A 3D inversion method of TEM combining PSO-NLCG optimization and adaptive regularization.

ABSTRACT To meet the increasing demand for raw materials and to foster the energy transition, advances in exploration techniques are required that extend capabilities to deeper and/or more densely populated areas while having a low environmental and societal impact. Although active electromagnetic (EM) geophysical techniques have a long tradition in mineral exploration, magnetotellurics (MT) as a passive EM method has recently attracted more attention in this context, as it naturally offers a wider range of survey depths while being even less invasive. Here, we present the results of an MT study conducted at the Stonepark Zn–Pb deposit, an exploration project hosted in Mississippian carbonates and volcanics in the Irish Orefield. The survey was carried out in late 2022, using a novel experimental layout of 33 five‐component broadband MT stations deployed in concert with 75 two‐component electric field only stations. Resulting data exhibit strong EM noise but reasonable MT transfer functions could be obtained in the frequency range of 10 4 –10 −2 Hz by using a combination of the robust remote reference (RR) technique, notch filtering and physical and statistical pre‐selection thresholds. Quality of full MT and E‐field stations is the same demonstrating the applicability of the hybrid surveying approach also in noisy environments. The electrical conductivity structure at Stonepark revealed by 2D and 3D inversion shows the host rocks as high resistivity material (500–5000 Ωm) typical of limestones with little lateral variations in resistivity across the survey area. Although the MT models do not provide a direct image of economically relevant mineralized zones, they are in excellent agreement with resistivity measurements on rock samples from nearby drill holes. This is remarkable because laboratory measurements are made on centimetre‐sized samples, whereas MT soundings sample resistivities over volumes of 10–100 s of meters. Integration of the MT results with the drill hole data, a 3D geological model of the Stonepark mineral system and a reprocessed seismic profile allowed the continuation of the horizons to the south where they are sparsely documented by drill holes. In particular, the MT data revealed alteration of volcanic material in the upper 200–800 m as well as a subvertical conductive feature that is spatially coincident with a hidden fault zone inferred from the seismic data, suggesting that the fault zone may be fluid enriched.

3D SPT-seismic full-waveform inversion of S- and P-wave fields with correlation to SPT-N in karstic terrain of Florida, USA

3D gravity inversion using basement well depths applied to the Recôncavo Basin

Abstract Electrical resistivity imaging (ERI) shows promise for investigating earthen flood barriers. We are interested in the applicability of ERI for aiding maintenance and construction efforts on agricultural dikes in the upper Bay of Fundy, Canada. An overarching research goal is to develop efficient survey and data processing procedures for detecting internal issues in such dikes ahead of more detailed investigations. We are developing detailed field survey and inverse modelling protocols to help us reliably target hypothetical subsurface features, and this article presents a critical step in that process: we performed a detailed investigation to assess and adapt existing common practices for three‐dimensional (3D) inversion mesh design for seawater‐adjacent agricultural dike‐imaging scenarios. We worked in 3D because dikes have significant 3D geometry and hence two‐dimensional inversion is unlikely to be adequate. We used a representative 3D dike model for the region and worked with the specifics of our surveying equipment, yet our conclusions may have wider ranging relevance. We investigated the effect of various mesh generation parameters on forward modelling accuracy, and we conclude by providing mesh design recommendations for dike investigations.

This study combines airborne magnetic and magnetotelluric (MT) surveys to enhance deep subsurface geologic characterization and groundwater prospecting in the Voltaian Sedimentary Basin (VSB) in Ghana. A comprehensive analysis of airborne magnetic data delineated potential subsurface lineaments, which subsequently informed the selection of sites for magnetotelluric (MT) field investigations. Resistivity maps derived from both 3D depth slices and 1D inversions of the magnetotelluric profiles, and jointly interpreted with lineaments from airborne magnetic, defined a deep zone of low resistivity formation indicating a probable groundwater-bearing zone. The resistivity signature of the structure is comparable to that of a unit in which a 132 l/min-yielding research borehole in close proximity is completed. This promising conductive layer is approximately 30 m thick at a depth of 260 m below ground level and is about 45 km long. This has presented new insights into the stratigraphic thickness of the two regionally significant geological formations: the Panabako and Poubogou Formations. The new data suggest thicker units than are currently known in the current literature. This finding is significant and will affect the current understanding of the geology and groundwater resource productivity of the Voltaian Basin.

One of the most remarkable features of the eastern Australian continent are Cenozoic age (65 Ma to present) volcanoes over 3,000 km span north-south, with no significant age progression, even though the continent has moved at a rate of up to 75 km/Ma in an NNE direction over much of this time. Three theorems have been advanced to explain age-independent volcanism: (1) decompression melting from the transition zone due to volatile content from subducted slab stagnation; (2) edge-driven convection at the margins of steps in lithospheric thickness that drive upwelling from volatile-rich mantle reservoirs; and (3) melting of low-viscosity pockets of sub-lithospheric mantle due to asthenospheric shear. In this paper we have undertaken a 3D inversion of magnetotelluric (MT) sites across the Tasmanide accretionary orogens in eastern Australia (~ 800 locations) to define the broad-scale resistivity of the lithosphere and asthenosphere. The upper mantle directly beneath the age-independent Cenozoic volcanoes has anomalously low-resistivity (~ 100 Ohm.m) below 125 km depth, compatible with dry lherzolite-harzburgite-wehrlite compositions at temperatures of ~ 1,400 °C. The resistivity model suggests that inland from the volcanos, there is a step-like increase in lithospheric thickness. In the lower crust, beneath volcanic centres a resistivity of ~ 50 Ohm.m is consistent with a hydrated clinopyroxene-orthopyroxene-plagioclase composition at ~ 900 °C. No additional conduction mechanisms (such as graphite, sulphides or partial melt) are required. Results suggest that decompression melting from the transition zone raises the geotherm to adiabatic below 125 km and efficiently removes volatiles to the crust over a wide area (hundreds of kilometres) of subducted slab. Edge-driven convention may also occur at the step in the lithosphere thickness, particularly for the New Volcanic Zone and the leucite Cosgrove Track volcanoes. However, surface volcanism occurs in a much narrower zone where the melt solidus is intersected at lower crustal depths.

Abstract Aircraft-based surveying to collect airborne electromagnetic data is a key method to image large swaths of the Earth’s surface in pursuit of better knowledge of aquifer systems. Despite many years of advancements, 3D inversion still poses challenges in terms of computational requirements, regularization selection, hyperparameter tuning and real-time inversion. A new approach is introduced for the inversion of airborne electromagnetic data that leverages machine learning to overcome the computational burden of traditional 3D inversion methods, which implicitly includes learned regularization and is applicable in real-time. The method combines 1D inversion results with geostatistical modeling to create tailored training datasets, enabling the development of a specialized neural network that predicts 2D conductivity models from airborne electromagnetic data. This approach requires 3D forward modeling and 1D inversion up front, but no forward modeling during inference. The workflow is applied to the Kaweah Subbasin in California, where it successfully reconstructs conductivity models consistent with real-world data and geological drill hole information. The results highlight the method’s capability to deliver fast and accurate subsurface imaging, offering a valuable tool for groundwater exploration and other near-surface applications.

Abstract Transient electromagnetic (TEM) inversion aims to estimate subsurface resistivity structures from measured transient responses. Although one-dimensional (1D) inversion methods are widely used, accurately resolving deep resistivity variations remains challenging due to the complex temporal dependence between early- and late-time responses. To address this issue, we propose a deep learning framework for efficient 1D TEM resistivity inversion. The model combines a multiscale convolution module to capture response features at different temporal scales, a Transformer encoder to model long-range dependencies, and a U-Net-based encoder-decoder architecture with skip connections to preserve structural information during feature reconstruction. Experiments on synthetic datasets show that our proposed method outperforms conventional CNN, Transformer, and MLP models in both accuracy and robustness. Quantitative results indicate RMSE values of 0.0633 and 0.0442 and RE values of 0.0358 and 0.0404 on the 7-layer and 30-layer datasets, corresponding to RMSE reductions of 55.7% and 28.5% compared with the CNN baseline. In particular, the model shows enhanced capability in recovering deeper resistivity structures. Application to field TEM data further confirms its practical effectiveness. These results suggest that our proposed framework provides an effective data-driven strategy for improving the efficiency and reliability of 1D TEM resistivity inversion.

We propose and test a method to reduce the dimensionality of Full Waveform Inversion (FWI) inputs as a computational cost mitigation approach. Given modern seismic acquisition systems, the data (as an input for FWI) required for an industrial-strength case is in the teraflop level of storage; therefore, solving complex subsurface cases or exploring multiple scenarios with FWI becomes prohibitive. The proposed method utilizes a deep neural network with a binarized sensing layer that learns by compressed learning seismic acquisition layouts from a large corpus of subsurface models. Thus, given a large seismic data set to invert, the trained network selects a smaller subset of the data, then by using representation learning, an autoencoder computes latent representations of the shot gathers, followed by K-means clustering of the latent representations to further select the most relevant shot gathers for FWI. This approach can effectively be seen as a hierarchical selection. The proposed approach consistently outperforms random data sampling, even when utilizing only 10% of the data for 2D FWI, and these results pave the way to accelerating FWI in large scale 3D inversion.

Cerebral adrenoleukodystrophy (CALD) is a common manifestation of adrenoleukodystrophy (ALD) in men. Early detection of CALD lesions through MRI screening is critical to allow for therapeutic action preventing severe disability and death. While the frequency of brain MRI monitoring has been addressed by international recommendations, no consensus currently exists regarding which MRI sequences should be used in a real-world setting for screening and follow-up of CALD lesions. The aim of this study was to establish guidelines for the MRI protocol in clinical practice and to identify priority sequences for research use, thereby promoting intercenter harmonization. A modified Delphi procedure was used to achieve consensus on MRI protocols for ALD screening, lesion monitoring, and research applications among experts with experience in brain imaging in ALD. Questionnaires allowed experts to indicate whether they considered sequences as core, optional, or research, or to express agreement (5-point scale ranging from completely disagree to completely agree) with specific statements. Topics where no agreement was reached were discussed during online consensus meetings. Thirty experts from 9 countries participated and agreed that the core screening protocol for ALD in adults and children should include at least 3D T1-weighted, spin-echo T2-weighted, 3D fluid-attenuated inversion recovery, and diffusion-weighted imaging (DWI). Postcontrast T1-weighted imaging should be performed systematically in specific clinical scenarios. Experts supported using DWI alongside the Loes score and postcontrast imaging to assess lesion progression. A research protocol was defined, prioritizing diffusion tensor imaging, MR perfusion, and quantitative volumetric analyses. This international project harmonizes the ALD MRI protocol, thus offering a practical framework to screen and monitor lesions, which will improve clinical decision making. It also identifies MRI sequences that should be prioritized in future research. Future research on MRI in ALD should focus on topics where no consensus has yet been reached in this project.

Abstract The archaeological site of Buto (Tell el-Fara'in) in the northwestern Nile Delta, Egypt, is a multilayered settlement with a complex occupational history spanning from the Predynastic period to the Early Islamic era. The workflow of this study is mainly focused on using_ not on developing new methods_ SAR (Sentinel-1 GRD) satellite imagery along with the electrical resistivity tomography (ERT) measurements, and excavation process to investigate the site's settlement phases, particularly in relation to subsurface architectural remains in the archaeological Tell of Buto (Kom C). Sentinel-1 (C-band) satellite imagery captured on May 5, 2018, was processed using SNAP software (version 9.0.0) to identify large-scale anomalies indicative of buried structures. Based on these detected anomalies, the locations for ERT profiles were strategically selected to maximize the likelihood of indicating significant subsurface features. ERT was applied in a quasi-3D survey mode across Kom C, using parallel 2D survey lines. A comparison between quasi-3D images—created by merging inverted 2D sections—and fully 3D inverted resistivity models highlights the superior accuracy of the 3D inversion algorithm in enhancing subsurface imaging and improving the interpretation of buried archaeological features, even in a site as intricate as Buto. Both horizontal resistivity depth slices (tomoplanes) and voxel-based 3D resistivity models provide critical insights into the subsurface architectural remains. The upper layers (0–3 m) exhibited diffuse resistivity patterns with scattered anomalies, indicating the presence of remnants of Ptolemaic or Roman archaeological material, including mudbricks, limestone debris, and pottery fragments, which were likely displaced by natural or human activity. Further, at a depth ranging from 3 to 6 m, a well-defined high-resistivity anomaly was identified as a Saite period (26th Dynasty, seventh–sixth century BCE) mudbrick structure, possibly a large tomb or shrine, resting on an artificial sand foundation or sandboxes. The sand layer of high-resistivity values was identified at a depth of 6–7 m, indicating intentional ground levelling during the Saite period, with a deeper layer potentially dating from the late eighth century BCE. Integrating Sentinel-1 data with ERT results provided multi-scale insights, guiding the excavation process, which was carried out over a 10 × 10 m area divided into four squares, revealing mudbrick walls and religious artefacts, further validating the geophysical and remote sensing interpretations. The results of this study demonstrate the effectiveness of combining geophysical measurements and remote sensing data, which gave a very accurate vision in detecting buried settlements in a complex region.

Development of quantitative geophysical proxy indicators for urban underground space use: An analysis of Stockholm and its city plan

Subsurface architecture of the Kachchh mainland and Katrol Hill Fault systems, Western India, using 3D constrained gravity and magnetic inversion

When a turbulent Bose-Einstein condensate (BEC) is driven out-of-equilibrium at a scale much smaller than the system size, nonlinear wave interactions transfer particles towards large scales in an inverse cascade process. In this work, we numerically study wave turbulence in a three-dimensional BEC in forced and dissipative inverse cascade settings. We observe that when the forcing rate increases, thereby increasing the particle flux, the turbulence spectrum gradually transitions from the weak-wave Kolmogorov-Zakharov cascade to a critical balance state characterized by a range of scales with balanced linear and nonlinear dynamic timescales. Further forcing increases lead to a coherent condensate component superimposed with Bogoliubov-type acoustic turbulence. The role of vortices in such a strongly forced state is marginal, which makes this new state distinct from the strongly turbulent state composed of a tangle of quantized vortex lines. We then use our predictions and numerical data to formulate a new out-of-equilibrium equation of state for the 3D inverse cascade.

This paper presents the results of the pilot airborne electromagnetic (AEM) and magnetic survey conducted over a sector of Great Salt Lake (GSL), Utah, aimed at investigating the potential presence of a large freshwater reservoir beneath the lake —the largest inland saline water body in the Western Hemisphere. A 3D inversion of AEM data reveals a laterally extensive resistive layer underlying the lake’s highly conductive brine. This resistive unit is interpreted as freshwater-saturated sediments or bedrock. Its identification provides evidence for substantial freshwater resources beneath the hypersaline layer. The presence of freshwater-saturated sediments beneath the saline water layer has been independently verified by direct measurements of water salinity and chemistry from cores collected in the wells. It has been demonstrated that the AEM method can overcome the challenge of imaging through a hypersaline surface layer to map deeper freshwater. Complementing the AEM data, the magnetic data and their inversion into remanent magnetization provided valuable insights into deeper basement structures. Based on the results of the pilot project, we anticipate that airborne geophysics will enable basin-scale surveys beneath the expansive (> 4,000 km2) footprint of the Great Salt Lake, as well as surveys of other terminal lakes and offshore freshened groundwater.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-40995-5.

This paper deals with the lithological association and geometrical relationship between highly related magnetized bodies in the Djadom area located in the southeastern Cameroon and the North-west (NW) part of the Congo craton, which is poorly documented. To address the issue, a combined approach that involves multiscale (Residual Magnetic Intensity (RMI) and horizontal Gradient magnitude(HGM)) at different altitudes, analytic signal and three-dimensional (3D) inversion modeling were performed on aeromganetic data via Geosoft software. The application of multiscale method to the RMI and HGM at different altitudes highlights structural features and geometrical magnetic source bodies from a shallow to a depth of 6 km. Furthermore, three major WSW – ENE, WNW – ESE and W – E striking faults correspond respectively to the Djadom, Eta, and Dja faults. The analytic signal revealed high magnetization bodies along the structures. The 3D inversion model highlights high susceptibility (S > 0.02 SI) dome-like bodies that are scattered and a lower susceptibility (S < 0.02 SI) sedimentary formation associated with Djadom and Eta faults. The highest susceptibility dome-like magnetized bodies could correspond to Banded Iron Formation-hosted (BIF) formations, whose geometry was inherited from Archean tectonics ascribed to vertical tectonics that affected the study area. During mineral exploration and subsurface geophysics, these aeromagnetic maps are an important tool for geological mapping and for identifying mining targets.

The identification and delineation of concealed mineralized zones in settings with weak or overlapping anomalies remains a critical challenge. Conventional geophysical methods provide limited resolution and reliability in such conditions. To overcome this limitation, this study introduces a systematic framework that integrates fractal clustering and geophysical inversion to enhance the accuracy of mineral exploration. Induced polarization (IP) chargeability data, acquired using a rectangular array at the Kabudan gold prospect in northeastern Iran-an area characterized by the Lack of outcrops and surface indications of mineralization-were analyzed using four well-established fractal models: Concentration-Area (C-A), Concentration-Perimeter (C-P), Concentration-Number (C-N), and Number-Size (N-S). To quantitatively evaluate the performance of each model, four statistical validation indices were employed: Silhouette, Davies-Bouldin, Calinski-Harabasz, and cluster stability. Among these models, The C-P fractal model exhibited the highest clustering quality, with the highest Silhouette index (closest to 1 among the models), the lowest Davies-Bouldin index, the highest Calinski-Harabasz index, and the lowest Silhouette index standard deviation (highest cluster stability). To verify the subsurface continuity of the identified anomalies, Four geoelectrical profiles were acquired over the anomalous zones, and two-dimensional (2D) inversion of the induced polarization (IP) and resistivity data was performed. The data were subsequently modeled, and the corresponding cross-sections were generated to illustrate the subsurface variations. The inverted sections revealed coherent chargeable structures that closely corresponded to the clusters derived from the fractal models. The results were further assessed and validated using borehole data, where the correspondence between a high-grade gold-bearing sulfide zone and the anomalies delineated in the profiles confirmed the reliability and accuracy of the interpretations. Overall, the proposed integration of fractal clustering, geophysical inversion, and statistical validation not only enhances the interpretability of subsurface data under complex geological conditions but also provides a scalable and transferable framework for next-generation mineral exploration.

Multiport microwave power splitters are key building blocks in high-frequency systems such as phased arrays, beamforming networks and measurement setups, but are usually designed using fixed circuit topologies that are difficult to adapt to many-port or unconventional layouts. This paper introduces a scalable inverse-design framework for multiport microwave power splitters that is directly compatible with three-dimensional printing. Here we combine gradient-based optimization with adjoint electromagnetic simulations to automatically shape a dielectric device that meets specified waveform targets at multiple output ports. The method is demonstrated on a four-port power splitter operating at ten gigahertz, fabricated using a polymer powder bed fusion process (multi jet fusion) with simple constraints on minimum feature size and material permittivity. Numerical simulations and waveguide measurements show close agreement in transmission, reflection, and port-to-port balance, indicating robust performance despite manufacturing tolerances. The approach is topology-agnostic and fabrication-aware, enabling economical prototypes and systematic scaling to devices with many ports. This work establishes a general route for integrating inverse design and three-dimensional printing in microwave engineering, and could be extended to other radio-frequency and millimetre-wave components.

FTCSEM—A FORTRAN-based parallelized 1D CSEM forward and inversion program for arbitrary source-receiver geometry