Computers & Geosciences最新文献

Single-sided magnetic resonance (SSMR) offers advantages of portability and noninvasive measurement for water detection, with significant potential applications in groundwater exploration, petroleum well logging, and soil moisture monitoring. However, the inherent highly inhomogeneous static magnetic field and radiofrequency (RF) field in SSMR necessitate the utilization of the Carr–Purcell–Meiboom–Gill (CPMG) sequence measurement scheme. To accelerate forward modeling during pulse excitation, we introduce a Geometric Analysis Method (GAM) and assess $T_2$ error using its primary parameters. The GAM involves applying spatial geometric rotations on the magnetization vector, leading to an analytical solution to the Bloch equation that disregards relaxation effects. Compared with the rotation matrix (RM) method, the GAM demonstrates high accuracy and reduces computational time by approximately 20.9%. By analyzing the primary parameters governing the magnetization vector in the analytical formula, we evaluated their impact on the transverse relaxation time ($T_2$) obtained through fitting the SE signal. Ultimately, the forward modeling results of the CPMG sequence within the region of interest (ROI) of a single-sided Halbach magnet array are validated. The $T_2$ fitting error increases as the primary parameters deviate from the ideal values, highlighting their significant role in the $T_2$ fitting results. This study provides a theoretical foundation for optimizing the design of SSMR magnets and RF coils.

The accurate taxonomic identification of fusulinid fossils holds significant scientific value in palaeontology, paleoecology, and palaeogeography. However, imbalanced image samples lead to the model preferring to learn features from categories with many samples while ignoring fewer sample categories, greatly reducing the prediction accuracy of fusulinid fossil identification. Moreover, the textual description of fusulinid fossils contains rich feature information. We collected and created an order fusulinid multimodal (OFM) dataset for research. We proposed a transformer-based multimodal integration framework (TMIF) using deep learning for fusulinid fossil identification. Compared to traditional neural networks, the transformer can create global dependencies between features at different locations. TMIF incorporates image and text branches dedicated to extracting features for both modalities, and a pivotal cross-modal integration module that allows visual features to learn textual semantic features sufficiently to obtain a more comprehensive feature representation. Experimental evaluation using the OFM dataset shows that TMIF achieves a prediction accuracy of 81.7%, which is a 2.8% improvement over the only image-based method. Further comparative analyses across multiple networks affirm that the TMIF performs optimally in addressing the taxonomic identification of fusulinid fossils with imbalanced samples.

A new CNN based approach supported by semantic segmentation, was proposed. This approach is frequently used to carry out regional-scale studies. The core of our method revolves around a CNN model, based on the famous U-Net architecture. Its purpose is to identify different classes of pixels on satellite images and later to automatically detect the coastline. The recently launched Coast Train dataset was used to train the CNN model. Traditional coastline detection was improved (“water/land” segmentation) by means of two new aspects the use of the Sobel-edge loss function and the segmentation of the satellite images into several categories like built-up areas, vegetation and land besides beach/sand and water classes. The approach used ensures a more precise coastline extraction, distinguishing water pixels from all other categories. Our model adeptly identifies features, such as cliff vegetation or coastal roads, that some models might overlook. In this way, coastline localization and its drawing for regional scale study, have minor uncertainties. The performance of the CNN-based method, achieving 85% accuracy and 80% IoU (Intersection over Union) in the segmentation process. The ability of the model to extract the coastline was validated on a Sicilian case study, notably the San Leone beach (Agrigento). The model's results align closely with the ground truth, moreover, its reliability was further confirmed when it was tested on other Sicilian coastal regions.

Beyond robustness, the model offers a promising avenue for enhanced coastal analysis potentially applicable to coastal planning and management.

Seismology has witnessed significant advancements in recent years with the application of deep learning methods to address a broad range of problems. These techniques have demonstrated their remarkable ability to effectively extract statistical properties from extensive datasets, surpassing the capabilities of traditional approaches to an extent. In this study, we present SAIPy, an open-source Python package specifically developed for fast seismic data processing by implementing deep learning. SAIPy offers solutions for multiple seismological tasks, including earthquake signal detection, seismic phase picking, first motion polarity identification and magnitude estimation. We introduce upgraded versions of previously published models such as CREIME_RT capable of identifying earthquakes with an accuracy above 99.8% and a root mean squared error of 0.38 unit in magnitude estimation. These upgraded models outperform state-of-the-art approaches like the Vision Transformer network. SAIPy provides an API that simplifies the integration of these advanced models, including CREIME_RT, DynaPicker_v2, and PolarCAP, along with benchmark datasets. It also, to the best of our knowledge, introduces the first fully automated deep learning based pipeline to process continuous waveforms. The package has the potential to be used for real-time earthquake monitoring to enable timely actions to mitigate the impact of seismic events. Ongoing development efforts aim to further enhance SAIPy’s performance and incorporate additional features that enhance exploration efforts, and it also would be interesting to approach the retraining of the whole package as a multi-task learning problem. A detailed description of all functions is available in a supplementary document.

Reconstructing geochemical data for anomaly detection using Generative Adversarial Networks (GANs) has become a prevalent method in identifying geochemical anomalies. However, injecting random noise into GANs can induce model instability. To mitigate this issue, we propose a novel anomaly detection model, Geo-Hgan, which integrates a dual adversarial network architecture with a Latent Space Adversarial Module (LSAM) to learn the distribution of latent variables from arbitrary data and optimize the sample reconstruction process, thereby alleviating instability during GAN training. Additionally, an encoder guided by the LSAM-pretrained GAN is employed to extract variational features, facilitating rapid and effective sample mapping into the latent space defined by LSAM. Experimental results demonstrate that under unsupervised conditions, Geo-Hgan achieves an Area Under the Curve (AUC) score of 85% across three geochemical datasets, outperforming similar models in accuracy and reconstruction capabilities. To assess its versatility and generalization ability, we extend Geo-Hgan to anomaly detection tasks in computer vision, where it achieves an average AUC score of 98.7% on the MvtecAD dataset, setting a new state-of-the-art performance in the domain. Furthermore, we propose AnomFilter, a method for setting anomaly thresholds based on the clustering hypothesis. AnomFilter identifies high-confidence anomaly samples identified by Geo-Hgan in the source domain and iteratively transfers them to the target domain. These high-confidence anomaly samples, combined with a small number of known positive samples in the target domain, enhance the accuracy of supervised geochemical anomaly detection in the target domain, which achieved an AUC score of 94%. The utilization of anomaly detection models for sample transfer learning offers a novel perspective for future work.

Fault dating plays an essential role in understanding deformation histories and modeling the tectonic evolution of orogenic belts. However, direct fault dating methods via different isotope geochronological techniques are expensive, and their use is often limited in many cases, making it essential to develop a fast and low-cost fault relative dating method. Therefore, on the basis of knowledge graphs and knowledge reasoning technology, this study proposes an automatic method to relatively date periods of fault activity using the principle of cross-cutting relationships between faults and strata. The method mainly involves (1) generating the knowledge graph based on a digital geological map; (2) using the knowledge reasoning algorithm to interpret the cross-cutting relationships amongst faults and generating the temporal sequence of fault activity; (3) relative dating the faults based on the cross-cutting relationships between faults and strata; and (4) according to the temporal sequence of fault activity, the relationship between faults can be revealed, and relative dating can be optimized. Results for cases in western Nevada and Qixia Hill of Nanjing illustrate the effectiveness of this method for interpreting the period of fault activity. The accuracy rates of the recognition results in the two cases were 90.24% and 80.77%, respectively, which means that the proposed method has the potential to relatively date fault activity across large areas. The algorithm is an effective supplement to the existing direct method of fault dating. The algorithm can efficiently infer the development sequence and the age of fault activity based on geological maps and geological cross-sections, which is of great significance for understanding regional tectonic history, evaluating earthquake disasters, and modeling tectonic evolution processes.

The water content in the Earth's interior is of great significance for material circulation and the dynamic evolution of the planet. The water content in mantle minerals significantly affects their conductivities. By measuring the variations in conductivity within the Earth, we can infer the water content in the mantle and study the movement and processes of materials within the Earth. The geomagnetic depth sounding is a widely used method for imaging the mantle conductivity as it has large sounding depth. However, the ocean induction effects can seriously impact geomagnetic data that can't be well corrected using conventional methods. Here, we present a novel three-dimensional inversion method for geomagnetic depth sounding to overcome the ocean induction effects by directly adopt the real earth surface conductivity into the inverse model. In this method, the unstructured tetrahedral grids are used to represent the model in multi-scale and the vector finite-element method is adopted to accurately compute the geomagnetic responses. The synthetic model tests show that the earth surface conductivity has serious effect on the inversion results, but it can be well suppressed by directly modeling it in the inverse model. We further invert the data from 128 geomagnetic stations around the world and obtain a more accurate new model of global mantle conductivity.

Modeling gas flow through fractures of subsurface rock is a particularly challenging problem because of the heterogeneous nature of the material. High-fidelity simulations using discrete fracture network (DFN) models are one methodology for predicting gas particle breakthrough times at the surface but are computationally demanding. We propose a Bayesian machine learning method that serves as an efficient surrogate model, or emulator, for these three-dimensional DFN simulations. Our model trains on a small quantity of simulation data with given statistical properties and, using a graph/path-based decomposition of the fracture network, rapidly predicts quantiles of the breakthrough time distribution on DFNs with those statistical properties. The approach, based on Gaussian Process Regression (GPR), outputs predictions that are within 20%–30% of high-fidelity DFN simulation results. Unlike previously proposed methods, it also provides uncertainty quantification, outputting confidence intervals that are essential given the uncertainty inherent in subsurface modeling. Our trained model runs within a fraction of a second, considerably faster than reduced-order models yielding comparable accuracy (Hyman et al., 2017; Karra et al., 2018) and multiple orders of magnitude faster than high-fidelity simulations.

Directional drilling of longer drillholes is becoming increasingly important as resources are exploited at greater depths. As drillholes lengthen, the choice of desurveying method becomes more crucial as the assumptions that are inherent to all methods are compounded. The aim of this study is to first discuss the assumptions involved in each desurveying method and their potential implications for plotting drillhole pathways, and secondly to compare the established desurveying methods to find the most precise one for plotting the drillhole pathway, using examples from Mount Isa, Australia.

The orientations (azimuth and plunge) of drillholes are required to orient drill core (also known as rock or well core), which can be used to measure the orientations of geological structures at any point. Knowledge of the 3D positions for points of interest along the drill core are required to locate drillhole intersections with geological boundaries, faults or underground mine workings. New computer code has been developed to estimate the orientations and positions of drillholes at any point along their length using the existing desurveying methods. Such orientation and location estimates from the computer codes allow the original orientations of geological structures observed in drill core to be calculated. The codes are available in both R and Python languages in an easy access repository. Results from the codes show that the Basic Tangent method is consistently the least precise, whilst the industry standard Minimum Curvature method has a high precision compared to the other desurveying methods. The impact of rock anisotropy and drillhole length on the precision of the desurveying methods was investigated. Distances between end-of-hole points for each desurveying method increase with increasing drillhole length and angle between the drillhole and anisotropy.

We propose an automatic global search algorithm based on the Variable Neighborhood Search (VNS) metaheuristic for tuning the hyperparameters of a generative adversarial network (GAN) seismic interpolator. We perform an exhaustive search to study the influence of each hyperparameter in the training process, and compare the proposed method with Random search and Bayesian Search. The seismic data set used for this study was synthetically modeled from a typical velocity model, estimated from a pre-salt field of the Brazilian cost. We also employ the proposed method with a real field data to show the importance and applicability of the search for optimum hyperparameters of GAN. The training data was constructed with decimated seismic data and the results were tested by comparing the reconstructed data with the original one. We performed two hyperparameter impact analyses: the first consists of an exhaustive grid exploration and the second consists of our proposed automatic exploration method using the VNS algorithm, comparing it with the other two algorithms. We concluded that the proposed method, which has a user-friendly usage, as it is almost parameter-free, can reach solutions with very good quality quickly, in any range of hyperparameter values. When compared with other methods of hyperparameter tuning, the one we propose proves to be better in the ease of configuration, while being efficient in the search process.