Applied Computing and Geosciences最新文献_第7页

Sinkhole susceptibility analysis using machine learning for west central Florida 利用机器学习对佛罗里达州中西部进行地陷敏感性分析

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2025-06-26 DOI: 10.1016/j.acags.2025.100262

Olanrewaju Muili, Hassan A. Babaie

This study examined the feasibility and accuracy of applying machine learning for sinkhole classification and prediction and using the results in automated sinkhole susceptibility mapping for west central Florida. A two-stage processing pipeline was developed. In the first stage, we assessed the predictive power of five exemplary machine learning algorithms: random forest (RF), logistic regression (LR), k-nearest neighbor (KNN), support vector machine (SVM), and multilayer perceptron (MLP), and select the best-performing model. The top-performed model was then chosen to develop a sinkhole susceptibility map (SSM) in the second step of the process. Nine feature layers were derived from the collected geospatial data and utilized as conditional variables. Several statistical metrics and receiver operating characteristic curves were utilized to evaluate the accuracy of the models. The results showed that the RF model, with a ROC of 0.984, had the highest prediction capability in the research area.

We generated a susceptibility map using the RF model, and the study area was classified into high susceptibility (H) and low susceptibility (L) areas. Confusion Matrix (CM) and Matthews Correlation Coefficient (MCC) were used to confirm the results of the sinkhole susceptibility map's classification. We present a model that predicts sinkhole distribution in the study area, and the output of our model is consistent with the sinkhole hazard map that the Florida Division of Emergency Management had previously created. This work can assist the government, community, and land managers in creating plans for mitigating hazards and land degradation.

本研究考察了应用机器学习进行天坑分类和预测的可行性和准确性，并将结果用于佛罗里达州中西部的自动天坑敏感性测绘。开发了两阶段加工流水线。在第一阶段，我们评估了五种典型机器学习算法的预测能力：随机森林（RF）、逻辑回归（LR）、k近邻（KNN）、支持向量机（SVM）和多层感知器（MLP），并选择了表现最佳的模型。然后，在该过程的第二步中，选择表现最好的模型来开发天坑敏感性图（SSM）。从收集到的地理空间数据中得到9个特征层，并将其作为条件变量。利用一些统计指标和受试者工作特征曲线来评估模型的准确性。结果表明，该模型预测能力最强，ROC值为0.984。利用RF模型绘制了敏感性图，并将研究区划分为高敏感性区(H)和低敏感性区(L)。利用混淆矩阵（Confusion Matrix， CM）和Matthews相关系数（Matthews Correlation Coefficient， MCC）对地陷敏感性图的分类结果进行了验证。我们提出了一个预测研究区域天坑分布的模型，我们模型的输出与佛罗里达州应急管理部门先前创建的天坑危害图一致。这项工作可以帮助政府、社区和土地管理者制定减轻灾害和土地退化的计划。

{"title":"Sinkhole susceptibility analysis using machine learning for west central Florida","authors":"Olanrewaju Muili, Hassan A. Babaie","doi":"10.1016/j.acags.2025.100262","DOIUrl":"10.1016/j.acags.2025.100262","url":null,"abstract":"<div><div>This study examined the feasibility and accuracy of applying machine learning for sinkhole classification and prediction and using the results in automated sinkhole susceptibility mapping for west central Florida. A two-stage processing pipeline was developed. In the first stage, we assessed the predictive power of five exemplary machine learning algorithms: random forest (RF), logistic regression (LR), k-nearest neighbor (KNN), support vector machine (SVM), and multilayer perceptron (MLP), and select the best-performing model. The top-performed model was then chosen to develop a sinkhole susceptibility map (SSM) in the second step of the process. Nine feature layers were derived from the collected geospatial data and utilized as conditional variables. Several statistical metrics and receiver operating characteristic curves were utilized to evaluate the accuracy of the models. The results showed that the RF model, with a ROC of 0.984, had the highest prediction capability in the research area.</div><div>We generated a susceptibility map using the RF model, and the study area was classified into high susceptibility (H) and low susceptibility (L) areas. Confusion Matrix (CM) and Matthews Correlation Coefficient (MCC) were used to confirm the results of the sinkhole susceptibility map's classification. We present a model that predicts sinkhole distribution in the study area, and the output of our model is consistent with the sinkhole hazard map that the Florida Division of Emergency Management had previously created. This work can assist the government, community, and land managers in creating plans for mitigating hazards and land degradation.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"27 ","pages":"Article 100262"},"PeriodicalIF":2.6,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144514348","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Advanced identification of geological discontinuities with deep learning 基于深度学习的地质不连续面高级识别

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2025-06-21 DOI: 10.1016/j.acags.2025.100256

Rushan Wang , Martin Ziegler , Michele Volpi , Andrea Manconi

Rock mass characterization is essential for various applications in geosciences. Traditional methods, such as manual mapping and interpretation, are labor-intensive and prone to inconsistencies. Although machine learning has advanced in many fields, its application in structural geology, especially for distinguishing different discontinuity types, remains limited. This study presents a deep learning-based approach for identifying geological discontinuities in borehole images, classifying features such as intact walls, induced cracks, and tectonic fault planes, among others. We evaluate deep learning architectures, including standard Convolutional Neural Networks and Transformer-based models, and optimize segmentation performance with multi-scale training, tiling strategies, and tailored loss functions. Our results demonstrate that the Transformer model, particularly SegFormer, outperforms U-Net in detecting complex geological features. The combined use of weighted cross-entropy and focal loss further improves model robustness, especially for underrepresented and challenging features. In addition, the choice of the tiling size significantly affects the classification performance of different geological features. This research establishes an efficient and accurate pipeline for automated geological interpretation, with significant implications for subsurface exploration and geotechnical engineering.

岩体表征在地球科学的各种应用中是必不可少的。传统的方法，如手工映射和解释，是劳动密集型的，并且容易产生不一致。尽管机器学习在许多领域都取得了进展，但它在构造地质学中的应用，特别是在区分不同的不连续类型方面的应用仍然有限。本研究提出了一种基于深度学习的方法，用于识别钻孔图像中的地质不连续性，对完整壁、诱导裂缝和构造断裂面等特征进行分类。我们评估了深度学习架构，包括标准卷积神经网络和基于transformer的模型，并通过多尺度训练、平铺策略和定制损失函数优化分割性能。我们的研究结果表明，Transformer模型，特别是SegFormer，在检测复杂地质特征方面优于U-Net。加权交叉熵和焦点损失的结合使用进一步提高了模型的鲁棒性，特别是对于代表性不足和具有挑战性的特征。此外，瓦片尺寸的选择显著影响不同地质特征的分类性能。本研究建立了一条高效、准确的自动化地质解释管道，对地下勘探和岩土工程具有重要意义。

{"title":"Advanced identification of geological discontinuities with deep learning","authors":"Rushan Wang , Martin Ziegler , Michele Volpi , Andrea Manconi","doi":"10.1016/j.acags.2025.100256","DOIUrl":"10.1016/j.acags.2025.100256","url":null,"abstract":"<div><div>Rock mass characterization is essential for various applications in geosciences. Traditional methods, such as manual mapping and interpretation, are labor-intensive and prone to inconsistencies. Although machine learning has advanced in many fields, its application in structural geology, especially for distinguishing different discontinuity types, remains limited. This study presents a deep learning-based approach for identifying geological discontinuities in borehole images, classifying features such as intact walls, induced cracks, and tectonic fault planes, among others. We evaluate deep learning architectures, including standard Convolutional Neural Networks and Transformer-based models, and optimize segmentation performance with multi-scale training, tiling strategies, and tailored loss functions. Our results demonstrate that the Transformer model, particularly SegFormer, outperforms U-Net in detecting complex geological features. The combined use of weighted cross-entropy and focal loss further improves model robustness, especially for underrepresented and challenging features. In addition, the choice of the tiling size significantly affects the classification performance of different geological features. This research establishes an efficient and accurate pipeline for automated geological interpretation, with significant implications for subsurface exploration and geotechnical engineering.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"27 ","pages":"Article 100256"},"PeriodicalIF":2.6,"publicationDate":"2025-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144514346","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Integrating neuro-symbolic AI and knowledge graph for enhanced geochemical prediction in copper deposits 结合神经符号人工智能和知识图谱增强铜矿地球化学预测

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2025-06-18 DOI: 10.1016/j.acags.2025.100259

Weilin Chen, Jiyin Zhang, Wenjia Li, Xiang Que, Chenhao Li, Xiaogang Ma

The integration of machine learning (ML) and deep learning (DL) in geoscience has demonstrated great promise for mineral prediction. However, existing approaches are predominantly data-driven and often overlook expert geological knowledge, limiting their interpretability, accuracy, and practical applicability. This study introduces a new method that combines Large Language Models (LLMs), knowledge graphs (KGs), and Neuro-Symbolic AI (NSAI) models to predict mineralization systems in diverse copper deposits, significantly increasing the precision in prediction results. We utilize LLMs to generate KGs from geological literature, extracting symbolic rules that encode domain-specific insights about copper mineralization. These rules, derived dynamically from expert knowledge, are integrated into ML models as guidance during the training and prediction phases. By fusing symbolic reasoning with ML's computational power, our approach overcomes the limitations of black-box models, offering both improved accuracy and transparency in mineral prediction. To validate this method, we apply it to a comprehensive geochemical dataset of global copper deposits. The results show that rule-guided ML models achieve notable performance improvements, outperforming traditional ML methods in accuracy, precision, and robustness. Interpretability is further enhanced by using tools such as SHAP values, which explain the influence of individual geochemical features within the rule-based framework. This combination not only identifies critical geochemical elements like Cu, Fe, and S but also provides coherent, domain-aligned explanations for the predicted mineralization patterns. Our findings demonstrate the transformative potential of combining LLMs, KGs, and ML models for mineral prediction. This hybrid approach enables geoscientists to leverage both computational and expert knowledge, achieving a deeper understanding of mineralization systems.

地球科学中机器学习（ML）和深度学习（DL）的整合在矿物预测方面显示出巨大的前景。然而，现有的方法主要是数据驱动的，往往忽略了专家地质知识，限制了它们的可解释性、准确性和实际适用性。本研究提出了一种结合大语言模型（LLMs）、知识图（KGs）和神经符号人工智能（NSAI）模型的新方法，用于预测不同铜矿床的矿化系统，显著提高了预测结果的精度。我们利用llm从地质文献中生成KGs，提取编码有关铜矿化的特定领域见解的符号规则。这些从专家知识中动态导出的规则在训练和预测阶段被集成到机器学习模型中作为指导。通过将符号推理与机器学习的计算能力相融合，我们的方法克服了黑箱模型的局限性，提高了矿物预测的准确性和透明度。为了验证该方法的有效性，我们将其应用于全球铜矿床地球化学综合数据集。结果表明，规则导向的机器学习模型取得了显著的性能改进，在准确性、精密度和鲁棒性方面优于传统的机器学习方法。通过使用SHAP值等工具进一步提高了可解释性，这些工具可以在基于规则的框架内解释单个地球化学特征的影响。这种组合不仅确定了关键的地球化学元素，如Cu、Fe和S，而且还为预测的矿化模式提供了连贯的、域对齐的解释。我们的研究结果证明了结合llm、kg和ML模型进行矿物预测的变革潜力。这种混合方法使地球科学家能够利用计算和专家知识，对成矿系统有更深入的了解。

{"title":"Integrating neuro-symbolic AI and knowledge graph for enhanced geochemical prediction in copper deposits","authors":"Weilin Chen, Jiyin Zhang, Wenjia Li, Xiang Que, Chenhao Li, Xiaogang Ma","doi":"10.1016/j.acags.2025.100259","DOIUrl":"10.1016/j.acags.2025.100259","url":null,"abstract":"<div><div>The integration of machine learning (ML) and deep learning (DL) in geoscience has demonstrated great promise for mineral prediction. However, existing approaches are predominantly data-driven and often overlook expert geological knowledge, limiting their interpretability, accuracy, and practical applicability. This study introduces a new method that combines Large Language Models (LLMs), knowledge graphs (KGs), and Neuro-Symbolic AI (NSAI) models to predict mineralization systems in diverse copper deposits, significantly increasing the precision in prediction results. We utilize LLMs to generate KGs from geological literature, extracting symbolic rules that encode domain-specific insights about copper mineralization. These rules, derived dynamically from expert knowledge, are integrated into ML models as guidance during the training and prediction phases. By fusing symbolic reasoning with ML's computational power, our approach overcomes the limitations of black-box models, offering both improved accuracy and transparency in mineral prediction. To validate this method, we apply it to a comprehensive geochemical dataset of global copper deposits. The results show that rule-guided ML models achieve notable performance improvements, outperforming traditional ML methods in accuracy, precision, and robustness. Interpretability is further enhanced by using tools such as SHAP values, which explain the influence of individual geochemical features within the rule-based framework. This combination not only identifies critical geochemical elements like Cu, Fe, and S but also provides coherent, domain-aligned explanations for the predicted mineralization patterns. Our findings demonstrate the transformative potential of combining LLMs, KGs, and ML models for mineral prediction. This hybrid approach enables geoscientists to leverage both computational and expert knowledge, achieving a deeper understanding of mineralization systems.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"27 ","pages":"Article 100259"},"PeriodicalIF":2.6,"publicationDate":"2025-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144331437","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Application of machine learning-based post-processing to improve crowd-sourced urban rainfall categorizations 基于机器学习的后处理应用于改进众包城市降雨分类

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2025-06-01 DOI: 10.1016/j.acags.2025.100255

Mohammad Ashar Hussain , Venkatesh Budamala , Rajarshi Das Bhowmik

In recent years, citizen science has gained significant attention in the hydrometeorological sciences as an alternative to traditional monitoring systems while also raising awareness of natural processes. Crowd participation in reporting rainfall, known as crowdsourcing rainfall, has the potential to provide insights into the spatio-temporal variability of urban rainfall. However, crowdsourcing often suffers from inaccuracies in rainfall classification due to inadequately trained participants. This study investigates whether machine learning models can reduce misclassification in crowd-sourced rainfall reports under a synthetic framework. A state-of-the-art stochastic rainfall generator is deployed to simulate high-resolution rainfall over Bangalore, India, traditionally monitored by only two rain gauge stations. The study assumes that the 'synthetic' crowd reports qualitative descriptions of two rainfall characteristics—intensity and duration—based on which a categorization of a rainfall event (normal/moderate/severe) is issued. Ten scenarios are introduced to represent varying degrees of misclassification in the crowd reports. Two machine learning models, random forest and logistic regression, are employed to address these misclassifications and improve the resulting rainfall categorization. The findings indicate that while the random forest model outperforms logistic regression, its performance declines as misclassification rates increase. Moreover, the study highlights that increasing the number of participants significantly enhances the post-processing performance, emphasizing the importance of properly training the crowd for accurate reporting.

近年来，公民科学作为传统监测系统的替代方案，在水文气象科学领域获得了极大的关注，同时也提高了人们对自然过程的认识。群众参与降雨报告，被称为众包降雨，有可能提供对城市降雨时空变化的见解。然而，由于参与者训练不足，众包在降雨分类方面经常存在不准确的问题。本研究探讨了在合成框架下，机器学习模型是否可以减少众包降雨报告中的错误分类。部署了最先进的随机降雨发生器来模拟印度班加罗尔的高分辨率降雨，传统上只有两个雨量站监测。研究假设“合成”人群报告两种降雨特征（强度和持续时间）的定性描述，并以此为基础发布降雨事件的分类（正常/中等/严重）。引入了十个场景来表示人群报告中不同程度的错误分类。两种机器学习模型，随机森林和逻辑回归，被用来解决这些错误分类，并改进最终的降雨分类。研究结果表明，虽然随机森林模型优于逻辑回归，但其性能随着误分类率的增加而下降。此外，该研究强调，增加参与者的数量显著提高后处理性能，强调了正确训练人群准确报告的重要性。

{"title":"Application of machine learning-based post-processing to improve crowd-sourced urban rainfall categorizations","authors":"Mohammad Ashar Hussain , Venkatesh Budamala , Rajarshi Das Bhowmik","doi":"10.1016/j.acags.2025.100255","DOIUrl":"10.1016/j.acags.2025.100255","url":null,"abstract":"<div><div>In recent years, citizen science has gained significant attention in the hydrometeorological sciences as an alternative to traditional monitoring systems while also raising awareness of natural processes. Crowd participation in reporting rainfall, known as crowdsourcing rainfall, has the potential to provide insights into the spatio-temporal variability of urban rainfall. However, crowdsourcing often suffers from inaccuracies in rainfall classification due to inadequately trained participants. This study investigates whether machine learning models can reduce misclassification in crowd-sourced rainfall reports under a synthetic framework. A state-of-the-art stochastic rainfall generator is deployed to simulate high-resolution rainfall over Bangalore, India, traditionally monitored by only two rain gauge stations. The study assumes that the 'synthetic' crowd reports qualitative descriptions of two rainfall characteristics—intensity and duration—based on which a categorization of a rainfall event (normal/moderate/severe) is issued. Ten scenarios are introduced to represent varying degrees of misclassification in the crowd reports. Two machine learning models, random forest and logistic regression, are employed to address these misclassifications and improve the resulting rainfall categorization. The findings indicate that while the random forest model outperforms logistic regression, its performance declines as misclassification rates increase. Moreover, the study highlights that increasing the number of participants significantly enhances the post-processing performance, emphasizing the importance of properly training the crowd for accurate reporting.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"26 ","pages":"Article 100255"},"PeriodicalIF":2.6,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144243509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Comparison of ETAS parameter estimates across different time windows within the North and East Anatolian Fault Zones, Turkey 土耳其北安纳托利亚断裂带和东安纳托利亚断裂带不同时间窗内ETAS参数估计的比较

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2025-06-01 DOI: 10.1016/j.acags.2025.100253

Suchanun Piriyasatit , Ercan Engin Kuruoglu , Mehmet Sinan Ozeren

Located at the intersection of major lithospheric plates, Turkey is characterized by significant seismic activity, particularly along the North Anatolian Fault (NAF) and East Anatolian Fault (EAF). This paper employs the Epidemic-Type Aftershock Sequence (ETAS) model, fitted using the BFGS quasi-Newton method, to study earthquake triggering processes along these faults from 1990 to 2023. Our findings show distinct temporal variations in seismicity parameters along these faults. Along the NAF, the ETAS model highlighted a lower background seismicity rate (

μ

) and aftershock productivity (

K_{0}

) compared to the EAF. In contrast, the EAF exhibits lower magnitude sensitivity (

α

), indicating that smaller earthquakes are more likely to trigger aftershocks, due to weaker dependence on mainshock magnitude. The aftershock decay rate (

p

) is notably faster in the NAF, suggesting quicker post-event stabilization. Our analysis across different time windows reveals significant non-stationarities in ETAS parameters, indicating that seismic behaviors along these faults do not strictly follow historical patterns. This temporal variability highlights the challenges in short-term seismic forecasting using historical data alone. A detailed comparison of ETAS parameters across time frames showcases the necessity for incorporating dynamic modeling approaches to improve earthquake forecasting in seismically active regions.

土耳其位于主要岩石圈板块的交汇处，地震活动频繁，特别是沿北安纳托利亚断层（NAF）和东安纳托利亚断层（EAF）。本文采用流行型余震序列（ETAS）模型，采用BFGS准牛顿方法拟合，研究了1990 - 2023年沿这些断裂的地震触发过程。我们的发现表明沿这些断层的地震活动性参数有明显的时间变化。在NAF上，ETAS模型显示背景地震活动性（μ）和余震生产力（K0）比EAF低。相比之下，东震场表现出较低的震级敏感性（α），表明由于对主震震级的依赖性较弱，较小的地震更有可能引发余震。余震衰减率(p)在NAF中明显更快，表明震后稳定更快。我们对不同时间窗的分析揭示了ETAS参数的显著非平稳性，表明沿这些断层的地震行为并不严格遵循历史模式。这种时间变异性突出了仅使用历史数据进行短期地震预报的挑战。ETAS参数跨时间框架的详细比较显示了采用动态建模方法来改进地震活跃地区地震预报的必要性。

{"title":"Comparison of ETAS parameter estimates across different time windows within the North and East Anatolian Fault Zones, Turkey","authors":"Suchanun Piriyasatit , Ercan Engin Kuruoglu , Mehmet Sinan Ozeren","doi":"10.1016/j.acags.2025.100253","DOIUrl":"10.1016/j.acags.2025.100253","url":null,"abstract":"<div><div>Located at the intersection of major lithospheric plates, Turkey is characterized by significant seismic activity, particularly along the North Anatolian Fault (NAF) and East Anatolian Fault (EAF). This paper employs the Epidemic-Type Aftershock Sequence (ETAS) model, fitted using the BFGS quasi-Newton method, to study earthquake triggering processes along these faults from 1990 to 2023. Our findings show distinct temporal variations in seismicity parameters along these faults. Along the NAF, the ETAS model highlighted a lower background seismicity rate (<span><math><mi>μ</mi></math></span>) and aftershock productivity (<span><math><msub><mrow><mi>K</mi></mrow><mrow><mn>0</mn></mrow></msub></math></span>) compared to the EAF. In contrast, the EAF exhibits lower magnitude sensitivity (<span><math><mi>α</mi></math></span>), indicating that smaller earthquakes are more likely to trigger aftershocks, due to weaker dependence on mainshock magnitude. The aftershock decay rate (<span><math><mi>p</mi></math></span>) is notably faster in the NAF, suggesting quicker post-event stabilization. Our analysis across different time windows reveals significant non-stationarities in ETAS parameters, indicating that seismic behaviors along these faults do not strictly follow historical patterns. This temporal variability highlights the challenges in short-term seismic forecasting using historical data alone. A detailed comparison of ETAS parameters across time frames showcases the necessity for incorporating dynamic modeling approaches to improve earthquake forecasting in seismically active regions.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"26 ","pages":"Article 100253"},"PeriodicalIF":2.6,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144279364","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

3D clay microstructure synthesis using Denoising Diffusion Probabilistic Models 基于去噪扩散概率模型的三维粘土微观结构综合

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2025-06-01 DOI: 10.1016/j.acags.2025.100248

Ali Aouf , Eric Laloy , Bart Rogiers , Christophe De Vleeschouwer

This work is concerned with the challenging task of generating 3D-consistent binary microstructures of heterogeneous clay materials. We leverage denoising diffusion probabilistic models (DDPMs) to do so and show that DDPMs outperform two classical generative adversarial networks (GANs) for a 2D generation task. Next, our experiments demonstrate that our DDPMs can produce high-quality, diverse realizations that well capture the spatial statistics of two distinct clay microstructures. Moreover, we show that DDPMs can be implicitly trained to generate porosity-conditioned samples. To the best of our knowledge, this is the first study that addresses clay microstructure generation with DDPMs.

这项工作涉及到产生三维一致的非均质粘土材料二元微结构的挑战性任务。我们利用去噪扩散概率模型（ddpm）来做到这一点，并表明ddpm在2D生成任务中优于两个经典的生成对抗网络（gan）。接下来，我们的实验表明，我们的ddpm可以产生高质量的，多样化的实现，很好地捕获两种不同粘土微观结构的空间统计。此外，我们表明ddpm可以隐式训练来生成孔隙率条件的样本。据我们所知，这是第一个用ddpm处理粘土微观结构生成的研究。

引用次数: 0

Enhancing Indian summer monsoon prediction: Deep learning approach for skillful long-lead forecasts of rainfall 加强印度夏季风预测：深度学习方法用于熟练的长期降雨预测

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2025-06-01 DOI: 10.1016/j.acags.2025.100257

Kalpesh R. Patil, Takeshi Doi, J.V. Ratnam, Swadhin K. Behera

The prediction of the Indian summer monsoon rainfall (ISMR) in the June–September (JJAS) season at long-lead times is challenging. The state-of-the-art dynamical models often fail to capture the sign and amplitude of the rainfall anomalies in the extreme rainfall seasons, limiting the overall skill of the models. We attempted to address this issue using a deep learning model based on convolutional neural networks (CNN). An ensemble of JJAS rainfall predictions using the CNN model with a unique custom function showed high skills in predicting ISMR at a long-lead time of 12 months. The predictions had an anomaly correlation coefficient (ACC) exceeding 0.5 at all the lead times from 2 to 17 months. The CNN model predictions could capture the sign and phase of the extreme rainfall events in the study period realistically. Analysis of saliency-based heatmaps indicated the high skill to be due to the model capturing the leading modes of climate variability, such as the Indian Ocean Dipole and El Niño-Southern Oscillation, realistically. The ensemble of CNN ISMR predictions can supplement the predictions of the forecasting centers.

6 - 9月（JJAS）季节的印度夏季季风降雨（ISMR）的长期预测是具有挑战性的。最先进的动力模式往往不能捕捉极端降雨季节降雨异常的信号和幅度，限制了模式的整体技能。我们尝试使用基于卷积神经网络（CNN）的深度学习模型来解决这个问题。使用CNN模型和独特的自定义函数的JJAS降雨预测集合显示出在12个月的长提前期预测ISMR的高技能。2 ~ 17个月的预测异常相关系数（ACC）均大于0.5。CNN模型预测能够真实地捕捉研究时段极端降雨事件的信号和阶段。对基于显著性的热图的分析表明，高技能是由于该模式实际捕获了气候变率的主要模式，如印度洋偶极子和El Niño-Southern振荡。CNN ISMR预测集合可以补充预报中心的预测。

{"title":"Enhancing Indian summer monsoon prediction: Deep learning approach for skillful long-lead forecasts of rainfall","authors":"Kalpesh R. Patil, Takeshi Doi, J.V. Ratnam, Swadhin K. Behera","doi":"10.1016/j.acags.2025.100257","DOIUrl":"10.1016/j.acags.2025.100257","url":null,"abstract":"<div><div>The prediction of the Indian summer monsoon rainfall (ISMR) in the June–September (JJAS) season at long-lead times is challenging. The state-of-the-art dynamical models often fail to capture the sign and amplitude of the rainfall anomalies in the extreme rainfall seasons, limiting the overall skill of the models. We attempted to address this issue using a deep learning model based on convolutional neural networks (CNN). An ensemble of JJAS rainfall predictions using the CNN model with a unique custom function showed high skills in predicting ISMR at a long-lead time of 12 months. The predictions had an anomaly correlation coefficient (ACC) exceeding 0.5 at all the lead times from 2 to 17 months. The CNN model predictions could capture the sign and phase of the extreme rainfall events in the study period realistically. Analysis of saliency-based heatmaps indicated the high skill to be due to the model capturing the leading modes of climate variability, such as the Indian Ocean Dipole and El Niño-Southern Oscillation, realistically. The ensemble of CNN ISMR predictions can supplement the predictions of the forecasting centers.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"26 ","pages":"Article 100257"},"PeriodicalIF":2.6,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144291535","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Data-driven dynamic friction models based on Recurrent Neural Networks 基于递归神经网络的数据驱动动态摩擦模型

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2025-06-01 DOI: 10.1016/j.acags.2025.100249

Gaëtan Cortes, Joaquin Garcia-Suarez

In this concise contribution, it is demonstrated that Recurrent Neural Networks (RNNs) based on Gated Recurrent Unit (GRU) architecture, possess the capability to learn the complex dynamics of rate-and-state friction (RSF) laws from synthetic data. The data employed for training the network is generated through the application of traditional RSF equations coupled with either the aging law or the slip law for state evolution. A novel aspect of this approach is the formulation of a loss function that explicitly accounts for the direct effect by means of automatic differentiation. It is found that the GRU-based RNNs effectively learns to predict changes in the friction coefficient resulting from velocity jumps (with and without noise in the target data), thereby showcasing the potential of machine learning models in capturing and simulating the physics of frictional processes. Current limitations and challenges are discussed.

在这个简洁的贡献中，证明了基于门控循环单元（GRU）架构的递归神经网络（RNNs）具有从合成数据中学习复杂动态速率和状态摩擦（RSF）定律的能力。用于训练网络的数据是通过将传统的RSF方程与状态演化的老化律或滑移律相结合来生成的。这种方法的一个新颖方面是通过自动微分明确地说明直接影响的损失函数的公式。研究发现，基于gru的rnn有效地学习预测速度跳跃（目标数据中有或没有噪声）导致的摩擦系数变化，从而展示了机器学习模型在捕获和模拟摩擦过程物理方面的潜力。讨论了当前的限制和挑战。

引用次数: 0

Prediction of carbon dioxide phase at bottomhole by adaptive factorization network considering well geometry 考虑井形的自适应分解网络预测井底二氧化碳相

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2025-05-21 DOI: 10.1016/j.acags.2025.100254

Sungil Kim , Tea-Woo Kim , Yongjun Hong , Hoonyoung Jeong

Accurate carbon dioxide (CO₂) phase prediction at the bottomhole of injection wells is essential for ensuring safe and efficient CO₂ storage and enhanced gas recovery (EGR). Phase misclassification can cause operational inefficiencies, equipment failure, and compromised storage integrity, posing significant risks to CO₂ injection projects. While previous studies have contributed to CO₂ phase prediction, they have overlooked well geometry effects, which can impact reliability in real-world applications. This study addresses these challenges by introducing a deep learning framework based on the adaptive factorization network (AFN), which enhances CO₂ phase prediction accuracy by leveraging feature interactions. The AFN model was trained on ∼43,000 wells across seven major North American shale gas basins, covering a wide range of well geometries and injection conditions. CO₂ phases were classified into supercritical and dense categories, reflecting prevailing flow conditions. To enhance practical applicability, we incorporated real-field wellbore data, ensuring alignment with actual injection environments. The standard AFN model achieved an F1-score of 0.94, with data augmentation further improving performance by reducing false predictions by 50 % and increasing the F1-score to 0.97. Rigorous validation demonstrated the model's robustness for optimizing wellhead temperature to achieve the desired CO₂ phase transition. By explicitly considering well geometry effects and real-field conditions, this study advances data-driven CO₂ injection modeling, providing a scalable, high-accuracy framework for evaluating CO₂ storage and EGR feasibility.

注水井井底准确的二氧化碳（CO2）相预测是确保安全高效的CO2储存和提高气采（EGR）的关键。阶段分类错误会导致操作效率低下、设备故障和存储完整性受损，给二氧化碳注入项目带来重大风险。虽然之前的研究对CO2相预测做出了贡献，但它们忽略了井的几何形状效应，这可能会影响实际应用中的可靠性。本研究通过引入基于自适应分解网络（AFN）的深度学习框架来解决这些挑战，该框架通过利用特征交互来提高CO2相位预测的准确性。AFN模型在北美7个主要页岩气盆地的约43,000口井中进行了训练，涵盖了各种井的几何形状和注入条件。CO2相分为超临界和致密两类，反映了主流的流动条件。为了提高实际适用性，我们结合了现场井眼数据，确保与实际注入环境一致。标准AFN模型的f1得分为0.94，数据增强进一步提高了性能，减少了50%的错误预测，并将f1得分提高到0.97。严格的验证证明了该模型在优化井口温度以实现所需的CO2相变方面的鲁棒性。通过明确考虑井的几何效应和现场条件，该研究推进了数据驱动的二氧化碳注入建模，为评估二氧化碳储存和EGR可行性提供了可扩展的、高精度的框架。

{"title":"Prediction of carbon dioxide phase at bottomhole by adaptive factorization network considering well geometry","authors":"Sungil Kim , Tea-Woo Kim , Yongjun Hong , Hoonyoung Jeong","doi":"10.1016/j.acags.2025.100254","DOIUrl":"10.1016/j.acags.2025.100254","url":null,"abstract":"<div><div>Accurate carbon dioxide (CO<sub>2</sub>) phase prediction at the bottomhole of injection wells is essential for ensuring safe and efficient CO<sub>2</sub> storage and enhanced gas recovery (EGR). Phase misclassification can cause operational inefficiencies, equipment failure, and compromised storage integrity, posing significant risks to CO<sub>2</sub> injection projects. While previous studies have contributed to CO<sub>2</sub> phase prediction, they have overlooked well geometry effects, which can impact reliability in real-world applications. This study addresses these challenges by introducing a deep learning framework based on the adaptive factorization network (AFN), which enhances CO<sub>2</sub> phase prediction accuracy by leveraging feature interactions. The AFN model was trained on ∼43,000 wells across seven major North American shale gas basins, covering a wide range of well geometries and injection conditions. CO<sub>2</sub> phases were classified into supercritical and dense categories, reflecting prevailing flow conditions. To enhance practical applicability, we incorporated real-field wellbore data, ensuring alignment with actual injection environments. The standard AFN model achieved an F1-score of 0.94, with data augmentation further improving performance by reducing false predictions by 50 % and increasing the F1-score to 0.97. Rigorous validation demonstrated the model's robustness for optimizing wellhead temperature to achieve the desired CO<sub>2</sub> phase transition. By explicitly considering well geometry effects and real-field conditions, this study advances data-driven CO<sub>2</sub> injection modeling, providing a scalable, high-accuracy framework for evaluating CO<sub>2</sub> storage and EGR feasibility.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"26 ","pages":"Article 100254"},"PeriodicalIF":2.6,"publicationDate":"2025-05-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144123572","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Soil organic carbon retrieval using a machine learning approach from satellite and environmental covariates in the Lower Brazos River Watershed, Texas, USA 基于卫星和环境协变量的机器学习方法在美国德克萨斯州下布拉索斯河流域土壤有机碳检索

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2025-05-16 DOI: 10.1016/j.acags.2025.100252

Birhan Getachew Tikuye, Ram Lakhan Ray

Soil is critical in global carbon storage, holding more carbon than terrestrial vegetation and the atmosphere combined. Accurate soil organic carbon (SOC) estimation is essential for improving agricultural productivity and mitigating climate change. This study aims to explore the retrieval of SOC using a machine learning (ML) approach, leveraging remote sensing data and environmental covariates, focusing on the Lower Brazos River Watershed, southern Texas, USA. The study used Sentinel 2A satellite data-derived indices such as vegetation and water indices, topographic features, soil properties, and climatic factors. Three ML models, namely Gradient Boosting (GB), Random Forest (RF), and eXtreme Gradient Boosting (XGBoost), were deployed, with performance assessed using the R², RMSE, and MAE. All explanatory variables are geospatial gridded datasets, except for the point-based measurement of SOC on the Prairie View A&M University (PVAMU) research farm plot used to train the model. The RF model demonstrated the best performance in model testing, with the lowest root mean square error (RMSE = 4.17) and mean absolute error (MAE = 3), as well as the highest coefficient of determination (R² = 0.78). GB was the second-best performing model, achieving an RMSE of 4.23 and an MAE of 3.12, with similar R² values to the RF model. The average SOC throughout the watershed is 45.5 tons/ha, while the total amount of SOC in the watershed is around 4,278,263 tons. These results suggest that integrating satellite data with environmental covariates and machine learning models holds excellent potential for SOC prediction and supports climate change mitigation efforts by improving carbon stock assessments.

土壤在全球碳储存中起着至关重要的作用，它所储存的碳比陆地植被和大气加起来还要多。准确的土壤有机碳（SOC）估算对于提高农业生产力和减缓气候变化至关重要。本研究以美国德克萨斯州南部下布拉索斯河流域为研究对象，利用遥感数据和环境协变量，探索利用机器学习（ML）方法检索土壤有机碳。该研究使用哨兵2A卫星数据衍生的指数，如植被和水指数、地形特征、土壤性质和气候因素。部署了三种ML模型，即梯度增强（GB），随机森林（RF）和极端梯度增强（XGBoost），并使用R2， RMSE和MAE评估性能。所有解释变量都是地理空间网格数据集，除了用于训练模型的基于点的草原视图A&；M大学（PVAMU）研究农场地块的SOC测量。RF模型在模型检验中表现最好，具有最低的均方根误差（RMSE = 4.17）和平均绝对误差（MAE = 3），决定系数最高（R2 = 0.78）。GB是表现第二好的模型，RMSE为4.23，MAE为3.12，R2值与RF模型相似。整个流域的SOC平均为45.5吨/公顷，而流域的SOC总量约为4278263吨。这些结果表明，将卫星数据与环境协变量和机器学习模型相结合，在有机碳预测方面具有很大的潜力，并通过改进碳储量评估来支持减缓气候变化的努力。

{"title":"Soil organic carbon retrieval using a machine learning approach from satellite and environmental covariates in the Lower Brazos River Watershed, Texas, USA","authors":"Birhan Getachew Tikuye, Ram Lakhan Ray","doi":"10.1016/j.acags.2025.100252","DOIUrl":"10.1016/j.acags.2025.100252","url":null,"abstract":"<div><div>Soil is critical in global carbon storage, holding more carbon than terrestrial vegetation and the atmosphere combined. Accurate soil organic carbon (SOC) estimation is essential for improving agricultural productivity and mitigating climate change. This study aims to explore the retrieval of SOC using a machine learning (ML) approach, leveraging remote sensing data and environmental covariates, focusing on the Lower Brazos River Watershed, southern Texas, USA. The study used Sentinel 2A satellite data-derived indices such as vegetation and water indices, topographic features, soil properties, and climatic factors. Three ML models, namely Gradient Boosting (GB), Random Forest (RF), and eXtreme Gradient Boosting (XGBoost), were deployed, with performance assessed using the R<sup>2</sup>, RMSE, and MAE. All explanatory variables are geospatial gridded datasets, except for the point-based measurement of SOC on the Prairie View A&M University (PVAMU) research farm plot used to train the model. The RF model demonstrated the best performance in model testing, with the lowest root mean square error (RMSE = 4.17) and mean absolute error (MAE = 3), as well as the highest coefficient of determination (R<sup>2</sup> = 0.78). GB was the second-best performing model, achieving an RMSE of 4.23 and an MAE of 3.12, with similar R<sup>2</sup> values to the RF model. The average SOC throughout the watershed is 45.5 tons/ha, while the total amount of SOC in the watershed is around 4,278,263 tons. These results suggest that integrating satellite data with environmental covariates and machine learning models holds excellent potential for SOC prediction and supports climate change mitigation efforts by improving carbon stock assessments.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"26 ","pages":"Article 100252"},"PeriodicalIF":2.6,"publicationDate":"2025-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144117054","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0