In recent years, passive seismic imaging has emerged as a cost-effective exploration tool, particularly valuable in complex geological settings and at depths beyond the reach of conventional active methods. These approaches are increasingly demonstrating their potential not only for geothermal and subsurface storage applications, but also for mining exploration.
This session invites contributions that advance passive seismic methodology and modeling for applications to subsurface imaging, as well as case studies showcasing applications to critical raw material exploration, mining, geothermal energy, hydrogen storage, and CCS. We particularly encourage studies highlighting integration of passive seismic techniques into industrial exploration workflows, and contributions spanning ambient-noise and/or earthquake-based approaches.
In the pursuit of deep geothermal energy (depths exceeding 3 km), the limitations of traditional surface exploration often render the subsurface "invisible." This study presents an integrated seismic exploration framework: the GEOthermal SEISmic AI Platform (GEOSEIS-AI). This platform leverages high-density microseismic monitoring networks and advanced deep-learning (DL) techniques to resolve the "four pillars" essential for geothermal development: heat sources (temperature), stress states (pressure), fracture distributions (pathways), and fluid properties. Building upon the architecture of the Real-Time Microearthquake Monitoring System (RT-MEMS) (Sun et al., 2025), GEOSEIS-AI utilizes DL phase picking and earthquake localization to accelerate the processing of massive datasets. Key seismic observables—including seismicity, focal mechanism, shear-wave splitting, and seismic tomography—are employed to directly characterize these four parameters. We demonstrate the platform's capabilities through two distinct case studies: a metamorphic region in Taiwan focusing on deep geothermal potential (Huang et al., 2023) and a volcanic region in Japan targeting supercritical energy (Tsuji et al., 2025). By mapping the spatial distribution of microearthquakes, we identify the Brittle-Ductile Transition (BDT) interface. Since seismic activity ceases as rocks transition from brittle to plastic states at high temperatures (350-400°C), the "seismic-quiet zone" serves as a proxy for the top of the heat source. Identifying these thermal upwellings is essential for targeting high-enthalpy drilling sites. By analyzing P-wave first motions with DL techniques, we resolve the local stress field and faulting styles. This provides vital data for assessing wellbore stability and distinguishing between dilated, fluid-conductive faults and compressed, sealing structures. Utilizing shear-wave splitting technique, we quantify the density and orientation of subsurface fracture networks. This provides a "pre-drilling ultrasound" that identifies high-permeability zones and informs hydraulic fracturing strategies for Enhanced Geothermal Systems (EGS). Through Vp/Vs ratio analysis derived from seismic tomography, we can differentiate between solid lithology and fluid-filled pores, and more critically, the identification of fluid phases (liquid water, steam, or melt), where low and high Vp/Vs ratios act as indicators of geothermal steam and fluids, respectively. The results show that GEOSEIS-AI significantly enhances the resolution of reservoir imaging and also provides critical insights into induced seismicity monitoring for future geothermal hydrofracturing and CO2 injection of CCS operation.
Keywords: GEOSEIS-AI; Deep Geothermal Energy; Supercritical Energy; CCS; Deep Learning; Microseismic Monitoring; Seismicity; Focal Mechanism; Shear-Wave Splitting; Vp/Vs; Seismic Tomography.
References:
Sun, W.-F., S.-Y. Pan, Y.-H. Liu, H. Kuo-Chen, C.-S. Ku, C.-M. Lin, and C.-C.Fu (2025). A Deep-Learning-Based Real-Time Microearthquake Monitoring System (RT-MEMS) for Taiwan. Sensors, 25(11), 3353. https://doi.org/10.3390/s25113353.
Tsuji ,T., R. Andajani, M. Kato, A. Hara, N. Aoki, S. Abe, H. Kuo-Chen, Z.-K. Guan, W.-F. Sun, S.-Y. Pan, Y.-H. Liu, K. Kitamura, J. Nishijima, and H. Inagaki (2025) Supercritical fluid flow through permeable window and phase transitions at volcanic brittle–ductile transition zone, Commun. Earth Environ. https://doi.org/10.1038/s43247-025-02774-4.
Huang S.-Y., W.-S. Chen, L.-H. Lin, H. Kuo-Chen, C.-W. Lin, W.-H. Hsu, Y.-H. Liou (2023). Geothermal characteristics of the Paolai Hot Spring area, Taiwan. 45th New Zealand Geothermal Workshop, Auckland, New Zealand.
How to cite: Kuo-Chen, H., Sun, W.-F., Guan, Z.-K., Pan, S.-Y., Chang, C.-J., Liu, Y.-H., and Tsuji, T.: GEOthermal SEISmic AI Platform (GEOSEIS-AI) for Deep and Supercritical Geothermal Exploration, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-15245, https://doi.org/10.5194/egusphere-egu26-15245, 2026.
Geothermal systems provide a low-carbon, renewable source of heat, whose performance depends on the presence of permeable, fluid-filled rock at depth. Tomographic images of compressional and shear-wave velocities, Vp and Vs, and their ratio, Vp/Vs, are typically used to constrain the lithology, porosity, fluid content and extent of fracturing in such systems: contrasts in seismic velocity delineate lithological boundaries, identify zones of fracture damage or fluid saturation, and thereby indicate areas of elevated permeability.
Passive seismic acquisition is attractive for geothermal exploration, because it is minimally invasive and can exploit microseismicity recorded by dense nodal seismological arrays. Combining data recorded from microseismic events with Bayesian joint inversion of seismic velocity and source location – here implemented with stochastic Stein Variational Gradient Descent (sSVGD) and double difference tomography – yields relocated earthquake events and three-dimensional estimates of Vp, Vs, and Vp/Vs together with their respective uncertainty. sSVGD approximates the statistical description of all possible models that fit the data, referred to as the posterior distribution, using an ensemble of particles or samples. These are initialized from a prior distribution, which encodes the prior information about the domain, and driven toward the posterior by iterative transforms that minimise the Kullback-Leibler divergence between the particle density and the posterior.
We apply this workflow to the Eden Project geothermal site (Cornwall, UK), using microseismic events recorded by an array of 450 STRYDE nodes deployed around the injection site. The objective is to recover mean models of Vp and Vs, and Vp/Vs with their corresponding uncertainty from passive sources alone, enabling probabilistic assessment of the subsurface structure and potential future well-placement targets.
Owing to the nodal geometry and the spatial distribution of microseismic sources, ray-path coverage is highly heterogeneous across the survey volume. Consequently, the posterior uncertainty is large over much of the domain and decreases substantially where ray coverage is dense – mostly around the geothermal well. Within this region, we observe velocity anomalies consistent with fractured and fluid-saturated rock, while regions distant from the well remain poorly constrained. By providing a clearer understanding of uncertainties inherent to tomographic inversions, the probabilistic imaging framework enables more robust and reliable analysis of the results, which is crucial in geothermal exploration.
How to cite: Dromigny, S., Yang, H., Koelemeijer, P., Curtis, A., Hudson, T., Kendall, M., and Zhang, X.: Probabilistic body wave tomography in a geothermal setting in Cornwall, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6450, https://doi.org/10.5194/egusphere-egu26-6450, 2026.
Imaging the Earth’s subsurface is fundamental to a wide range of geophysical applications, including natural hazard assessment and mitigation, geothermal and mineral exploration, and crustal characterization. However, achieving reliable seismic images in strongly heterogeneous media remains a significant challenge. In such environments, conventional seismic imaging approaches, including tomography and migration, often perform poorly due to the prevalence of multiple scattering and high attenuation, which obscures primary reflections and degrades image quality.
While multiple scattering has traditionally been regarded as a major impediment to seismic imaging, recent advances have demonstrated that this scattered energy can instead be exploited to extract valuable information. One such approach is Reflection Matrix Imaging (RMI). RMI involves using seismic interferometry to construct a reflection matrix that contains the full wavefield response between virtual source–receiver pairs, allowing for the analysis of reflected energy generated by subsurface heterogeneities. From this, the distortions undergone by the incident and reflected waves can be isolated and compensated for even with a rough estimate of the background seismic velocity. RMI has been shown to enhance imaging in complex geological settings, including volcanic environments, and has also been seen to be effective in 3D imaging applications in fields such as optical microscopy and medical ultrasound.
In this study, RMI is adapted to data from a dense seismic array deployed in the Hengill Geothermal Field, Iceland. The subset of the array considered here comprises 267 stations distributed over a rectangular approximately 5X10km2 , with continuous recordings spanning 2.5 months. Reflection matrices are constructed, and the applicability and performance of RMI in this highly heterogeneous geothermal setting are systematically evaluated.
How to cite: Berry-Walshe, L., Korta Martiartu, N., and Obermann, A.: Reflection Matrix Imaging of the Hengill Geothermal Field, Iceland, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13211, https://doi.org/10.5194/egusphere-egu26-13211, 2026.
Mapping of fracture networks is critical to the exploration and responsible exploitation of geothermal resources. Fractures provide the permeable pathways required for efficient heat extraction and knowledge of their subsurface distribution is necessary for optimal well placement and reservoir modelling. Additionally, fractures play a significant role in induced seismic hazards both by decreasing rock strength and by providing hydraulic connections between fluid injection/extraction sites and surrounding fault networks that may slip in response to pore pressure perturbations. However, constraining fracture distributions in 3D can be challenging. Geologic mapping provides limited information regarding how these systems evolve with depth and exploratory drilling is expensive and only provides point-wise constraints that may not reflect larger-scale trends. Seismic imaging utilising local earthquakes provides a cost-effective means to overcome these issues and map fractures at the reservoir scale. In this contribution, we constrain the anisotropic P-wave velocity structure of the Hengill Geothermal Field (Iceland) using arrival times from natural and induced seismicity. A Bayesian Monte Carlo sampling approach is used to construct likely velocity models and posterior parameter distributions from which we evaluate hypotheses for fracture properties. The imaged slow P-wave propagation directions constrain the average 3D fracture plane orientations while the degree of alignment and extent of fracturing is inferred from the strength of velocity anisotropy. Our models reveal significant spatial heterogeneity in these fracture properties throughout the Hengill geothermal system. We explore possible mechanisms behind this heterogeneity (e.g. deformation related to topographic loading, tectonic and magmatic stresses, and geothermal energy production) and its relationship to local seismicity patterns.
How to cite: VanderBeek, B., Nowacki, A., and de Ridder, S.: Exploring fracture networks beneath the Hengill Geothermal Field (Iceland) through probabilistic anisotropic P-wave tomography, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13292, https://doi.org/10.5194/egusphere-egu26-13292, 2026.
Understanding the structure of the crust and subsurface fluid distribution in volcanic systems is critical for geothermal energy development, volcanic hazard monitoring and mineral exploration. Seismic travel time tomography provides high-resolution images of the subsurface by mapping variations in P- and S-wave velocity structures and their ratio (Vp/Vs), offering insights into the internal structure of the volcano. In this study, we apply local earthquake travel time tomography at Aluto volcano, located in the central Main Ethiopian Rift (MER), and Ethiopia's first pilot site for geothermal energy development.
We analyse seismic data recorded between January 2012 and January 2014, identifying 2,393 local earthquakes mainly along the central part of the caldera and the Wonji Fault Belt (WFB) using non-linear location methods. We selected events with low spatial errors and a signal-to-noise ratio threshold of three or higher for the 3D travel time tomography. By resolving P- and S-wave velocity variations, as well as Vp/Vs anomalies, we aim to delineate zones of fluid saturation and structural heterogeneity. We compute the complete model resolution matrix using direct sparse methods, enabling us to assess the reliability of the tomographic model.
The results of this study are compared with previous studies on the attenuation and conductivity structure of Aluto, collectively providing new insights into the magmatic-hydrothermal system of the Aluto volcano. This study will help to refine geothermal exploration strategies and enhance our understanding of subsurface processes beneath the volcano.
How to cite: Yemane, T., Kendall, J. M., and Bogiatzis, P.: Travel time tomography of Aluto-Langano Geothermal field in the Main Ethiopian Rift, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-23038, https://doi.org/10.5194/egusphere-egu26-23038, 2026.
A sustainable supply of critical and strategic raw materials, like copper, cobalt, lithium, or fluorite, amongst others, is a critical prerequisite for the decarbonisation of the economy and the successful implementation of the green transition. However, Europe currently lacks sufficient knowledge, exploration activity, and domestic supply of these commodities. To overcome these limitations, the DEXPLORE project aims at developing surface to subsurface sustainable cost-effective geological and geophysical techniques for mineral exploration using three pilot zones in Spain and Estonia. Within the framework of this project, we present an innovative non-invasive passive seismic exploration approach and a field test conducted to optimize acquisition and processing parameters. The main objective was to achieve sufficient resolution at prospect depths of 500-1000 m, enabling the identification of ore-associated geological structures in one of the pilot zones, corresponding to the Villabona Fluorite deposit in Asturias (northern Iberian Peninsula, Spain).
The passive seismic methodology relies on the recording and processing of ambient seismic noise acquired by seismic nodes. We designed a preliminary configuration and workflow based on an extensive review of the passive seismic method to run a five-day small-scale field test in the Minersa Pilot Zone, located in the central part of Asturias (N Spain). In this area, the currently active Villabona Mine produces fluorite hosted by Mesozoic sediments affected by extensional faults on an epigenetic Mississippi-Valley-type ore deposit. The fieldwork encompassed the deployment of 38 seismic nodes along a profile with a total length of 3300 meters, with a sensor spacing of 90 meters. Five days of continuous passive data were acquired. Processing methods included the Extended Spatial Autocorrelation (ESPAC) methodology and the Ambient Noise Interferometry (ANI) procedure. The inversion of 31 dispersion curves enabled the construction of a 2-D S-wave velocity model extending to a maximum depth of 700 m. The model shows two velocity sectors separated by a low velocity corridor and identifies velocity anomalies that correspond with structural variations and major fault systems. These results validate the proposed ambient seismic noise workflow for imaging geological and structural features to depths of approximately 700 meters. Additionally, this study demonstrates that the ESPAC processing method enhances survey efficiency and flexibility, particularly when using irregular array configurations. The ESPAC method provided the most reliable results for developing an S-wave velocity model, with lateral resolution dependent on the number and spacing of seismic nodes. Future works include the development of additional passive seismic profiles in the Villabona Pilot Zone, together with planned tests in two additional pilot areas in Spain and Estonia. The main aim is to further validate and apply the passive seismic methodology across diverse geological settings characterized by variable ore deposit distribution and structural configurations.
How to cite: Cadenas, P., Olona, J., Acevedo, J., Fernández Martín, E., Mazaira, A., Gallardo, P., Villa, L., Cueto, M., Marín, J. A., Agüera, M., Rodríguez, R., and Olivar-Castaño, A.: Optimization of a land-based non-invasive passive seismic approach for the exploration of deep-seated critical raw materials in the North of Spain, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14029, https://doi.org/10.5194/egusphere-egu26-14029, 2026.
Deep mineral exploration in high-altitude permafrost regions, such as the Qinghai-Tibet Plateau, faces severe challenges due to complex topography, fragile ecosystems, and intense industrial noise. While passive seismic reflection imaging offers an eco-friendly alternative to active sources, its reliability is often compromised in active mining areas where the ambient noise field is strongly directional and non-stationary, violating the stationary phase assumption required for interferometry.
In this study, we present a successful application of passive seismic reflection imaging at the Huoshaoyun super-large lead-zinc deposit (>5,000 m elevation) in Xinjiang, China. To overcome the artifacts induced by strong directional noise (e.g., heavy mining trucks and machinery), we propose a novel wavefield reconstruction method based on Directional Energy Balancing in the frequency-wavenumber (f−k) domain. Unlike traditional linear stacking, our approach introduces a Directionality Index (DI) to quantify the energy asymmetry of noise slices. We implement a "bucket balancing" weighting strategy that actively screens and balances the noise energy flux, constructing a virtual isotropic illumination environment. This process effectively suppresses spurious artifacts and significantly enhances the signal-to-noise ratio of body-wave reflections.
Utilizing 31 days of continuous waveform data from a dense linear array of 500 short-period seismometers, we retrieved high-resolution reflection profiles reaching 2 km depth. The imaging results clearly reveal the spatial geometry of ore-controlling syncline structures and interlayer fracture zones. These geophysical interpretations were validated by subsequent drilling, demonstrating a high consistency with geological facts. Our findings indicate that the proposed directional balancing strategy can turn "noise into signal" even in strongly heterogeneous noise environments, providing a robust, low-cost, and non-invasive solution for deep resource exploration in extreme environments.
How to cite: Jin, Z. and Wang, Z.: Passive Seismic Reflection Imaging in Active Mining Environments: A Directional Energy Balancing Strategy Applied to the Huoshaoyun Deposit, Tibet Plateau, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16335, https://doi.org/10.5194/egusphere-egu26-16335, 2026.
Deep-derived carbon dioxide (CO₂) degassing is a globally important process linking crust–mantle fluid transport with atmospheric carbon budgets. Matched Field Processing-Bartlett Beamformer (MFP-BB) method offers a seismic approach for detecting tremor signals generated by these degassing centers (mofette). Its principle relies on comparing recorded wavefields with modeled replicas to identify the most likely source locations.
This study applies the MFP–BB technique to dense-array seismic noise data from three key mofette areas in the Cheb Basin, western Eger Rift—Bublák, Hartoušov, and Soos. We combine field observations with numerical simulations to evaluate the method’s performance. Synthetic tests with interfering noise-embedded sources (SNR = 5 dB) demonstrate that accurate localization is achievable with appropriate frequency selection, and that even 20% perturbations in the velocity model introduce only minor degradation.
Field data were processed through segmentation, noise filtering, and spectral analysis to determine persistent frequency bands used in the algorithm. Across all sites, MFP-BB energy concentrates near the surface, coinciding with known mofette fields and CO₂ discharge zones. These shallow anomalies reflect microtremors generated as ascending CO₂ interacts with groundwater and unconsolidated sediments; additional, weaker anomalies at depths < 200 m may also represent active gas migration.
How to cite: Karimi, K., Fischer, T., Vlček, J., Mazanec, M., Vilhelm, J., and Masihi, A.: Matched-Field Processing for Detecting Mofette Activity in the Western Eger Rift, Czechia, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10793, https://doi.org/10.5194/egusphere-egu26-10793, 2026.
Establishing a real-time and high-resolution earthquake catalog is crucial for understanding the development process of earthquake sequences and conducting disaster risk assessment. This study developed a real-time microearthquake monitoring system (RT-MEMS) that integrates deep learning technology (Sun et al., 2025). After testing and verification, it was confirmed that the system can quickly and reliably provide earthquake activity information through a fully automated process. The main data processing process of the system includes: (1) using SeedLink to receive continuous waveform data from four broadband seismic networks, maintained by the Institute of Earth Sciences of Academia Sinica, the National Center for Research on Earthquake Engineering, National Chung Cheng University, and National Taiwan University, and store and build a continuous waveform database; (2) using a deep learning model trained with Taiwan earthquake arrival data to identify and select P- and S-wave arrival times and store them in an arrival database; (3) selecting appropriate seismic station combinations according to the monitoring area, extracting corresponding P- and S-wave arrival times to associate and locate earthquake events, and generating a preliminary deep learning earthquake catalog; (4) preparing daily earthquake reports and sending them to relevant personnel via email, LINE, Discord etc. Compared with the existing seismic observation network, this system has shown advantages in microseismic detection and analysis capabilities and processing efficiency. It is particularly suitable for specific areas or fields that require intensive monitoring. Currently, three real-time microseismic monitoring systems have been established: 1. Chihshang real-time microearthquake monitoring system (2022CSN-RT-MEMS), which observes the background microseismic activity of the creeping segment of the Chihshang fault, including the 2022 M6.9 Chihshang earthquake sequence (Sun et al., 2024); 2. Hualien earthquake real-time microseismic monitoring system (2024HL-RT-MEMS), which continuously observes the changes in the aftershock sequence of the 2024 M7.2 Hualien earthquake; 3. the Chia-Nan real-time microseismic monitoring system (2025CN-RT-MEMS), that this system was established in early 2025 to observe the main aftershock sequence of medium and large earthquakes in the area including the 2025 M6.4 Dapu earthquake sequence (Kuo-Chen et al., 2025). RE-MEMS can quickly provide changes in seismic activity and establish a long-term earthquake catalog. After further data processing (such as absolute or relative relocation), the earthquake catalog will help the subsequent interpretation of earthquake tectonic structures and other earthquake parameter studies, such as focal mechanism, earthquake magnitude, and three-dimensional velocity model inversion. In summary, RT-MEMS serves as an effective reinforcement for the current earthquake observation network, significantly improving the timeliness and resolution of earthquake observation.
Keywords: real-time microearthquake monitoring system; deep learning; SeedLink; automated workflow; earthquake catalog
References
Kuo-Chen H., et al. (2025). Real-time earthquake monitoring with deep learning: A case study of the 2025 M6.4 Dapu earthquake and its fault system in southwestern Taiwan. The Seismic Record, 5(3), 320-329, https://doi.org/10.1785/0320250023.
Sun, W. F., et al. (2024). Deep learning-based earthquake catalog reveals the seismogenic structures of the 2022 MW 6.9 Chihshang earthquake sequence. Terr. Atmos. Ocean. Sci., 35, 5, https://doi.org/10.1007/s44195-024-00063-9.
Sun, W. F., et al. (2025). A Deep-Learning-Based Real-Time Microearthquake Monitoring System (RT-MEMS) for Taiwan. Sensors, 25(11), 3353. https://doi.org/10.3390/s25113353.
How to cite: Sun, W.-F., Pan, S.-Y., Liu, Y.-H., Kuo-Chen, H., Ku, C.-S., Lin, C.-M., Fu, C.-C., Wen, S., and Kuo, Y.-T.: GEOthermal SEISmic AI Platform (GEOSEIS-AI): A Deep-Learning-Based Real-Time Microearthquake Monitoring System (RT-MEMS) for Taiwan, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16026, https://doi.org/10.5194/egusphere-egu26-16026, 2026.
Shear-wave splitting analysis module is part of the GEOSEIS-AI platform, primarily utilized to characterize stress states and subsurface fracture distributions in geothermal sites. However, microseismic data in geothermal sites often face on inherent limitations, including low signal-to-noise ratios (SNR), cycle skipping, fast/slow wave misidentification, and null measurements, all of which compromise the accuracy of automated processing.
To solve these limitations, this study optimizes the pre-processing stage by utilizing adaptive time-window selection to maximize SNR. Furthermore, an automated quality-controlling workflow was developed, based on three diagnostic metrics: (1) peak-picking determination of fast and slow waves; (2) cross-correlation (CC) coefficients; and (3) the energy variation rate between the principal S-wave component and perpendicular component. These tests facilitate the robust identification and remove low-quality seismic events.
This methodology was validated using microseismic monitoring data from the geothermal site in Miaoli, Taiwan. The results reveal two predominant fracture sets oriented NW-SE and N-S. The NW-SE orientations align with the regional focal mechanism solutions, reflecting stress states, while the N-S trends correspond to surface-mapped fault orientations. This workflow was integrated into the GEOSEIS-AI Platform—alongside AI catalogs, focal mechanisms, and seismic tomography—to establish a reliable microseismic monitoring system for geothermal exploration.
Keywords: GEOSEIS-AI; Geothermal Energy; Microseismic Monitoring; Shear-Wave Splitting; fracture distribution.
How to cite: Chang, C.-J., Sun, W.-F., Liu, Y.-H., Pan, S.-Y., and Kuo-Chen, H.: GEOthermal SEISmic AI Platform (GEOSEIS-AI): Shear-wave Splitting Analysis Module and A Case Study of Geothermal Site in Miaoli, Taiwan, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-15833, https://doi.org/10.5194/egusphere-egu26-15833, 2026.
Geothermal exploration in tectonically active regions requires reliable imaging of subsurface structures, fracture systems, and potential heat sources. Seismic methods play a critical role in providing key constraints on buried fault geometry and geothermal-related structures.
This study applies an AI-assisted seismic workflow to seismic tomography for evaluating geothermal potential in the Western Foothills of Taiwan.Earthquake catalogs generated using AI-based detection and phase-picking algorithms were used as inputs for finite-difference travel-time tomography to construct three-dimensional P- and S-wave velocity models from the surface to 8 km depth, with an approximate spatial resolution of 1 km in the upper 6 km.
Two geothermal areas were investigated: the Tai’an area in central Taiwan and the Baolai area in southwestern Taiwan, both characterized by prominent hot spring outcroppings. A total of 63 and 49 seismic stations, respectively, recorded one month of continuous data in each area. The tomography results reveal shallow seismicity mainly distributed between 3 and 7 km depth, closely associated with mapped active faults from geological investigations. High-velocity anomalies (Vp > 5.2 km/s) observed at depths of 2–5 km are interpreted as uplifted crystalline basement or competent metamorphic rocks related to orogenic processes.
These shallow high-velocity bodies likely act as geothermal heat sources and structural controls for fluid circulation, explaining the development of surface hot springs. Our results demonstrate that AI-assisted seismic tomography provides an efficient and practical framework for geothermal exploration in complex tectonic environments.
How to cite: Guan, Z.-K., Kuo-Chen, H., Sun, W.-F., and Pan, S.-Y.: GEOthermal SEISmic AI Platform (GEOSEIS-AI):AI-assisted Seismic Tomography for Geothermal Exploration in the Western Foothills of Taiwan, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17354, https://doi.org/10.5194/egusphere-egu26-17354, 2026.
Focal mechanism solutions serve as an effective observational tool for fracture detection in geothermal exploration and monitoring induced seismicity, aiding in the understanding of subsurface stress states. In these monitoring tasks, often involving high-density, small-scale networks, there is a critical need to generate real-time focal mechanism solutions for a large volume of microseismic events characterized by low signal-to-noise ratios. In this study, we develop an automated workflow integrating deep learning models to determine focal mechanisms. To resolve smaller seismic events (especially magnitude < 3), the P-wave first motion method is employed. Validation tests demonstrate that the workflow can rapidly provide a reliable catalog of focal mechanism solutions. The workflow includes: (1) performing signal-to-noise ratio threshold on P-waves to exclude phases with ambiguous polarities; (2) utilizing a suitable deep learning model, RPNet, to determine first-motion polarity, ensuring accurate identification even with arrival time offsets (about 0.02s), which is characteristic of deep learning-based seismic catalogs; and (3) calculating focal mechanisms using three distinct methods: HASH, FPFIT, and FOCMEC, to ensure solution stability, with the Kagan angle used to quantify consistency (smaller differences indicate higher stability). This workflow has been implemented at the Miaoli geothermal field in Taiwan. The resulting focal mechanisms are predominantly strike-slip; the P-axes exhibit a NW-SE orientation, while the T-axes show a NE-SW orientation, aligning with shear wave splitting results. This workflow has been integrated into the GEOSEIS-AI Platform, aiming to get focal mechanisms rapidly and reliably, enhancing our understanding to the seismogenic structure.
Keywords: GEOSEIS-AI; Deep Geothermal Energy; focal mechanisms; deep learning; automated workflow
How to cite: Pan, S.-Y., Sun, W.-F., Liu, Y.-H., and Kuo-Chen, H.: GEOthermal SEISmic AI Platform (GEOSEIS-AI): P-wave First Motion Focal Mechanism Determination Module, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18228, https://doi.org/10.5194/egusphere-egu26-18228, 2026.
Seismic ambient noise analysis has become an important tool for subsurface characterization, offering a cost-effective alternative to active sources and enabling continuous monitoring. Array-based techniques such as beamforming are central to ambient noise analysis, allowing the estimation of wavefield properties such as propagation direction and phase velocity. Traditionally, beamforming has been applied either to vertical-component array data, particularly for surface-wave analysis, or to three-component (3C) seismic arrays, which allow for polarization and wave-type discrimination.
More recently, distributed acoustic sensing (DAS) has emerged as a powerful tool for ambient noise studies, providing dense spatial sampling and large apertures at relatively low per-channel cost. However, DAS measurements are primarily sensitive to axial strain and can therefore be interpreted as effectively single-component observations. As a result, DAS arrays deployed along a single line cannot leverage the benefits of 3C beamforming, such as polarization analysis and wave-type identification. Conversely, sparse 3C arrays provide polarization information but are often limited in wavenumber resolution due to restricted aperture and station spacing.
In this study, we develop and test a joint beamforming approach that combines DAS and 3C seismic observations in a unified framework. The joint beamformer is constructed by combining normalized beam power estimates from DAS-only and 3C-only beamforming, enhancing coherent signals that are consistent across both datasets while suppressing incoherent or aliased energy. The performance of the joint approach is evaluated using numerical simulations in layered elastic media. Systematic tests are carried out for different array geometries and station spacings to investigate their effects on aliasing, resolution, and information gain. The results show that the joint beamformer improves the stability of the results, particularly in cases where DAS-only or 3C-only beamforming suffers from aliasing or limited resolution. Finally, the method is applied to a real test dataset to demonstrate its applicability under realistic noise conditions.
Our study suggests that joint DAS–3C beamforming provides a robust framework for ambient noise analysis, offering improved wavefield characterization compared to single-sensor approaches and highlighting the potential of hybrid array designs for future seismic monitoring applications
How to cite: Kaur, S., Finger, C., and Saenger, E. H.: Ambient Noise Beamforming with Joint DAS and Three-Component (3C) Seismic Arrays, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5919, https://doi.org/10.5194/egusphere-egu26-5919, 2026.
How to cite: Maatouk, N., Cazenave, F., Gerbaud, L., Noble, M., and Velmurugan, N.: Seismic While Drilling as a Passive Source for Imaging Deep Geothermal Systems , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11305, https://doi.org/10.5194/egusphere-egu26-11305, 2026.
Accurate subsurface characterization is fundamental to the successful development of geothermal systems. Such comprehensive knowledge allows determining geological structures that govern local fluid circulation and heat transport. As drilling represents one of the largest cost factors in geothermal development, ensuring that wells target zones of high hydraulic conductivity and permeability can substantially reduce exploration risk and overall project costs. Passive seismic techniques, being both inexpensive and non-invasive, have proven to be effective tools for both geothermal exploration and monitoring. Among them, Horizontal-to-Vertical spectral ratios (H/V) are often used to map subsurface topography. Their interpretation and inversion, however, often rely on prior knowledge of local shear-wave velocity or subsurface layering.
In this case study we employ a trans-dimensional Bayesian framework to invert H/V curves from more than 70 survey points across a prospective aquifer thermal energy storage system in rural Belgium, which will supply about 160 housing units. Our approach enables the generation of pseudo-2D shear-wave velocity profiles across the site without requiring additional information to constrain the inversion. Low velocity zones are identified which can be related to karstification and geological layering suggested by geological maps. The results are further validated with direct field measurements. Borehole logs from exploration wells drilled on the basis of our results indicate high hydraulic conductivity and are supported by water table measurement from cone penetration testing. The derived profiles offer valuable information to guide well placement and optimize drilling decisions by reducing uncertainty in subsurface conditions. Our findings demonstrate that passive seismological techniques, combined with probabilistic inversion approaches can serve as a cost-effective tool in support of the energy transition.
How to cite: Pätzel, J., Yates, A., Depoorter, M., and Caudron, C.: Bayesian Inversion of H/V Spectral Ratios for Constraining Shallow Subsurface Structure in Geothermal Exploration, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18122, https://doi.org/10.5194/egusphere-egu26-18122, 2026.
The Kuusamo Belt of northern Finland formed as an Early Proterozoic rift system within the Archean crust of the Fennoscandian Shield between ~2.5 and 2.0 Ga. Following its formation, it has recorded several major shortening episodes, resulting in a heavily folded medium-grade metamorphic belt. More recently, this region has garnered interest due to the presence of Au and Co deposits within the belt. As a part of the multidisciplinary project UNDERCOVER, we present a preliminary ambient noise tomography model of the region focused on the crustal architecture around these deposits, using 493 nodal seismometers that were deployed over a 35x35 km region from June to August 2025. This dense array was inset within a larger network of 35 broadband stations spanning 150x170 km. Combining these arrays, we extract path‐averaged Rayleigh wave phase‐velocity dispersion measurements for >130,000 interstation paths to constrain the shear wave velocity structure of the upper 40 km. Our preliminary model resolves the large-scale crustal structure with the bulk crust characterized by shear wave velocities >3.0 km/s up to very shallow depths, consistent with wavespeeds sampling Archean greenstones and Paleoproterozoic mafic rocks which outcrop within the study region. To first order, broad-scale velocity perturbations align well with the trend of large-scale folding in the region. Going forward, we hope to take advantage of the dense nodal array and incorporate high-frequency phase velocities in a single model to refine the shallow subsurface structure and better characterize the relationship between velocity anomalies, structural features, and mineralization.
This research has received funding from the European Union through the Horizon Europe project UNDERCOVER (Grant agreement No. 101177528).
How to cite: Rodriguez, E., Sippl, C., Luhta, T., Hill, G., and Kamm, J.: High-resolution imaging of the Kuusamo Belt, northern Finland, from ambient noise tomography, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18875, https://doi.org/10.5194/egusphere-egu26-18875, 2026.
How to cite: Latallerie, F., Hjörleifsdóttir, V., Isken, M., Biondi, E., Obermann, A., and Peidong, S.: Characterising vp/vs ratios beneath geothermal systems of the Hengill volcano in Iceland using a global-local approach, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-1163, https://doi.org/10.5194/egusphere-egu26-1163, 2026.
In the summer of 2025, the Passive Array for Critical Minerals on the Island of Newfoundland (PACMIN) project was launched. The deployed array comprises 22 broadband seismograph stations, which will record local, regional and global earthquakes as well as ambient ground vibrations for a period of two years. This experiment will yield the first ever detailed 3D lithospheric structure models of the entire island of Newfoundland from multiple types of seismic analysis. From these models, we will be able to investigate how the region was shaped by Appalachian mountain-building processes, while also exploring tectonic controls on the distribution of key mineral deposits such as gold and critical minerals. The onshore seismicity of Newfoundland, while low, will also be investigated to better understand and mitigate mining exploration/exploitation hazards. Improved detection and locating of small local earthquakes will also allow fault networks in the shallow crust to be mapped and assessed in terms of their potential as fluid pathways that may carry critical minerals.
How to cite: Welford, J. K., Darbyshire, F., and Long, M.: Passive Array for Critical Minerals on the Island of Newfoundland (PACMIN), EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5405, https://doi.org/10.5194/egusphere-egu26-5405, 2026.
