Marine controlled-source electromagnetic (CSEM) inversion is a widely used technique for imaging subsurface structures beneath the seafloor and exploring hydrocarbon or mineral resources. Although essential for data interpretation, the inversion process is computationally intensive, particularly due to the high memory demands and time consumption associated with sensitivity calculations, even in 2D cases. This paper presents recent advances in frequency-domain marine CSEM inversion algorithms, including a novel method for approximating sensitivities that significantly accelerates 2D inversion without relying on exact sensitivity matrices. A 2.5D frequency-domain CSEM forward modeling solver is employed as a testbed, where the secondary-field formulation is adopted to mitigate source singularities and enhance numerical accuracy. In the developed 2D inversion routine, the sensitivity matrix is computed via the adjoint-state method, integrating the inner product of the adjoint and forward electric fields over each cell during each iteration. The inversion is performed using the Gauss–Newton optimization scheme. To improve computational efficiency, adjoint fields are evaluated over a simplified layered model rather than the full multidimensional true model. Numerical experiments demonstrate that the proposed sensitivity approximation achieves nearly identical inversion performance to that of the exact sensitivities. Results from both synthetic and experimental datasets confirm that the approximate sensitivities are sufficiently accurate to ensure convergence toward geologically meaningful solutions.

To address the challenges of low detection efficiency and limited accuracy in identifying contamination on offshore photovoltaic platforms, this study proposes an enhanced YOLOv8-based algorithm for detecting salt deposit on photovoltaic modules. The SimAM parameter-free attention mechanism is integrated at the end of the backbone network and within the neck layers to improve feature representation of salt deposits under complex environmental conditions, thereby enhancing detection accuracy. In addition, the WIoU loss function is employed in place of the original CIoU loss to alleviate harmful gradients caused by low-quality data and to strengthen the generalization capability of the model. A dedicated dataset of salt accumulation images from offshore photovoltaic panels is constructed to support this targeted detection task. Experimental results demonstrate that the proposed algorithm achieves an mAP50 of 85.8%, a 3% improvement over YOLOv8, while maintaining a detection speed of 67 frames per second. These findings confirm that the proposed approach meets both the accuracy and efficiency requirements for automated detection of salt deposition on offshore photovoltaic modules.

Assessing the propagation of the marine controlled-source electromagnetic(EM)fields in shallow waters presents a formidable challenge.This paper addresses this complexity by applying marine controlled-source EM theory and asymptotic expression for the airwave.Our methodology utilizes a vector subtraction technique to separate the electric field,allowing for a thorough investigation of marine electric field propagation.The numerical tests in this study reveal salient phenomena including total reflection,half-wave loss,and interference cancellation as the electric field propagates through a wave field with multiple paths.In shallow waters,the interface multiple reflection should not be overlooked.To better understand the subseafloor conductivity structures in marine controlled-source EM surveys,effective airwave suppression strategies are required,particularly those that travel through seawater and the sea-air interface.Moreover,for successful underwater target detection,careful consideration must be implemented to suppress guided waves along the seafloor interface,particularly in near-source field data.In contrast,for field data collected distant from the source,the airwaves,and multiple reflections become strong signals,providing helpful insights for underwater target detection.

Accurate positioning is important for improving the efficiency of repairing submarine cables and reducing the related repair costs. The magnetic anomaly produced by a submarine cable can be used to estimate its vertical and horizontal positions. A novel approach using magnetic data for estimating the position of submarine cables based on the 1D residual convolutional neural network (RCNN) is investigated. Infinitely long ferromagnetic cylinder models with different parameters are used to generate data sets for model training and testing. Tests on noisy synthetic data sets show that the developed 1D RCNN method can capture detailed features related to the magnetic source position information, which is more accurate than the conventional Euler method in estimating the position of submarine cables. The developed 1D RCNN method has also been successfully applied to processing field data. Furthermore, the processing workflow of our 1D RCNN method is less noise-sensitive compared with the conventional Euler method. The proposed 1D RCNN method and its workflow open a new window for estimating the position of submarine cables using magnetic data.

A stable and fast inversion is essential for marine controlled-source electromagnetic(CSEM)data interpretation.In this paper,a 2D marine CSEM inversion algorithm based on the improved Gauss-Newton optimization algorithm is developed.For the 2.5D CSEM forward solver used in inversion,the staggered finite-difference(SFD)discretization is applied and the linear system of equations is solved by a direct matrix factorizing solver,which could solve multi-transmitter electromagnetic(EM)responses efficiently with only one time of the matrix factorization per frequency.Furthermore,the improved interpolating algorithm is applied for accurately calculating the EM fields for arbitrarily located seafloor receivers in a more efficient way.The adjoint-equation method is used for calculating the sensitivities implicitly and the Gauss-Newton optimization algorithm with quasi-quadratic convergence is applied which makes the 2D inversion developed converges in a more efficient way.A way using the sensitivity matrix is also applied for selecting the regularization parameter automatically to make the inversion converge in a more stable way.Numerical tests demonstrate the efficiency and stability of the inversion algorithm developed.

The electromagnetic (EM)-field components are commonly used in marine controlled-source EM (CSEM) inversion in which accurate orientations, i.e., the heading, tilt, and roll of the sources and receivers, are required for reliable data processing and inversion. However, potential uncertainties may be introduced in inversion and interpretation, especially when the orientation is lost during the field data collection when the compass and tilt recordings are unavailable. This study presents a new marine CSEM inversion algorithm using orientation-independent rotational invariants instead of EM fields, which permits not considering the source or receiver orientations in inversion. The application of the rotational invariants for the frequency-domain marine CSEM data inversion is presented. The Gauss-Newton optimization is carried out in the frequency domain for inverting the CSEM data, which iteratively searches for the seafloor resistivity compatible with the data. The sensitivity related to the rotational invariants is also computed. Numerical tests for both 1-D and 2-D cases using the rotational invariants and conventional EM-field components are validated and compared.

The Pearl River Mouth Basin (PRMB) is located on the northern margin of the South China Sea (SCS). Heat flow measurements indicate that the PRMB is a typical 'hot basin' characterized by a high background heat flow. However, the tectono-thermal evolution of the PRMB and the mechanisms involved are still controversial due to the different input parameters used in tectono-thermal models, small datasets and the limited area of the basin studied. We used tectono-thermal modelling with a multi-stage finite stretching model, constrained by extensive well data, to systematically analyse the tectono-thermal evolution of the PRMB since the onset of rifting. The study area was expanded relative to previous studies to cover both the proximal and hyperextended regions of the northern SCS margin, providing a more comprehensive understanding of its tectono-thermal history. Numerous wells and seismic data covering the entire basin were used, along with appropriate input parameters, to address the gaps in previous studies and to enhance the accuracy of the results. Our findings indicate that the PRMB underwent two phases of heating, primarily due to thinning caused by lithospheric extension during rifting. By the end of rifting, the Northern Depression Zone and the Kaiping Sag had reached peak heat flow, while the Panyu Lower Uplift and Baiyun Sag saw their highest heat flow since the initial rifting. The PRMB experienced thermal decay during the post-rift stage, but a significant reheating event occurred from 23.03 to 13.82 Ma, likely due to northwards lower crustal flow from the Baiyun Sag. The Panyu Lower Uplift and the Baiyun Sag had reached their peak heat flow by 13.82 Ma and the heat flow generally increased by 1-3 mW m(-2) in the other studied areas during this stage. The basin's thermal decay slowed at 5.33 Ma and localized heating events linked to magmatic activities in the southern Dongsha Uplift were observed. In general, the proximal domain of the northern SCS margin reached peak heat flow by the end of rifting, whereas the hyperextended domain reached peak heat flow by 13.82 Ma. The heat flow in the hyperextended domain was generally higher than that in the proximal domain as a result of the thinner crust of the hyperextended domain caused by intense multiple lithospheric detachment-extensional thinning processes. During the rifting stage, the lithospheric extensional thinning process was the most important factor influencing the tectono-thermal evolution of the northern SCS margin, while lower crustal flow and magmatism were the most important factors during the post-rift stage.

To process magnetic anomaly data, appropriate parameters for field separation, denoising, and Euler deconvolution must be manually selected. The traditional workflow is inefficient and cannot fulfill the rapid detection of submarine cables due to complex processing and manual parameter tuning. This study presents an end-to-end deep learning approach for the identification and positioning of submarine cables based on magnetic anomalies. The proposed approach effectively establishes a direct mapping correlation between the magnetic field data and the position of the submarine cable. Synthetic tests suggest that our method performs better in terms of positioning accuracy than the conventional Euler method. Our results for the field data are comparable to those obtained using conventional techniques. Furthermore, the proposed method achieves an optimal solution by employing a clustering technique and selecting the solution with the maximum confidence, which avoids spurious solutions associated with traditional methods. The proposed method can directly determine the position of the submarine cables using the raw magnetic field data. Contrary to the traditional processing workflow, field separation and denoising are not necessary in this novel approach, resulting in higher processing efficiency and a simpler processing process.

In shallow waters, the interpretation of the frequency-domain marine controlled-source electromagnetic (CSEM) data is challenging due to the airwave. The airwave dominates and will lead to misinterpretation of the data set. A differential-field approach is presented to attenuate the airwave for the shallow-water marine CSEM data. The difference of the fields between two consecutive receivers or source points is calculated and weighted by the geometric spreading related factor. By using differential fields, the detectability given by the field ratio between the models with and without the target is enhanced, which indicates that the impact of airwaves is suppressed.

The recognition of submarine cable magnetic anomaly (SCMA) signals is a challenging task in magnetic signal data processing. In this study, a multi-task convolutional neural network (MTCNN) model is proposed to simultaneously recognize abnormal signals and locate abnormal regions. The residual block is added to the shared feature backbone to improve the ability of the network to extract high-level features and maintain the gradient stability of the model in the training process. The long short-term memory (LSTM) block is added to the classification branch task to learn the internal relationship of the magnetic anomaly time series, so as to improve the network’s ability to recognize magnetic anomalies. Our proposed model can accurately recognize the SCMA signals collected in the East China Sea and the South China Sea. The classification accuracy and the ability to locate the abnormal regions are close to the manual labeling of human analysts. The newly developed model can help analysts reduce the probability of missing and misjudging submarine cable magnetic anomalies, improve the efficiency and accuracy of interpretation, and could even be deployed to an unmanned platform to realize the automatic detection of SCMAs.

Seismic attenuation can be used to investigate the structural and compositional characteristics of the lithosphere. We studied the seismic attenuation characteristics of the continent-ocean transition zone in the Northeastern South China Sea using ocean bottom seismometer (OBS) data from the survey line OBS2016-2. A P-wave attenuation (Q(P)) transect was obtained by the Q(P) forward modeling. The results reveal a high attenuation zone in the upper crust of the continental slope that is about 40 km wide and 4 similar to 5 km thick. This low Q(P) (280 similar to 410) and the high-attenuation zone are also low in V-P (5. 5 similar to 6. 3 km . s (1)), V-S(3.1 similar to 3. 6 km . s (1)), and V-P/V-S (1. 72 similar to 1. 80), which may be related to faults and volcanic activity in this area. The high attenuation area with the Q(P) of 300 similar to 400 in the upper crust from the continent-ocean transition zone to the oceanic domain is characterized by high V-P/V-S(1. 90 similar to 1. 96), possibly related to faults and fluid migration. Moreover, the lower crust high-velocity anomaly zone in the continent-ocean transition area is characterized by relatively low Q(P)(550 similar to 600), high V-P (7. 0 similar to 7. 8 km . s (1)) and V-S (3. 5 similar to 3. 8 km . s (1)), and high V-P/V-S (1. 85 similar to 1. 96), which is related to possible serpentinization. Serpentinization can increase the porosity of rocks and lead to more fluid migration, causing high seismic wave attenuation in the ocean-continent boundary area. The Qp model, P- and S-wave velocity, and VP/Vs ratio can help us better understand the geological structure and petrophysical parameters around the continental margin, which has important reference value for further mining OBS data information.

Seismic forward modeling is fundamental in sensitivity analysis and full-waveform inversions. In finite-difference acoustic wavefield simulation, the absorption boundary, especially the perfectly matched layer (PML), is widely used, but the setting of PML parameters is empirical. An optimized complex frequency-shift PML (CFS-PML) for the modeling of acoustic fields has been developed. It refines the selection of parameters for improving the artificial attenuation. The improved CFS-PML boundary condition is applied to a 2D frequency-domain acoustic wave simulation using an optimal 17-point finite-difference scheme. Numerical results are compared with a conventional nine-point scheme in terms of computational time and physical memory consumption. These tests indicate that our CFS-PML absorption boundary can effectively improve numerical accuracy without increasing the computational burden remarkably.

The South China Sea (SCS), located at the junction of the Eurasian Plate, the Philippine Sea Plate and the IndoAustralian Plate, developed as the result of intra-continental rifting and seafloor spreading on the South China margin in the Cenozoic. Faults were developed extensively in the Pearl River Mouth Basin (PRMB), northern SCS margin, during the post-rift stage, but spatiotemporal evolution characteristics and geodynamics of the faulting were not resolved effectively. In this study, three stages of faulting evolution controlled by different geodynamics since the continental breakup were systematically analyzed and revealed based on the latest high resolution 2D/ 3D seismic data, the drilling data, the regional tectonic evolution, and the comparison with the previous studies. During Stage I, faulting showed tendency of migration from south to north (Zhu IV Depression to Baiyun Sag) and weak-intensive activities, which was mainly controlled by the southeastward mantle flow stemming from IndoEurasian collision. During Stage II, faulting showed tendency of migration from south to north (Baiyun Sag to Zhu I Depression) once again and intensive-weak activities, which was mainly controlled by the continental-ward lower crust flow stemming from the Baiyun Sag due to sedimentary loading. During Stage III, faulting showed tendency of migration from east to west and weak-intensive-weak activities, which was mainly controlled by the collision between the Luzon Arc and the Eurasian continental margin. From Stage I to Stage III, quantities of faults increased from nearly 400 to over 1000, but major lengths showed declined tendency from 3 to 15 km to 0-6 km. Striking of faults showed clockwise rotation from approximately E-W trending with azimuth of N90-100 degrees to NW-SE trending with azimuth of N110-125 degrees, and fault dip angles increased significantly from 35 to 45 degrees to 55-65 degrees and even 70-80 degrees. During Stage I and Stage II, faults were mainly developed in the extension stress field, showing roughly parallel distribution on plan view and were arranged in steps on the sections. During Stage III, faults were mainly developed in the trans-tensional stress field. Many en echelon arrangement, x-conjugatejoint and horsetail-type fault systems were developed on plan view and Y- or composite Y-type, as well as flowerlike structure fault systems on the seismic sections. Various geodynamics of the post-rift faulting in the northern SCS margin show transition from the interaction between plates to the west (the Indo-Eurasian Plates) during Stage I and Stage II, to the interaction between plates to the east (Philippine Sea Plate and the Eurasian Plate) during Stage III.

For geophysical electromagnetic (EM) forward modeling problems, the accuracy of solutions mainly depends on the numerical modeling method used and the corresponding boundary conditions. Most multi-dimensional EM studies deal with numerical methods for discretisation (e.g., finite-difference, finite-element, integral equation, etc.) and pay less attention to the boundaries. This review presents the recent research on optimizing boundary conditions for the frequency-domain marine controlled-source EM (CSEM) forward modeling algorithm. Current geophysical EM field simulation techniques usually utilize the truncated Dirichlet boundary condition, which requires the modeling domain boundaries to be far away from the area of interest and field values to be zero at the boundaries to mitigate artificial reflections/refractions resulting from truncated boundaries. The perfectly matched layer (PML) approach with few additional absorbing layers can serve as an alternative boundary to supress these truncated boundary effects. In this review, the application of the PML boundary condition to marine CSEM using a staggered finite-difference scheme for the 2.5D problem in vertical transverse isotropic (VTI) conductivity structures is introduced. This new algorithm utilizes the complex frequency-shifted PML (CFS-PML) boundary condition. The selection of optimal PML parameters are also further investigated for numerical stability. Numerical tests for several Earth conductivity models show that the CFS-PML approach is of similar high accuracy compared to using traditional Dirichlet boundary condition and exhibits additional advantages in terms of computational time and memory usage. Furthermore, the numerical tests indicate that the proposed forward modeling algorithm using CFS-PML boundary condition works well for both shallow and deep water cases, including the application to real field example from the Troll Field in Norway. The detectability of subsurface-related EM fields in airwave dominated shallow waters can be enhanced by using the weighted difference fields for mitigating the effect of airwaves on the models.

Gas hydrate is seen as a kind of new energy resources, yet it may also be one of the main greenhouse gases as its dissociation may release methane into the atmosphere. Furthermore, a severe hazard to offshore infrastructures may also be introduced by extensive gas hydrate dissociation associated with the stability of the geological structures after gas production. Therefore, it is essential to investigate the gas hydrate as well as its environmental impacts before drilling and extracting it. The geophysical seismic reflection data is usually used for exploring the gas hydrate. The gas hydrate can be effectively identified by the bottom simulating reflectors (BSRs) on seismic reflection data. However, the BSR is only for identifying the bottom boundary and it is difficult to estimate its space distribution and saturation within the hydrate stability zone. The marine controlled-source electromagnetic (CSEM) data is suitable for detecting the gas hydrate as the resistivity of the seafloor increases significantly in the presence of gas hydrate or free gas. In this study, a weighted differential-field method is applied to improve the detectivity for identifying the gas hydrate. Numerical tests show that the difference of the EM fields can effectively suppress the airwaves in shallow waters. Therefore, the detectivity given by the field ratio between the models with and without the gas hydrate target is enhanced.

The South China Sea, which is located to the southeast of Eurasian continent, developed as the result of intra-continental rifting and seafloor spreading on the South China margin. Passive margins are traditionally classi-fied as one of two end-member types, the magma-rich and magma-poor margins, based on the relative abundance or scarcity of magmatism during rifting and breakup. Previous studies suggest that the northern margin of the South China Sea is a magma-poor margin for lacking abundant magmatic activities during the breakup. A growing body of work is beginning to recognize that significant, widespread magmatism may also be present on margins that are presently considered to be magma-poor margins. In this study, we use high resolution 2D/3D seismic profiles, industrial well data, basalt geochemistry (major oxides, trace elements and isotopes) and published results to outline post-rift magmatism developed within the northern margin of the South China Sea. Four magmatic stages occurred since lithospheric breakup. The first stage, from 32 to 23.6 Ma and mainly within the distal margin of the northern South China Sea, was dominated by magma intrusions and corresponded to the spreading of the East Sub-basin. The second stage, from 23.6-19.1 Ma, was mainly in the Baiyan Sag and sur-rounding uplifts and explosive eruptions dominated. The third Stage, from 19.1-10 Ma, was east of the Zhu III Depression and west of the Enping Sag and dominated by quiet eruptions. The fourth stage, since 10 Ma, and widely distributed in the southern Dongsha uplift, experienced scattered volcanic eruptions. Two basalt samples from 23.6-19.1 Ma (industrial well HJ1, Stage 2) and seven samples from 19.1-10 Ma (industrial well EP1, Stage 3) are analyzed for major oxides, trace elements and Sr-Nd-Pb-Hf isotope compositions. Geochemical features of trace elements show that these samples are characterized by OIB-like basalts, being highly enriched in LREEs (light rare earth elements) relative to HREEs (heavy rare earth elements). Geochemical features of Sr-Nd-Pb-Hf isotope compositions show that the samples all resemble ocean-island basalts with two mixing endmembers: depleted mid-ocean ridge basalt mantle (DMM) and enriched mantle II (EMII). Pb isotopic characteristics show the Dupal isotope anomaly in the northern margin of the South China Sea. And geochemistry data of all the samples signals the contribution of Hainan Plume. Previous studies have revealed the existing of southeastward mantle flow from Tibet to South China Sea and a branch of Hainan Plume existing beneath the northern South China Sea using geophysical methods by different researchers. Based on our latest research and published geological evidences by other researchers, we propose that Stage 1 magmatism was caused by southeastward mantle flow stemming from Indo-Eurasian collision, the stage 2 and stage 3 magmatism was caused by Hainan Plume and the activation of Yangjiang-Yitongansha Faut Zone, whereas the last stage magmatic activity was mainly related with the combination of Hainan Plume and activation of the faults caused by the subduction of the SCS beneath the Luzon Arc at the Manila trench.

PreviousNext No AccessSEG 2021 Workshop: 4th International Workshop on Mathematical Geophysics: Traditional & Learning, Virtual, 17–19 December 2021Active-source ocean bottom seismometer data denoising based on multiresolution convolutional neural networksAuthors: Yutao LiuChun-Feng LiGang LiYutao LiuZhejiang UniversitySearch for more papers by this author, Chun-Feng LiZhejiang UniversityQingdao National Laboratory for Marine Science and TechnologySearch for more papers by this author, and Gang LiZhejiang UniversitySearch for more papers by this authorhttps://doi.org/10.1190/iwmg2021-05.1 SectionsAboutPDF/ePub ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InRedditEmail Abstract In order to eliminate the noise of the active-source ocean bottom seismometer (OBS) data, the multiresolution convolutional neural networks (MCNN) for denoising the seismic data is presented. The datasets are made of both field noise and synthetic records. MCNN uses noisy seismic data as the input, obtains noise through residual learning, and then outputs clean seismic data in the output layer. We compare the MCNN training results with the denoising convolutional neural networks (DnCNN) training results. The denoising results of synthetic noisy seismic records and field seismic records show that MCNN can effectively remove the noise and has a better denoising performance than DnCNN, which verifies the feasibility and advantages of MCNN denoising. Keywords: neural networks, deconvolution, mapping, noise, acquisitionPermalink: https://doi.org/10.1190/iwmg2021-05.1FiguresReferencesRelatedDetails SEG 2021 Workshop: 4th International Workshop on Mathematical Geophysics: Traditional & Learning, Virtual, 17–19 December 2021ISSN (online):2159-6832Copyright: 2022 Pages: 193 publication data© 2022 Published in electronic format with permission by the Society of Exploration GeophysicistsPublisher:Society of Exploration Geophysicists HistoryPublished: 24 Feb 2022 CITATION INFORMATION Yutao Liu, Chun-Feng Li, and Gang Li, (2022), "Active-source ocean bottom seismometer data denoising based on multiresolution convolutional neural networks," SEG Global Meeting Abstracts : 17-20. https://doi.org/10.1190/iwmg2021-05.1 Plain-Language Summary Keywordsneural networksdeconvolutionmappingnoiseacquisitionPDF DownloadLoading ...

We present three-dimensional (3-D) modeling method of marine controlled-source electromagnetic (CSEM) fields in general anisotropic media using an adaptive finite element approach based on the vector-scalar potential. The modeling is based on the governing Helmholtz equations in the vector-scalar potential system. Unstructured tetrahedral grids are employed, which can exactly simulate the terrain relief and complex electrical structures. Moreover, based on the gradient recovery technology, the adaptive finite element approach is used to drive the mesh refinement, and make the finite element solutions converge gradually to the exact solutions. The primary-secondary field approach is used to improve the numerical accuracy of CSEM fields near the source point, where the primary field is calculated by using the quasi-analytical formula. The accuracy of this approach is verified by a one-dimensional model. Two 3-D models are used to demonstrate the effectiveness of the adaptive mesh refinement and the influences of dipping anisotropy layer on the marine CSEM responses for both inline and broadside geometries. The complex synthetic model is simulated to show the capability and flexibility of the approach for geometrically complex situations.

Accurate seismic velocity can determine the depth, dip angle and location of the stratum, and study the properties of rocks and pore fluid, such as rock density, reservoir location, gas hydrate anomaly and deep tectonic characteristics. The South China Sea (SCS) has special tectonic location and complex evolution history, which has always been the focus of geologists and geophysicists. At the same time, the SCS is rich in oil/gas resources, and is one of the four major marine oil-gas accumulation zones in the world. In this paper, three-dimensional velocity modeling was carried out in northwest Pearl River Mouth Basin of the SCS based on the three-dimensional geological framework model. Minimum Curvature Interpolation was used in the three-dimensional velocity modeling. Results showed that the three-dimensional velocity model can visually display sedimentary boundary and basin basement. The geological framework model makes the lateral velocity change of strata more in line with the actual geological structure. Considering the key role of the underground layer velocity in the exploration of oil, gas, hydrate and lower-crustal structure, it will be of great scientific and resources significance to carry out the three-dimensional velocity modeling research in the SCS.