Computers & Geosciences最新文献

{"title":"An integrated neighborhood and scale information network for open-pit mine change detection in high-resolution remote sensing images","authors":"Zilin Xie ,&nbsp;Kangning Li ,&nbsp;Jinbao Jiang ,&nbsp;Jinzhong Yang ,&nbsp;Xiaojun Qiao ,&nbsp;Deshuai Yuan ,&nbsp;Cheng Nie","doi":"10.1016/j.cageo.2025.105880","DOIUrl":"10.1016/j.cageo.2025.105880","url":null,"abstract":"<div><div>The open-pit mine change detection (CD) in high-resolution remote sensing images plays a crucial role in mineral development and environmental protection. Recent advancements in deep learning have significantly promoted the open-pit mine CD. However, the existing deep-learning-based CD methods encounter challenges in effectively integrating neighborhood and scale information from high-resolution remote sensing images, resulting in insufficient performance. Therefore, according to exploration of the influence patterns of neighborhood and scale information, this paper proposed an integrated neighborhood and scale information network (INSINet) dedicated to open-pit mine CD in high-resolution remote sensing images. Specifically, INSINet introduces 8-neighborhood-image information to expend the receptive field, which improves the recognition of boundary regions in center images. Drawing on techniques of skip connection, deep supervision, and attention mechanism, the multi-path deep supervised attention module is designed to enhance multi-scale information for fusion and change feature extraction. Experimental results demonstrate that incorporating neighborhood and scale information increases the F1 score of INSINet by 6.40%, with improvements of 3.08% and 3.32% respectively. INSINet outperforms existing methods, achieving an overall accuracy of 97.69%, an intersection over union of 71.26%, and an F1-score of 83.22%. INSINet shows significance for open-pit mine CD in high-resolution remote sensing images.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"196 ","pages":"Article 105880"},"PeriodicalIF":4.2,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143093410","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Ensemble Kalman, adaptive Gaussian mixture, and particle flow filters for optimized earthquake occurrence estimation","authors":"Hamed Ali Diab-Montero ,&nbsp;Andreas S. Stordal ,&nbsp;Peter Jan van Leeuwen ,&nbsp;Femke C. Vossepoel","doi":"10.1016/j.cageo.2024.105836","DOIUrl":"10.1016/j.cageo.2024.105836","url":null,"abstract":"<div><div>Probabilistic forecasts are regarded as the highest achievable goal when predicting earthquakes, but limited information on stress, strength, and governing parameters of the seismogenic sources affects their accuracy. Ensemble data-assimilation methods, such as the Ensemble Kalman Filter (EnKF), estimate these variables by combining physics-based models and observations. While the EnKF has demonstrated potential in perfect model experiments using earthquake simulators governed by rate-and-state friction (RSF) laws, challenges arise from the non-Gaussian distribution of state variables during seismic cycle transitions. This study investigates the Adaptive Gaussian Mixture Filter (AGMF) and the Particle Flow Filter (PFF) as alternatives for improved stress and velocity estimation in earthquake sequences compared to Gaussian-based methods like the EnKF. We test the AGMF and the PFF’s performance using Lorenz 96 and Burridge–Knopoff 1D models which are, respectively, standard simplified atmospheric and earthquake models. This approach, using widely recognized and commonly used testbed models in their fields, makes the methods and findings accessible to both the data assimilation and seismology communities, while supporting comparisons and collaboration. We test these models in periodic, and aperiodic conditions, and analyze the impact of assuming Gaussian priors on the estimates of the ensemble methods. The PFF demonstrated comparable performance in chaotic scenarios, yielding lower RMSE for the estimates of the Lorenz 96 models and stronger resilience to underdispersion for the Burridge–Knopoff 1D models. This is vital given the limited and sparse historical earthquake data, underscoring the PFF’s potential in enhancing earthquake forecasting. These results emphasize the need for careful data assimilation method selection in seismological modeling.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"196 ","pages":"Article 105836"},"PeriodicalIF":4.2,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143093411","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SeisSegDiff: A label-efficient few-shot texture segmentation diffusion model for seismic facies classification","authors":"Tobi Ore,&nbsp;Dengliang Gao","doi":"10.1016/j.cageo.2024.105823","DOIUrl":"10.1016/j.cageo.2024.105823","url":null,"abstract":"<div><div>Traditional seismic facies analysis, which depends on manual interpretation of seimic amplitude, encounters difficulties because of the complexity, volume, and limited resolution of the seismic data. To tackle these problems, seismic texture based deep learning has emerged as a highly promising technique. However, the reliance on extensive labeled datasets poses a significant hurdle. Here we introduce SeisSegDiff, an innovative approach to seismic texture classification that combines diffusion probabilistic models with deep learning models. We use diffusion models to enhance the generalization capabilities and accuracy of deep learning models for seismic facies segmentation. The proposed method utilizes the feature maps obtained from the intermediate layers of a Unet denoising network that estimates the Markov phase of the reverse diffusion process, serving as seismic image representations. These feature maps serve as high-level semantic information for addressing deep learning seismic interpretation challenges, hence reducing the requirement for large, labeled datasets. We assess the effectiveness of SeisSegDiff by conducting experiments on two seismic benchmark datasets from the Netherlands F3 and the Parihaka Basin. The results demonstrate the exceptional performance of our method in defining subsurface facies boundaries and structures. The ability of SeisSegDiff to operate with minimal labeled datasets (∼&lt;1% inlines) further emphasizes its potential for practical field deployments. Our work will draw the geophysical deep learning community closer to the goal of creating a unified global seismic texture model for automatic seismic interpretation and cost effective subsurafe characterization.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"196 ","pages":"Article 105823"},"PeriodicalIF":4.2,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143093821","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluating key parameters impacting the performance of Seis Seg Diff model for seismic facies classification","authors":"Tobi Ore,&nbsp;Dengliang Gao","doi":"10.1016/j.cageo.2024.105829","DOIUrl":"10.1016/j.cageo.2024.105829","url":null,"abstract":"<div><div>Facies are a body of rock that is distinct from adjacent rock units based on observable characteristics such as composition and texture. They are sought out in subsurface characterization tasks because of the valuable information they provide about past environments and geological processes. In seismic data, facies express distinct reflection patterns and are traditionally interpreted manually using seismic attributes. However, manual interpretation is typically time-consuming and biased by the interpreter. Automatic interpretation methods that capitalize on the predictive ability of deep learning have been proposed with relative success. However, these methods are data-intensive with practical deployment limitations. SeisSegDiff is a novel model that draws from the representations learned by diffusion models to classify the facies accurately with limited training data. In this paper, we investigate the quality of the representations learned by the diffusion model and the impact of the model hyperparameters on its performance. We found that for a diffusion denoising encoder-decoder network, the middle decoder blocks [5–13] at the later time steps of the diffusion process [0–250] had the most informative representations for the facies discrimination. For the few shot capability, the model had a mIoU of 0.75 when it was trained with only 3 inlines and its performance consequently increased for more training cross sections with 0.83 when trained with 5 inlines and crosslines, outperforming the state-of-the-art with only ∼2% training data. Furthermore, we found that the model is robust in the presence of faults but struggles with regions with complex salt structures. Our results demonstrate that well designed SeisSegDiff model parameters can greatly speed up subsurface characterization tasks in practical field settings with real seismic and well data. We anticipate the model to be a starting point for more sophisticated applications of the diffusion model for geophysical data interpretation and processing. For example, the learned representations from the diffusion model can lend themselves to the development of a global reservoir property inversion model.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"196 ","pages":"Article 105829"},"PeriodicalIF":4.2,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143093822","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Siamese topographic generation model: A deep learning model for generating Antarctic subglacial topography with fine details","authors":"Yiheng Cai ,&nbsp;Yanliang He ,&nbsp;Shinan Lang ,&nbsp;Xiangbin Cui ,&nbsp;Xiaoqing Zhang ,&nbsp;Zijun Yao","doi":"10.1016/j.cageo.2025.105857","DOIUrl":"10.1016/j.cageo.2025.105857","url":null,"abstract":"<div><div>The ongoing accumulation of radio-echo sounding (RES) measurements in Antarctica in recent years has significantly expanded our understanding of subglacial structures. The effective use of RES-collected data to construct accurate Antarctic subglacial topography has emerged as a vital component of contemporary polar research. Various methods, including conventional interpolation, inversion techniques, and even deep learning methods, have been used to recreate Antarctic bed topography. However, these bed topographies are often plagued by over-smoothing, loss of small-scale features, low precision, and instability.</div><div>The Siamese topographic generation model (STGM) is proposed here to address the above mentioned issues. After being trained on ArcticDEM, this model can generate Antarctic subglacial topography with stability and accuracy by merging the advantages of deep learning-based generative models, Siamese networks, kernel prediction, and deformable convolutions. In terms of evaluation, both quantitative and qualitative comparisons with current Antarctic subglacial digital elevation models demonstrate that our method can generate topographical features, such as mountains, ice streams, and valleys, with high precision and minimal artifacts. In quantitative validation, our model achieves over 20% improvement in both Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) compared to the previously best-performing method (GEI), surpassing existing models in terms of accuracy and detail.</div><div>Moreover, an error analysis specifically focusing on the effect of varying track intervals has been conducted, offering a benchmark for future investigations into the influence of track density on model errors. Finally, using STGM based on the RES data, the subglacial topography of Princess Elizabeth Land has also been successfully generated. In this area, the topography generated by STGM at a resolution of 500 m clearly depicts subglacial lakes and valleys, revealing the complexity and diversity of the subglacial topography.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"196 ","pages":"Article 105857"},"PeriodicalIF":4.2,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143102345","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Time reversal imaging and transfer learning for spatial and temporal seismic source location","authors":"Anna Franczyk,&nbsp;Damian Gwiżdż","doi":"10.1016/j.cageo.2024.105843","DOIUrl":"10.1016/j.cageo.2024.105843","url":null,"abstract":"<div><div>This article presents the application of Time Reversal Imaging (TRI) and transfer learning methods for spatial and temporal seismic wave location. The study applies the ResNet-50 model, pre-trained on the basis of ImageNet images, and later retrained using seismic wave field component images. The objective of the study was to provide an accurate classification of seismic source areas and determine the temporal localization of seismic events.</div><div>The research involved training the ResNet-50 model based on datasets of wave field component images obtained through the backpropagation of reversed waveforms in simplified geological models. The classification was evaluated using performance metrics. Additionally, to assess its effectiveness in realistic scenarios the methodology was applied to the complex Marmousi velocity model.</div><div>The results show that the combined TRI and transfer learning approach is highly effective in classifying seismic source areas. The trained model successfully identifies patterns unique to seismic wave components, enabling precise spatial localization. Additionally, the method accurately determines the focusing time, which is essential for the temporal localization of seismic events. The article includes research on the influence of receiver network geometry on localization outcomes. By examining various receiver configurations, valuable insights have been gained, further improving the practical applicability of the method.</div><div>The study highlights the potential for further advances by extending the methodology to three-dimensional models, although there remains a need to address various computational challenges. Three-dimensional modeling would enhance the accuracy of source localization, especially in the case of seismic sources characterized by dominant isotropic components.</div><div>In conclusion, the combination of TRI and transfer learning presents a promising approach for ensuring precise spatial and temporal seismic wave location. This methodology has the potential to enhance seismic monitoring, early warning systems, and make a significant contribution to earthquake engineering and hazard assessment.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"196 ","pages":"Article 105843"},"PeriodicalIF":4.2,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143093101","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"TranSeis: A high precision multitask seismic waveform detector","authors":"Yuxin Zhou ,&nbsp;Huai Zhang ,&nbsp;Shi Chen ,&nbsp;Zheng Yuan ,&nbsp;Chuanqi Tan ,&nbsp;Fei Huang ,&nbsp;Yicun Guo ,&nbsp;Yaolin Shi","doi":"10.1016/j.cageo.2025.105867","DOIUrl":"10.1016/j.cageo.2025.105867","url":null,"abstract":"<div><div>This work introduces a highly efficient multitask parallel Artificial Intelligence model designed for weak seismic signal detection and phase picking, leveraging the capabilities of the conventional AI-powered Transformer architecture. By integrating a multi-part data extraction strategy, a multi-GPU parallel processing framework, and a multi-layer network schedule, we significantly enhance the accuracy of detecting P- and S-phases while optimizing the model's efficiency. The accuracy attained for the P and S phases was 92% and 76% when employing only a segment of the dataset. When we incorporated the entire dataset, the precision improved to 97% for P phases and 87% for S phases. Notably, our model demonstrates higher accuracy compared to existing deep-learning and traditional detection algorithms. When applied to extensive seismic phase observation data collected from 2020 to 2023 in mainland China, our model consistently demonstrated high accuracy, confirming its generalizability across various spatiotemporal contexts. It also exhibited exceptional sensitivity to subtle changes in waveform data, highlighting its promising potential for detecting smaller seismic events with even greater resolution in future applications.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"196 ","pages":"Article 105867"},"PeriodicalIF":4.2,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143093407","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Non-linear non-parametric geostatistical rock-physics inversion of elastic attributes for petrophysical properties using direct multivariate simulation","authors":"Leandro Passos de Figueiredo ,&nbsp;Dario Grana ,&nbsp;Bruno B. Rodrigues ,&nbsp;Alexandre Emerick","doi":"10.1016/j.cageo.2025.105862","DOIUrl":"10.1016/j.cageo.2025.105862","url":null,"abstract":"<div><div>The estimation of subsurface petrophysical properties plays an essential role in the reservoir characterization and forecasting process. In this work, we present a novel algorithm for geostatistical rock-physics inversion of elastic properties that assumes a non-linear forward model and a non-parametric multivariate joint distribution. The inversion method is based on the numerical solution for data conditioning of the joint probability distribution and it combines statistical rock-physics models and stepwise conditional transformations applied to non-parametric geostatistical simulations. Specifically, we apply a data conditioning approach of Direct Multivariate Simulation to obtain the petrophysical properties conditioned to the measured elastic properties. The approach can be applied to estimate median models or to simulate multiple geostatistical realizations conditioned on direct measurements. We validate the approach through two applications: a 1D study using real well logs for the estimation of petrophysical volumetric fractions using a 6-variate joint distribution and a synthetic time-lapse seismic study for the estimation of porosity and fluid changes using a 7-variate joint distribution. We discuss the computational advantages of the proposed implementation in terms of computational time and RAM usage. The efficient implementation makes this method applicable to high-dimensional problems. The algorithm effectively preserves the non-linear and heteroscedastic relationships among variables, providing accurate estimations of petrophysical properties while maintaining spatial correlations and incorporating hard data conditioning.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"196 ","pages":"Article 105862"},"PeriodicalIF":4.2,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143093488","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automatic detection of ditches and natural streams from digital elevation models using deep learning","authors":"Mariana Dos Santos Toledo Busarello ,&nbsp;Anneli M. Ågren ,&nbsp;Florian Westphal ,&nbsp;William Lidberg","doi":"10.1016/j.cageo.2025.105875","DOIUrl":"10.1016/j.cageo.2025.105875","url":null,"abstract":"<div><div>Policies focused on waterbody protection and restoration have been suggested to European Union member countries for some time, but to adopt these policies on a large scale the quality of small water channel maps needs considerable improvement. We developed methods to detect and classify small stream and ditch channels using airborne laser scanning and deep learning. The research questions covered the influence of the resolution of the digital elevation model on channel extraction, the efficacy of different terrain indices to identify channels, the potential advantages of combining indices, and the performance of a U-net model in mapping both ditches and stream channels. Models trained in finer resolutions were more accurate than models trained with coarser resolutions. No single terrain index consistently outperformed all others, but some combinations of indices had higher MCC values. Natural stream channels were not classified to the same extent as ditches. The model trained on the 0.5 m resolution had the most balanced performance using a combination of indices trained using the dataset with both types of channel separately. The deep learning model outperformed traditional mapping methods for ditches, increasing the recall from less than 10% to over 92%, while the recall for natural channels was around 71%. However, despite the successful detection of ditches, the models frequently misclassified streams as ditches. This poses a challenge, as natural channels are protected under land use management practices, while ditches are not.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"196 ","pages":"Article 105875"},"PeriodicalIF":4.2,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143093489","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"tonus: Detection, characterization and cataloguing of seismo-volcanic tonal signals","authors":"Leonardo van der Laat ,&nbsp;Mauricio M. Mora ,&nbsp;Javier Fco. Pacheco ,&nbsp;Philippe Lesage ,&nbsp;Esteban Meneses","doi":"10.1016/j.cageo.2024.105791","DOIUrl":"10.1016/j.cageo.2024.105791","url":null,"abstract":"<div><div>Observational seismology plays a crucial role in volcano monitoring programs. It enables the detection and understanding of various volcanic processes. Among the variety of seismic signatures, tonal coda and harmonic tremor stand out. They showcase at least one prominent spectral peak and appear at various phases of volcanic activity, during late stages of pre-eruptive periods and eruptions. Previous studies have shown that the analysis of these signals can, not only enhance the understanding of volcanic processes, but potentially contribute to eruption forecasting. This research introduces <em>tonus</em>, a software tool designed to detect, analyze, and catalogue tonal events in a volcano observatory context. The tool provides user-friendly graphical interfaces that facilitate data visualization and analysis, parameters adjustment, and querying of a standardized database. Developed using open-source and cross-platform systems, tonus uniquely detects and systematically catalogs relevant characteristics of tonal coda and harmonic tremor events. The detection algorithm, tested with pre-eruptive data from Turrialba volcano in April 2016, achieved 95% precision and 80% recall. The occurrence of thousands of tonal events in Costa Rican volcanoes inspired the development of this software, providing us with the ability to rapidly process tonal seismicity. Over the last three years, the use of this software enabled the identification of surges in tonal coda events, characterized by decreasing spectral frequencies, preceding eruptive activities at both Turrialba and Rincón de la Vieja volcanoes. tonus represents a significant contribution to volcano seismology research and monitoring, successfully bridging a gap between academic methodologies and practical observatory applications.</div></div>","PeriodicalId":55221,"journal":{"name":"Computers & Geosciences","volume":"196 ","pages":"Article 105791"},"PeriodicalIF":4.2,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143102328","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}