In repetition experiments, we could document high accuracies and precisions for specific diagnostic alloy elements, while few others show medium accuracies and precisions. The complementary information of elemental concentrations provides solid ground for an improved alloy detection and strategically points towards further options for dynamic thresholds in scrap processing procedures.
How to cite: Fuchs, M., Singh, A., Patil, R., Barry, M. O., Regulan, G., Madriz Diaz, Y. C., and Gloaguen, R.: Adaptive LIBS analysis for estimating concentrations of alloy elements in heterogeneous metal scrap recycling streams , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16870, https://doi.org/10.5194/egusphere-egu26-16870, 2026.
The Tamdroust copper ore deposit, located within the Bou Azzer–El Graara inlier (Central Anti-Atlas, Morocco), exemplifies Lower Cambrian carbonate–siliciclastic–hosted copper mineralization formed through the combined effects of stratigraphic, structural, and hydrothermal processes. The deposit lies within the Lower Cambrian Igoudine and Amouslek formations of the Tata Group. It is controlled by a major fault system trending N110°–N150°, which served as the main pathway for metalliferous fluids during Hercynian tectonic reactivation. Copper mineralization predominantly occurs in reduced green siltstones and dolostones deposited on a shallow, mixed carbonate–siliciclastic marine platform influenced by episodic terrigenous input. Two main styles of mineralization are recognized: (i) disseminated sulfides, including fine-grained bornite, chalcopyrite, and pyrite dispersed within permeable host rocks; and (ii) vein and veinlet stockworks along interconnected fracture corridors associated with the major fault zone. Textural and petrographic studies reveal a multi-stage paragenetic sequence evolution that comprises: (1) early disseminated and veinlet-type bornite–chalcopyrite–pyrite associated with quartz–calcite; (2) hydrothermal enrichment along faults marked by bornite replacement by chalcocite with digenite and covellite; and (3) supergene weathering producing native copper and secondary carbonates. Stable isotope geochemistry offers crucial insights into the origin and development of mineralizing fluids. Sulfur isotope compositions of bornite (δ³⁴S ≈ +10.2‰) suggest a mixed sulfur reservoir primarily formed by thermochemical sulfate reduction (TSR) of evaporitic sulfates, aligning with the presence of Lower Cambrian evaporite-rich formations. Carbon and oxygen isotope values measured in hydrothermal calcite (δ¹³C = –3.6 to –2.6‰ VPDB; δ¹⁸O = –15.8 to –15.2‰ VPDB, equivalent to +14.7 to +15.3‰ VSMOW) indicate moderate-temperature (~150–160°C) hydrothermal fluids originating from mixed meteoric–basinal brines that have isotopically equilibrated with carbonate–evaporite host rocks. The δ¹³C signatures further point to a dominant marine carbonate source with no significant biogenic carbon contribution, while minor meteoric or atmospheric mixing remains possible. These findings support a model of fluid–rock interaction in a mesothermal hydrothermal setting, where brines, partially modified by evaporites, played a key role in copper transport and sulfide formation. The spatial distribution of ores highlights the significance of redox-controlled mineralization, with the most notable mineral deposits forming at the boundary between oxidized hematite-bearing red beds and reduced green siltstones and carbonates. This redox boundary served as a chemical trap, allowing TSR-driven production of reduced sulfur species and subsequent copper sulfide deposition. In summary, geological, structural, and isotopic evidence indicate that the Tamdroust deposit is a carbonate-hosted copper system of epigenetic stratabound type in Cambrian evaporitic settings, formed during the Hercynian reactivation of Cambrian sedimentary basins. The Tamdroust system exhibits strong similarities with other Cambrian Cu ore deposits in the Anti-Atlas, particularly Jbel N’Zourk and Jbel Laassal, supporting a regional metallogenic model involving fault-controlled brine flow, evaporite involvement, and redox-driven sulfide formation. These findings offer a predictive framework for future copper exploration, focusing on structurally controlled brine pathways and redox boundaries as primary targets across the Central Anti-Atlas.
How to cite: Bouskri, I., Ilmen, S., Souhassou, M., Ikenne, M., Kharis, A.-A., Hibti, M., Gaouzi, A., Zouhair, M., Maacha, L., Maghfouri, S., Jabbour, M., Ouchchen, M., and Ghannami, M.: Sulfur, Carbon, and Oxygen Isotope Constraints on Fluid Sources at the Tamdroust Cu Ore Deposit (Central Anti-Atlas), EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-273, https://doi.org/10.5194/egusphere-egu26-273, 2026.
The Ni-Cu-Co mineralization in the Ringerike Municipality, Norway, is associated with a suite of magmatic intrusions occurring within the Eastern Kongsberg Complex (EKC). The complex formed during the Gothian Orogeny (1.6 - 1.5 Ga), and the most significant local Ni-Cu deposits are predominantly correlated with this magmatism (Orvik et al. 2025). Historically, nickel and copper were produced in this area until operations ceased in 1920 (Mathiesen & NGU, 1977).
In recent years, both industry and academia have shown renewed interest in the region. Current publications have advanced the understanding of the tectonic evolution of the EKC, and its implications for mineral exploration (Orvik et al. 2025). Mansur et al. (2025) discussed the formation and constraints of the most significant past producers, the Ertelien and Langedalen deposits. However, other than several master theses, there has been little to no focus on the other magmatic intrusions hosting the mineralization; the mineralization itself; and the local structural framework and controls on fluid flow. The current license holder, Kuniko Limited, carried out a range of exploration activities and defined a mineral resource estimate (MRE) for the Ni-Cu-Co Ertelien deposit (Kuniko Limited, 2024). The remaining magmatic intrusions received less attention, with large but disparate datasets being produced over the years.
This PhD aims at utilizing the collected data, supported by field and laboratory work, to understand the structural regime across the region and increase the understanding of the controls on mineralization. The integration of the available data will be undergone by application of python-based machine learning to generate mineral prospectivity mapping model. This would allow the identification of exploration targets and the development of hypotheses, which could be then tested by state-of-the-art exploration techniques, significantly enhancing the exploration efforts within the region.
References
Orvik, A. A., Mansur, E. T., Henderson, I., Slagstad, T., Huyskens, M. & Bjerkgård, T., 2025. Isotopic identification of paleo rift zones within the Sveconorwegian Province; implications for nickel sulphide utilisations in the SW Fennoscandian Shield. Precambrian Research 427, 107836.
Mansur, E., Orvik, A. A., Henderson, I., Miranda, A. C., Slagstad, T., Dare, S., Bjerkgard, T., Sandstad, J. S., 2025. Formation of the Ertelien and Langedalen magmatic Ni–Cu sulfide deposits in Norway: investigating the evolution of platinum-group-element-depleted systems at convergent margins. European Journal of Mineralogy 37, 841869
Mathiesen, C. O. & The Geological Survey of Norway, 1977. Vurdering Av Ringerike Nikkelfelter. NGU-RAPPORT, 21.
Kuniko Limited, 2024. ASX Release: Significant Mineral Resource Increase at Ertelien. https://kuniko.eu/asx-announcements/
How to cite: Mroz, R.: Understanding the regional structural framework and controls on Ni-Cu-Co mineralization, in the Ringerike Metallogenic Province, Norway; , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-1191, https://doi.org/10.5194/egusphere-egu26-1191, 2026.
Deep longwall mining in North China-type coalfields is increasingly threatened by water inrush from high-pressure karstic limestone aquifers beneath the coal seam floor. Conventional grouting from underground roadways often has low pressure, short diffusion distances and poor control of hidden faults and collapse columns, so residual water-conducting channels may still trigger serious inflows. This contribution presents an integrated control mode and a quantitative verification framework for deep coal seam floor water hazards. First, a GIS-based multi-criteria assessment of floor failure depth, aquifer pressure and structural complexity is used to delineate high-risk blocks at panel scale. These blocks are treated in advance through coordinated control of water-filled aquifers and water-conducting structures, combining high-capacity directional drilling from the surface with supplementary underground boreholes to grout target limestone aquifers and associated fracture zones ahead of mining. To evaluate the effectiveness of the treatment before face retreat, we establish a sequential verification method that links borehole pressure tests, calculated water-blocking coefficients, repeated mine DC-resistivity surveys, spatial analysis of grouting pressure and volumes, and inspection drilling and inflow monitoring. Application to a >800 m deep longwall panel mining the 11# coal seam shows that inflows from overlying and underlying limestone aquifers were reduced to tens of cubic metres per hour and no floor water inrush occurred during mining. The proposed control–verification scheme provides a transferable engineering model for designing and auditing floor water-hazard management in deep coal mines affected by high-pressure confined aquifers.
How to cite: Hu, Y.: Integrated control and sequential verification of deep coal seam floor water inrush hazards, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-352, https://doi.org/10.5194/egusphere-egu26-352, 2026.
Accurate and efficient rock mass characterization is crucial for achieving sustainable mineral exploration and resource evaluation, especially in the context of increasing global resource scarcity and the urgent need to reduce environmental and operational costs. The Rock Quality Designation (RQD) is a widely used indicator for assessing rock mass integrity in geological and geotechnical engineering. However, conventional RQD determination relies heavily on manual measurements of drill cores, which suffer from low efficiency, poor scalability, and limited integration into data-driven exploration workflows.
To address these limitations, this study proposes an automated approach for RQD computation of drill cores based on computer vision and deep learning. The method integrates image-based sensing with advanced object detection and image segmentation algorithms to achieve non-destructive and automated characterization of drill cores.
First, perspective correction is applied to field-acquired core images to ensure geometric consistency. The principle of perspective correction is to project the two-dimensional original image into a three-dimensional viewing space and then transform the three-dimensional space to the image processing plane. The formulas are as follows:
The 3D viewing space is then mapped to the image processing plane using:
Subsequently, the Segment Anything Model (SAM) is employed to automatically detect and extract core regions based on the similarity of color and texture features. In SAM, the prompt encoder partitions and encodes the image based on object color, texture, and other features using:
On this basis, a YOLOv8-based image segmentation model is constructed to identify gap features between core pieces, enabling precise segmentation of individual core segments. YOLOv8 selects positive samples using the TaskAlignedAssigner strategy, formulated as:
Furthermore, by establishing a mapping between image pixels and physical dimensions, the lengths of core pieces are automatically quantified, enabling RQD computation as follows:
Studies on practical cases indicate that this approach maintains high computational accuracy while significantly improving processing efficiency, highlighting its potential as an AI-driven tool for automated core characterization. This method provides a scalable, non-destructive, and efficient technique for digital and data-driven mineral exploration workflows, supporting more sustainable and scientifically informed decision-making in mineral exploration and resource evaluation.
How to cite: Jiang, J.: Non-destructive, AI-based Rock Core Characterization for Automated RQD Assessment in Mineral Exploration, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-3464, https://doi.org/10.5194/egusphere-egu26-3464, 2026.
3D point clouds of outcrops are digital representations of rock exposures used for geological surveying. These datasets often have high spatial density, up to a thousand points per square meter. By integrating georeferenced data into the 3D point cloud and applying remote sensing interpretation techniques, geoscientists can extract geological features and build 3D models. These models enable the integration of various types of georeferenced datasets, such as compositional, mineralogical, petrographic, structural, multi- and hyperspectral, geophysical, and petrophysical, across 1D, 2D, or 3D formats. However, manual interpretation of 3D point clouds remains labour-intensive, non-reproducible, and prone to human bias. Convolutional neural networks have been applied to segment the images used to build the 3D models, based on a few labelled training and testing subsets, to reduce the amount of human labour. This work used a U-Net encoder-decoder network architecture to segment images of sedimentary facies in reservoir analogue outcrop. The datasets vary in size from 500-1000 images with 40 MP resolution and in number of facies from 2-10. Different data processing pipelines were experimented with, including resizing and slicing due to memory constraints. Approximately 5-10 % of the images in each dataset were labelled by an expert interpreter, with half used for training and half for testing the model, yielding an overall accuracy of 70-85 %. The model was then retrained on the full labelled set and applied to the remaining unlabelled images. The final segmented outputs were processed through a photogrammetry pipeline to generate classified 3D point clouds, capturing the spatial distribution of architectural elements within the outcrop. This workflow allowed a reduction of 90% in manual labour with a high accuracy in the result.
How to cite: Gonçalves, Í., de Souza, E., Guadagnin, F., Roemers-Oliveira, E., Machado, E., Rangel, G., Freccia, A. C., Toledo, J., Schaffer, G., and Scherer, C.: Point cloud segmentation of sedimentary facies in outcrops with convolutional neural networks, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8405, https://doi.org/10.5194/egusphere-egu26-8405, 2026.
Mining activities are essential for the global energy transition, but they remain major drivers of land surface transformation and environmental degradation. Reliable, scalable monitoring of mining-induced land-use change is therefore critical for sustainable resource governance. In our earlier work, MineNetCD (2024) established the first global benchmark for mining change detection, enabling the identification of abrupt mining footprint changes from high-resolution bi-temporal imagery across 100 geographically diverse sites. While this provided a robust foundation for static change detection, sustainable mining oversight requires tracking the continuous and often gradual evolution of mining activities over time.
To address this limitation, we introduce EuroMineNet (2025), the first comprehensive multi-temporal mining benchmark designed for dynamic monitoring across the European Union. Leveraging a decade of Sentinel-2 multispectral imagery (2015–2024), EuroMineNet provides annual observations for 133 mining sites, enabling systematic analysis of both short-term operational dynamics and long-term land-use transformations.
The dataset supports two complementary, sustainability-oriented tasks: (1) Multi-temporal mining footprint mapping, producing temporally consistent annual delineations; and (2) Cross-temporal change detection, capturing gradual expansion, reclamation, and episodic disturbances.
To assess temporal consistency under evolving conditions, we propose a novel Change-Aware Temporal IoU (CA-TIoU) metric. Benchmarking 20 state-of-the-art deep learning models reveals that while current GeoAI methods perform well for long-term changes, they struggle with short-term dynamics crucial for early warning and mitigation. By advancing from global static detection to regional continuous monitoring, this work directly supports the European Green Deal and contributes to the development of transparent and explainable GeoAI tools for environmental resilience.
How to cite: Yu, W., Nwazelibe, V., Zhang, X., Zhu, X., Gloaguen, R., and Ghamisi, P.: EuroMineNet: Continuous Multitemporal Monitoring of Mining Dynamics in the European Union, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13973, https://doi.org/10.5194/egusphere-egu26-13973, 2026.
Tailings Storage Facilities (TSFs) represent one of the most critical and high-risk infrastructures in the mining sector, with failures leading to severe environmental, social, and economic consequences at local and transboundary scales. Increasing climate variability, ageing facilities, rising demand for mined products, and rising regulatory expectations necessitate more advanced TSF monitoring approaches. Existing TSF monitoring is often fragmented, as Earth observation, in-situ sensing, and risk assessment tools operate independently, limiting their effectiveness for continuous risk assessment. This underscores the need for integrated, multi-sensor monitoring approaches that can provide continuous, comprehensive, and predictive assessment of TSF stability and associated risks.
The GAIA-TSF (Geospatial Artificial Intelligence Analysis for Tailings Storage Facilities) project, led by an international consortium, aims to design and develop a prototype system. This system integrates satellite Earth Observation (EO) and ground-based sensor data with machine-learning (ML) algorithms to enable continuous, multi-level, and multi-scale characterization and monitoring of TSFs.
As a work in progress, the project has undertaken a comprehensive stakeholder engagement process to identify current gaps, operational needs, and priority monitoring requirements for TSFs. A review of the state of the art in available EO and ground-based monitoring technologies has been conducted, leading to the identification of key technologies and ML techniques. An extensive review of the literature, coupled with stakeholder input, led to the identification of key variables relevant to TSF monitoring. Such parameters include water quality, air quality, and slope stability. In parallel, potential test sites across different continents have been selected to support future calibration and validation of the prototype under diverse geographical and climatic conditions. The functional requirements and system architecture have been defined, identifying the key components of the prototype and how they are connected. The initial development phase of the GAIA-TSF prototype has commenced.
Integrated TSF monitoring supports risk-informed life-cycle management of TSF, enabling loss prevention and effective asset stewardship. It also strengthens decision-making for ESG compliance, the Global Industry Standard on Tailings Management (GISTM), and climate adaptation, ensuring safer and more sustainable mining operations.
The GAIA-TSF prototype offers a transferable and scalable continuous monitoring solution that enhances early anomaly detection and supports risk-informed decision-making. It thereby contributes to more sustainable and resilient TSF management.
How to cite: Desta, F., Růžička, J., Bouvier, R., Andreani, L., Brodský, L., Landa, M., Bouček, T., Buxton, M., Nwaila, G., Mahboob, M., Shimaponda, M., Chabala, M., Makondo, C. C., Quijano, L., and Diego Lozano, D.: Geospatial AI for Continuous Multi-Scale Risk Monitoring of Tailings Storage Facilities, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10886, https://doi.org/10.5194/egusphere-egu26-10886, 2026.
Surface wave signals between station pairs can be obtained by cross-correlating long-term continuous ambient noise recordings, from which group- and phase-velocity dispersion measurements at different periods are obtained and subsequently inverted for 3-D shear-wave velocity structures from shallow crust to upper mantle. This method does not rely on artificial seismic sources such as explosives, features relatively low exploration costs, and is well suited to complex topographic and environmental conditions. In recent years, it has been widely applied to image 3-D isotropic shear-wave velocity structures of mineral districts at different spatial scales (Hollis et al., 2018; Zheng et al., 2022; Jing et al., 2025). However, due to limitations in imaging resolution and the relatively small density contrast between ore-related rock bodies and surrounding host rocks, isotropic velocity structures alone are often insufficient for the effective identification and detailed characterization of ore-related rock bodies.
To address these limitations, we employed a direct surface wave tomography framework (Fang et al.,2015; Liu et al., 2019) to a selected mineral district using dense array ambient noise data. We first resolved the 3-D isotropic shear-wave velocity structure and subsequently retrieved the azimuthally anisotropic velocity structure in the very shallow crust. The results demonstrate that the isotropic velocity structure clearly delineates the major ore-controlling faults and structural framework of the mineral district, providing insights into its ore-forming tectonic regime. Besides, the azimuthally anisotropic shear-wave velocity structure shows strong spatial consistency with the distribution of known ore-related rock bodies and effectively highlights potential favorable mineralization targets. Overall, our study suggests that the combined interpretation of 3-D isotropic and azimuthally anisotropic velocity structures derived from ambient noise surface wave tomography provides an effective geophysical tool for mineral exploration and evaluation at both shallow and deep levels in mineral districts.
Reference
[1] Hollis D, McBride J, Good D, et al. 2018. Use of ambient-noise surface-wave tomography in mineral resource exploration and evaluation. SEG Technical Program Expanded Abstracts: 1937-1940.
[2] Zheng F, Xu T, Ai Y S, et al. 2022. Metallogenic potential of the Wulong goldfield, Liaodong Peninsula, China revealed by high-resolution ambient noise tomography. Ore Geology Reviews, 142: 104704.
[3] Jing J L, Chen G X, Li P, et al. 2025. Ambient noise seismic tomography of Tonglushan skarn-type Cu-Fe-Au deposit in Eastern China. Ore Geology Reviews, 184: 106718.
[4] Fang H J, Yao H J, Zhang H J, et al. 2015. Direct inversion of surface wave dispersion for three-dimensional shallow crustal structure based on ray tracing: methodology and application. Geophysical Journal International, 201(3): 1251-1263.
[5] Liu C M, Yao H J, Yang H Y, et al. 2019. Direct inversion for three-dimensional shear wave speed azimuthal anisotropy based on surface wave ray tracing: Methodology and application to Yunnan, southwest China. Journal of Geophysical Research: Solid Earth, 124(11): 11394-11413.
How to cite: Fang, J., Li, X., Yao, H., and Luo, X.: Azimuthal Anisotropy of Ambient Noise Rayleigh Waves Revealing Ore-Controlling Structures and Ore-Related Rock Bodies in a Mineral District, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7639, https://doi.org/10.5194/egusphere-egu26-7639, 2026.
Subsea sediment sampling is of great significance for marine geological research, resource exploration, environmental assessment, and geotechnical investigation. However, due to the common characteristics of high clay content, high water content, and under-consolidation of seabed sediments, conventional sampling techniques often cause severe sample disturbance, compression, or even loss. This leads to engineering challenges such as low core recovery and destruction of the original structure, which significantly compromises the in-situ characteristics and representativeness of the samples.Inspired by organisms (such as lotus leaves and earthworm) that maintain clean body surfaces in viscous environments, this study developed a material-structure coupled bionic anti-adhesion and drag-reduction surface by mimicking their micro-nano structure and low interfacial energy characteristics. This surface was constructed using a specific etching process combined with a low interfacial energy material coating technique and applied to the key contact parts of a subsea sediment sampling drill tool. Microstructural characterization and comparative sampling tests in typical clay and silty clay demonstrated that the bionic drill tool significantly reduces soil adhesion and frictional resistance during the sampling process. Consequently, it substantially increases the core recovery rate and effectively preserves the original stratigraphic sequence and moisture condition of the samples, markedly enhancing their in-situ fidelity.The bionic self-cleaning surface technology proposed in this study offers an innovative solution to the technical bottleneck of low-disturbance, high-fidelity sampling of highly viscous subsea sediments. Preliminary tests have verified the chemical stability and corrosion resistance of the surface coating in simulated seawater environments. Its long-term service reliability and large-scale engineering application processes require further research and optimization.
How to cite: Zhang, P., Guo, W., Wang, Y., and Jia, R.: Development of a Bionic Self-Cleaning Drill Tool toward Enhanced In-Situ Fidelity in Subsea Sediment Sampling, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8770, https://doi.org/10.5194/egusphere-egu26-8770, 2026.
