Based on broadband magnetotelluric (MT) array data, we have obtained the three-dimensional (3-D) electrical resistivity model in western Yunnan, the southeastern Tibetan Plateau where the Quaternary intraplate Tengchong volcanism and seismic activities occur. Our MT model clearly reveals three conductive bodies in the depth ranges of 10–30 km in the Tengchong volcano area, which we interpret as three middle-lower crustal magma chambers associated with the Tengchong volcanism. Seismogenic faults in the Gaoligong Shear Zone (GLGSZ) are characterized by subvertical conductive zones bounded by resistive upper crustal layer on both sides. Earthquakes of moderate magnitudes near the GLGSZ have all occurred within the conductive fault zones at the bottom of the upper resistive crust. Our model also suggests a bifurcation of the crustal flow in western Yunnan, with a southwestern branch running into the Tengchong Block north of the Dayingjiang Fault and a southeastern branch flowing into the Baoshan Block. The current study provides evidence from electrical resistivity structure for the middle-lower crustal magma chambers in the Tengchong volcano area and detailed 3-D electrical structure of crustal channel flow in this active tectonic region.

How to cite: Huang, Q., Ye, T., Chen, X., and Zhang, H.: Seismo-tectonics and magma chambers revealed by the 3D resistivity model in western Yunnan, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2628, https://doi.org/10.5194/egusphere-egu24-2628, 2024.

Ground-penetrating radar (GPR) is a widely acknowledged tool for imaging near-surface environments in various geological, hydrological, and sedimentological applications. In complex and heterogeneous settings, applying 3D GPR is crucial to correctly image subsurface architecture and, thus, to prevent misinterpretations. Recent advancements in GPR system design and instrumentation enable the collection of densely sampled 3D GPR datasets with superior resolution, establishing 3D GPR as a standard for near-surface structure imaging.

This study showcases the latest developments in analyzing and interpreting 3D GPR datasets. We demonstrate how GPR datasets and derived structural models contribute to a detailed understanding of complex near-surface environments. Using selected case studies, we present integrated interpretation approaches combining 3D GPR data and models, respectively, with the results of 3D electromagnetic induction surveying, 2D electrical resistivity tomography as well as 1D geophysical and geological borehole logging. Such a strategy allows for a more comprehensive and reliable near-surface characterization by integrating detailed 3D structural information with electrical/petrophysical property distributions and geological information, surpassing the limitations of typical 2D single-method interpretation approaches.

How to cite: Koyan, P., Guillemoteau, J., Klose, T., and Tronicke, J.: 3D ground-penetrating radar to characterize near-surface environments: Advances in data analysis and integrated geophysical interpretation, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3588, https://doi.org/10.5194/egusphere-egu24-3588, 2024.

The Beishan Orogen, located in the southernmost part of the Central Asian Orogenic Belt, comprises ophiolitic complexes, passive-margin strata, arc assemblages, and Precambrian basement rocks, recording oceanic subduction, accretion of oceanic materials onto continental margins, and continental-arc collision. However, debates surrounding its origin and evolution persist, attributed in part to the absence of high-resolution geophysical data, leading to varying interpretations in tectonic evolution models regarding the involved terranes and paleo-subduction polarity (Li et al., 2023). In this study, we present an electrical resistivity model of the crust and uppermost mantle in the Beishan Orogen from magnetotelluric (MT) data. The resistivity model suggests an overall resistive upper crust, with conductive features indicating paleo-suture zones and tectonic boundaries. The high resistivity lithosphere beneath Niujuanzi indicates both north- and southward subduction of the Hongliuhe-Xichangjing Ocean, potentially unveiling remnants of a cold fossil oceanic lithosphere. Conductors in mantle wedges on both sides of the high resistivity body are inferred to result from recycled carbonates introduced deep into the Earth through oceanic subduction, which is substantiated by recent laboratory measurements (Jing et al., 2023). These measurements demonstrate that even unmelted carbonates can enhance electrical conductivity through cation exchange reactions with silicates in the lower crust and uppermost mantle.

*This research is funded by the National Natural Science Foundation of China (42074089, 41774087, 41404060).

Reference

Jing, C., Hu, H., Dai, L., Sun, W., Wang, M., Hu, Z., 2023. Recycled carbonates elevate the electrical conductivity of deeply subducting eclogite in the Earth’s interior. Commun Earth Environ 4, 276. https://doi.org/10.1038/s43247-023-00936-w

Li, J., Wu, C., Chen, X., Zuza, A.V., Haproff, P.J., Yin, A., Shao, Z., 2023. Tectonic evolution of the Beishan orogen in central Asia: Subduction, accretion, and continent-continent collision during the closure of the Paleo-Asian Ocean. GSA Bulletin 135, 819–851. https://doi.org/10.1130/B36451.1

How to cite: Zhou, L., Zhang, L., and Jin, S.: Magnetotelluric Evidence for the Paleo-Subduction Polarity and Recycled Carbonates within the Beishan Orogen, Southern Central Asian Orogenic Belt , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-715, https://doi.org/10.5194/egusphere-egu24-715, 2024.

How to cite: Donoso, G. A., Rydman, O., Yu Smirnov, M., Juhojuntti, N., and Van Den Berg, H.: Seismic constrained magnetotelluric inversion for iron orebody at Gällivare, Sweden, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-20378, https://doi.org/10.5194/egusphere-egu24-20378, 2024.

We designed an computer efficient, probabilistis 3-D inversion algorithm for CSEM data. It is commonly known that multiple physical descriptions of the subsurface fit the geophysical data equally well, due to incomplete measurement coverage, incomplete physical description of the problem, and model overparameterization. A probabilistic inversion approach allows us to explicitly account for data and model uncertainty. Probabilistic inversion is a demanding process in terms of the number of forward model computations required to sample the posterior probability density function of model parameters. Therefore, we use an efficient sampling algorithm (DREAM(ZS)) combined with strategies to optimize the forward model by approximations and error estimation with deep learning techniques. Regarding the latter, we propose an alternative (approximate) approach to our forward model for simulating the electromagnetic (EM) response which reduces the computing time, and we quantify the modeling error committed directly in the inversion process. Thus, we create a statistical error-model related to the approximate EM response by training a Spatial Generative Adversarial Network (SGAN). In contrast to other neural networks, the SGAN training process has the particularity of being a competition between a Generator, which creates fake samples of the training set, and a Critic, which scores the quality of both true or fakesamples. After training the Generator results in a parametric model of the probability density function of the training set (modeling errors). This parametric error-model is incorporated into the inversion process as a complement to correct and quantify the error in our approximate forward model. To test our methodology we first proposed a synthetic experiment of a marine exploration environment. The implementation and subsequent training of the network allowed us to show that SGAN is useful to generate a statistical error-model. The comparison between a set of samples created with the Generator and the training set shows similarities in the statistical properties of both. Thus, we obtain a parameter-reduced error-model capable of representing the different components of the EM response at a considerable number of receivers and frequencies. In addition, the inversion process is significantly accelerated by introducing the forward model approximations, and the incorporation of the statistical error-model improved the determination of the true parameters in our synthetic test case. We then applied our methodology to the inversion of CSEM data acquired in a marine environment.

How to cite: Elias, M., Rosas-Carbajal, M., and Zyserman, F. I.: Deep learning techniques applied to 3-D probabilistic inversion of controlled-source electromagnetic data in a marine environment, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5910, https://doi.org/10.5194/egusphere-egu24-5910, 2024.

Imposing structural constraints between grids of variable sizes is problematic because different geophysical techniques employ different grid divisions. We propose a new collaborative inversion approach and utilize it to invert the proposed 2-D ATEM and 2-D Magnetic Methods with Induced Polarization (IP) effects, which overcomes the problem of imposing cross-gradient constraints at different grid sizes. The grid mapping technique is included into the iterative collaborative inversion process by the inversion strategy. The combined data inversion results show that the technique may successfully leverage data complements to improve the accuracy of the medium boundary description results. The concept is suitable to collaborative inversion of geophysical methods with arbitrary grid divisions and successfully solves the problem of mismatched grid divisions in collaborative inversion.

How to cite: Lei, D. and Ren, H.: A cooperative inversion of airborne transient electromagnetic and magnetic methods based on grid mapping, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7184, https://doi.org/10.5194/egusphere-egu24-7184, 2024.

Rising temperatures as well as accompanying changes in precipitation rates and periods due to climate change demand an efficient use of water resources for farming. The Yucatan peninsula (Mexico) is not only affected by the climatic changes per se but also by their consequences, such as the deepening of the groundwater table and seawater intrusion, enhanced by population growth. These threats call for the development of efficient irrigation methods to maintain farming activities. Moreover, the management of water resources needs to consider irregular water infiltration as well as groundwater flow and storage due to the presence of fractures, caves and low permeable limestone associated to the karstic geology of the peninsula. We evaluate the application of the transient electromagnetic (TEM) method to gain information about subsurface architecture, in particular to identify the presence of unknown cavities that may act as preferential flow paths. The TEM is a geophysical method well suited for hydrogeological investigations in karst environments as it resolves the subsurface electrical conductivity without galvanic contact with the ground and with a higher spatial resolution compared to borehole data. The TEM survey was planned for a depth of investigation of ca. 60 m across a citrus farm on the Yucatan peninsula where treated wastewater is used for irrigation. While such practice minimizes the exploitation of groundwater, some concerns have been raised about the possible contamination of the aquifer due to with the use of treated wastewater. The difference in the electrical conductivity between the treated wastewater and the groundwater due to their different chemical composition renders the TEM as a suitable method to delineate pathways of the irrigation water within the subsurface as well as hydraulic connections with the aquifer. In a first step, the TEM soundings were inverted independently to enhance the spatial variability associated to the complexity of the karst system. Interpretation of the resulting electrical models in terms of the aquifer geometry and preferential flow paths took into account the existing information from boreholes and irrigation points. We also conducted a sensitivity analysis of the resolved model parameters after the inversion of the data, to better evaluate the interpretation of our results. In a second step, we conducted a stochastic analysis of the TEM data to quantify the uncertainty of our results, in particular, regarding the resolved geometry of the aquifer. Our results reveal changes in the electrical conductivity at different depths and across the farmland. High conductivity values, which were observed close to the surface, are related to the infiltration of the treated wastewater. Deeper variations in electrical conductivity reveal the presence of caves and other preferential flow paths for groundwater. Areas revealing high resistivity values are associated with less karstified rocks that may act as hydraulic barriers.

How to cite: Flores Orozco, A., Aigner, L., Montes-Avila, I., Moser, C., Cerca-Ruiz, F., Giácoman-Vallejos, G., and Cardona-Benavides, A.: Application of transient electromagnetic to understand infiltration in farmlands in karst areas of Yucatan, Mexico, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11238, https://doi.org/10.5194/egusphere-egu24-11238, 2024.

The forward modelling for curl-curl equations is the fundament for time-harmonic electromagnetic (EM) problems in geophysics. The simulations with the discretized partial differential equations (PDE) are computationally intensive and crucial to practical geophysical EM problems like Magnetotelluric/Controlled Source EM. However, most published algorithms for curl-curl PDEs are still CPU-based and cannot utilize the rapid development of modern large-scale multi-GPU parallel architectures. Based on previously proposed CPU-based divergence-free modelling algorithm, we develop a hybrid parallel paradigm to exploit the high-throughput of interconnected heterogenous parallel systems equipped with multiple GPUs. The large sparse linear system derived from the staggered-grid finite-difference approximation of curl-curl problem is decomposed into sub-domains and solved efficiently with a mixed-precision Krylov subspace GPU algorithm in parallel. To demonstrate how the practical inversion problems can be substantially accelerated, we test the new algorithm with both the synthetic and real-world 3D models, for forward/adjoint calculations. The test results show a promising ~3.5x improvement regarding the computation speed on a small NVIDIA® HGX based system, over a conventional CPU-base server cluster with 12 nodes. This may significantly reduce the computation time and carbon footage for large-scale frequency domain EM inversion problems and brings the possibility of near real-time EM imaging, as in engineering and environmental applications.

How to cite: Dong, H. and Sun, K.: Multi-GPU accelerated parallel modelling for geophysical electromagnetic inversions , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13759, https://doi.org/10.5194/egusphere-egu24-13759, 2024.

The  Ernest Henry copper mine is one of the main Australian copper resources and a typical Iron Oxide Copper Gold (IOCG) deposit case. These deposits have specific geophysical signatures, e.g., considerable magnetic susceptibility and low resistivity due to magnetite mineralization. This research studies the geophysical models estimated via the inversion of magnetic and magnetotelluric in the Ernest Henry mine area. We also conducted these inversions in two other areas close to the Ernest  Henry mine, named A and B, which show magnetic anomalies similar to the Ernest Henry mine. Then, we compared the estimated models in these areas with those in  Ernest Henry's area. Three-dimensional inversion of magnetic data using the Li and Oldenburg algorithm revealed masses with magnetic susceptibilities higher than 0.019, 0.048, and 0.011 in SI units in the Ernest Henry, A and B areas, respectively. All the masses were extending from the surface to depths of one to three kilometres. Next, impedance analyses of magnetotelluric data indicated a two-dimensional behaviour of the Earth up to a frequency of 1 Hz in all the areas. Also,  we conducted two-dimensional inversions of these data along a profile for each area. A comparison of the estimated resistivities demonstrated a relatively conductive mass with a resistivity of less than 30 ohmm at depths of 5 km and beyond in all three areas. These models also demonstrated a decrease in the resistivity along the fault lines within the ranges, corresponding to the location of masses with relatively significant magnetic susceptibility identified during the magnetic data inversions. Geochemical analysis of copper grade variations in exploratory boreholes drilled in areas A and B indicated an increase in copper concentration exceeding 400 ppm in both areas. The structure of the estimated geophysical models in areas A and B and those estimated in the Ernest Henry area is similar. Besides, geochemical analysis of copper grade in exploratory boreholes drilled in areas A and B indicated an increase in copper concentration exceeding 400 ppm in both areas. Therefore, we deduce the possibility of IOCG copper mineralization deposits similar to the Ernst Henry mine in areas A and B. Meanwhile, since the concentrations of probable mineralizations have occurred along a fault zone in all the models, we suspect that the mineralization likely originated from hydrothermal solutions. These solutions spread from fault lines adjacent to highly resistive intrusive masses to the surface, causing the mineralization of magnetite and copper ores.

How to cite: Sakhaeyan, F., Abtahi Forooshani, S. M., and Sadeghisorkhani, H.: Exploration of IOCG deposits in Queensland, Australia using geophysical models of Ernest Henry copper mine, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14779, https://doi.org/10.5194/egusphere-egu24-14779, 2024.