Applied Computing and Geosciences最新文献

CloudSense: A model for cloud type identification using machine learning from radar data

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2024-12-01 DOI: 10.1016/j.acags.2024.100209

Mehzooz Nizar , Jha K. Ambuj , Manmeet Singh , S.B. Vaisakh , G. Pandithurai

The knowledge of type of precipitating cloud is crucial for radar based quantitative estimates of precipitation. We propose a novel model called CloudSense which uses machine learning to accurately identify the type of precipitating clouds over the complex terrain locations in the Western Ghats (WG) of India. CloudSense uses vertical reflectivity profiles collected during July–August 2018 from an X-band radar to classify clouds into four categories namely stratiform, mixed stratiform-convective, convective and shallow clouds. The machine learning (ML) model used in CloudSense was trained using a dataset balanced by Synthetic Minority Oversampling Technique (SMOTE), with features selected based on physical characteristics relevant to different cloud types. Among various ML models evaluated Light Gradient Boosting Machine (LightGBM) demonstrate superior performance in classifying cloud types with a BAC (Balanced Accuracy) of 0.79 and F1-Score of 0.8. CloudSense generated results are also compared against conventional radar algorithms and we find that CloudSense performs better than radar algorithms. For 200 samples tested, the radar algorithm achieved a BAC of 0.69 and F1-Score of 0.68, whereas CloudSense achieved a BAC of 0.8 and F1-Score of 0.79. Our results show that ML based approach can provide more accurate cloud detection and classification which would be useful to improve precipitation estimates over the complex terrain of the WG.

{"title":"CloudSense: A model for cloud type identification using machine learning from radar data","authors":"Mehzooz Nizar , Jha K. Ambuj , Manmeet Singh , S.B. Vaisakh , G. Pandithurai","doi":"10.1016/j.acags.2024.100209","DOIUrl":"10.1016/j.acags.2024.100209","url":null,"abstract":"<div><div>The knowledge of type of precipitating cloud is crucial for radar based quantitative estimates of precipitation. We propose a novel model called CloudSense which uses machine learning to accurately identify the type of precipitating clouds over the complex terrain locations in the Western Ghats (WG) of India. CloudSense uses vertical reflectivity profiles collected during July–August 2018 from an X-band radar to classify clouds into four categories namely stratiform, mixed stratiform-convective, convective and shallow clouds. The machine learning (ML) model used in CloudSense was trained using a dataset balanced by Synthetic Minority Oversampling Technique (SMOTE), with features selected based on physical characteristics relevant to different cloud types. Among various ML models evaluated Light Gradient Boosting Machine (LightGBM) demonstrate superior performance in classifying cloud types with a BAC (Balanced Accuracy) of 0.79 and F1-Score of 0.8. CloudSense generated results are also compared against conventional radar algorithms and we find that CloudSense performs better than radar algorithms. For 200 samples tested, the radar algorithm achieved a BAC of 0.69 and F1-Score of 0.68, whereas CloudSense achieved a BAC of 0.8 and F1-Score of 0.79. Our results show that ML based approach can provide more accurate cloud detection and classification which would be useful to improve precipitation estimates over the complex terrain of the WG.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"24 ","pages":"Article 100209"},"PeriodicalIF":2.6,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745714","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A machine learning approach for mapping susceptibility to land subsidence caused by ground water extraction 绘制地下水开采导致土地沉降易感性地图的机器学习方法

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2024-11-16 DOI: 10.1016/j.acags.2024.100207

Diana Orlandi , Esteban Díaz , Roberto Tomás , Federico A. Galatolo , Mario G.C.A. Cimino , Carolina Pagli , Nicola Perilli

Land subsidence is a worldwide threat that may cause irreversible damage to the environment and the infrastructures. Thus, identifying and mapping areas prone to land subsidence with accurate methods such as Land Subsidence Susceptibility Index (LSSI) mapping is crucial for mitigating the adverse impacts of this geohazard. Also, Machine Learning (ML) is now becoming a powerful tool to analyze vast and different datasets such as those necessary for LSSI mapping. In this study, we use the conventional Frequency Ratio (FR) method and ML models to generate LSSI maps of the region of Murcia (Spain) where land subsidence occurred in the past due to groundwater overdraft. A LSSI map was initially generated with known FR. Then, additional Conditioning Factors (CFs) with increased spatial resolution were used to train several ML models and generate a new LSSI map. The Extra-Trees Classifier (ETC) outperformed the other approaches, achieving the best performance with a weighted average precision and F1-Score of 0.96, after optimizing its hyperparameters. Then, a third LSSI map was calculated using the FR method and observations of land subsidence from InSAR (Interferometric Synthetic Aperture Radar). This study shows that the effectiveness of using several CFs depends on the added information of each layer. Moreover, the comparison between the different LSSI maps and InSAR data highlights the crucial role of the spatial resolution for accurate mapping, thus enhancing land subsidence risk assessment.

土地沉降是一种全球性威胁，可能对环境和基础设施造成不可逆转的破坏。因此，采用土地沉降易感指数（LSSI）绘图等精确方法识别和绘制易发生土地沉降的地区，对于减轻这种地质灾害的不利影响至关重要。此外，机器学习（ML）目前已成为分析大量不同数据集（如 LSSI 测绘所需的数据集）的有力工具。在本研究中，我们使用传统的频率比（FR）方法和 ML 模型来生成穆尔西亚地区（西班牙）的 LSSI 地图，该地区过去曾因地下水超采而发生过地面沉降。LSSI 地图最初是根据已知 FR 生成的。然后，使用空间分辨率更高的附加条件因子（CF）来训练多个 ML 模型，生成新的 LSSI 地图。优化超参数后，树外分类器（ETC）的表现优于其他方法，加权平均精度和 F1 分数均达到 0.96。然后，利用 FR 方法和 InSAR（干涉合成孔径雷达）的地面沉降观测数据计算了第三幅 LSSI 地图。这项研究表明，使用多个 CF 的有效性取决于每个层的附加信息。此外，不同的 LSSI 地图和 InSAR 数据之间的比较凸显了空间分辨率对精确制图的关键作用，从而加强了土地沉降风险评估。

{"title":"A machine learning approach for mapping susceptibility to land subsidence caused by ground water extraction","authors":"Diana Orlandi , Esteban Díaz , Roberto Tomás , Federico A. Galatolo , Mario G.C.A. Cimino , Carolina Pagli , Nicola Perilli","doi":"10.1016/j.acags.2024.100207","DOIUrl":"10.1016/j.acags.2024.100207","url":null,"abstract":"<div><div>Land subsidence is a worldwide threat that may cause irreversible damage to the environment and the infrastructures. Thus, identifying and mapping areas prone to land subsidence with accurate methods such as Land Subsidence Susceptibility Index (LSSI) mapping is crucial for mitigating the adverse impacts of this geohazard. Also, Machine Learning (ML) is now becoming a powerful tool to analyze vast and different datasets such as those necessary for LSSI mapping. In this study, we use the conventional Frequency Ratio (FR) method and ML models to generate LSSI maps of the region of Murcia (Spain) where land subsidence occurred in the past due to groundwater overdraft. A LSSI map was initially generated with known FR. Then, additional Conditioning Factors (CFs) with increased spatial resolution were used to train several ML models and generate a new LSSI map. The Extra-Trees Classifier (ETC) outperformed the other approaches, achieving the best performance with a weighted average precision and F1-Score of 0.96, after optimizing its hyperparameters. Then, a third LSSI map was calculated using the FR method and observations of land subsidence from InSAR (Interferometric Synthetic Aperture Radar). This study shows that the effectiveness of using several CFs depends on the added information of each layer. Moreover, the comparison between the different LSSI maps and InSAR data highlights the crucial role of the spatial resolution for accurate mapping, thus enhancing land subsidence risk assessment.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"24 ","pages":"Article 100207"},"PeriodicalIF":2.6,"publicationDate":"2024-11-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142699885","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Revolutionizing the future of hydrological science: Impact of machine learning and deep learning amidst emerging explainable AI and transfer learning 彻底改变水文科学的未来：机器学习和深度学习在新兴可解释人工智能和迁移学习中的影响

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2024-11-09 DOI: 10.1016/j.acags.2024.100206

Rajib Maity, Aman Srivastava, Subharthi Sarkar, Mohd Imran Khan

Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) are revolutionizing hydrology, driving significant advancements in water resource management, modeling, and prediction. This review synthesizes cutting-edge developments, methodologies, and applications of AI-ML-DL across key hydrological processes. By critically evaluating these techniques against traditional models, we highlight their superior ability to capture complex, nonlinear relationships and adapt to diverse environments. We further explore AI applications in precipitation forecasting, evapotranspiration estimation, groundwater dynamics, and extreme event prediction (floods, droughts, and compound events), showcasing their timely potential in addressing critical water-related challenges. A particular emphasis is placed on Explainable AI (XAI) and transfer learning as essential tools for improving model transparency and applicability, enabling broader stakeholder trust and cross-regional adaptability. The review also addresses persistent challenges, including data limitations, computational demands, and model interpretability, proposing solutions that integrate emerging technologies like quantum computing, the Internet of Things (IoT), and interdisciplinary collaboration. This review charts a strategic course for future research and practice by bridging AI advancements with practical hydrological applications. Our findings highlight the importance of embracing AI-driven approaches for next-generation hydrological modeling and provide actionable understandings for researchers, practitioners, and policymakers. As hydrology faces escalating challenges due to human-induced climate change and growing water demands, the continued evolution of AI-integrated models and innovations in data handling and stakeholder engagement will be imperative. In conclusion, the findings emphasize the critical role of AI-driven hydrological modeling in addressing global water challenges, including climate change adaptation, sustainable water resource management, and disaster risk reduction.

人工智能（AI）、机器学习（ML）和深度学习（DL）正在彻底改变水文学，推动水资源管理、建模和预测领域的重大进步。本综述综合了人工智能-ML-DL 在关键水文过程中的前沿发展、方法和应用。通过对这些技术与传统模型进行严格评估，我们强调了这些技术在捕捉复杂的非线性关系和适应不同环境方面的卓越能力。我们进一步探讨了人工智能在降水预报、蒸散估计、地下水动力学和极端事件预测（洪水、干旱和复合事件）中的应用，展示了它们在应对与水有关的重大挑战方面的及时潜力。其中特别强调了可解释人工智能（XAI）和迁移学习，将其作为提高模型透明度和适用性的重要工具，从而实现更广泛的利益相关者信任和跨区域适应性。综述还探讨了长期存在的挑战，包括数据限制、计算需求和模型可解释性，并提出了整合量子计算、物联网（IoT）和跨学科合作等新兴技术的解决方案。本综述将人工智能的进步与实际水文应用相结合，为未来的研究和实践指明了战略方向。我们的研究结果强调了采用人工智能驱动的方法进行下一代水文建模的重要性，并为研究人员、从业人员和决策者提供了可操作的理解。由于人类引起的气候变化和日益增长的水资源需求，水文学面临着不断升级的挑战，因此，人工智能集成模型的持续发展以及数据处理和利益相关者参与方面的创新势在必行。总之，研究结果强调了人工智能驱动的水文建模在应对全球水资源挑战（包括适应气候变化、可持续水资源管理和减少灾害风险）方面的关键作用。

{"title":"Revolutionizing the future of hydrological science: Impact of machine learning and deep learning amidst emerging explainable AI and transfer learning","authors":"Rajib Maity, Aman Srivastava, Subharthi Sarkar, Mohd Imran Khan","doi":"10.1016/j.acags.2024.100206","DOIUrl":"10.1016/j.acags.2024.100206","url":null,"abstract":"<div><div>Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) are revolutionizing hydrology, driving significant advancements in water resource management, modeling, and prediction. This review synthesizes cutting-edge developments, methodologies, and applications of AI-ML-DL across key hydrological processes. By critically evaluating these techniques against traditional models, we highlight their superior ability to capture complex, nonlinear relationships and adapt to diverse environments. We further explore AI applications in precipitation forecasting, evapotranspiration estimation, groundwater dynamics, and extreme event prediction (floods, droughts, and compound events), showcasing their timely potential in addressing critical water-related challenges. A particular emphasis is placed on Explainable AI (XAI) and transfer learning as essential tools for improving model transparency and applicability, enabling broader stakeholder trust and cross-regional adaptability. The review also addresses persistent challenges, including data limitations, computational demands, and model interpretability, proposing solutions that integrate emerging technologies like quantum computing, the Internet of Things (IoT), and interdisciplinary collaboration. This review charts a strategic course for future research and practice by bridging AI advancements with practical hydrological applications. Our findings highlight the importance of embracing AI-driven approaches for next-generation hydrological modeling and provide actionable understandings for researchers, practitioners, and policymakers. As hydrology faces escalating challenges due to human-induced climate change and growing water demands, the continued evolution of AI-integrated models and innovations in data handling and stakeholder engagement will be imperative. In conclusion, the findings emphasize the critical role of AI-driven hydrological modeling in addressing global water challenges, including climate change adaptation, sustainable water resource management, and disaster risk reduction.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"24 ","pages":"Article 100206"},"PeriodicalIF":2.6,"publicationDate":"2024-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142652751","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Generating land gravity anomalies from satellite gravity observations using PIX2PIX GAN image translation 利用 PIX2PIX GAN 图像转换从卫星重力观测中生成陆地重力异常点

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2024-11-06 DOI: 10.1016/j.acags.2024.100205

Bisrat Teshome Weldemikael , Girma Woldetinsae , Girma Neshir

Generative Adversarial Networks (GANs), specifically the Pix2Pix GAN, are used to effectively map gravity anomalies from satellite to ground, and adapt the Pix2Pix GAN model for large-scale data transformation. The impact of varying patch sizes on model performance is investigated using key metrics to ensure improved accuracy in gravity anomaly mapping. The model used 2728 satellite, and 2728 ground Bouguer gravity anomaly images from northern and northeast part of Ethiopia. 5456 images were used for training and 552 for testing. The findings indicate that Intermediate patch sizes, particularly 70 x 70 pixels, significantly enhanced model accuracy by capturing global features and contextual information. Additionally, models incorporating L2 loss with LcGAN demonstrated superior performance across qualitative metrics compared to those with L1 loss. The study will contribute to improve geophysical exploration by providing an alternative method that generates more accurate gravity maps, thereby enhancing the precision of geological models and related applications.

生成对抗网络（GAN），特别是 Pix2Pix GAN，用于有效绘制从卫星到地面的重力异常图，并使 Pix2Pix GAN 模型适用于大规模数据转换。利用关键指标研究了不同斑块大小对模型性能的影响，以确保提高重力异常绘图的准确性。该模型使用了埃塞俄比亚北部和东北部的 2728 幅卫星图像和 2728 幅地面布格重力异常图像。5456 幅图像用于训练，552 幅图像用于测试。研究结果表明，中间补丁大小，尤其是 70 x 70 像素，通过捕捉全局特征和上下文信息，显著提高了模型的准确性。此外，与采用 L1 损失的模型相比，采用 L2 损失和 LcGAN 的模型在质量指标方面表现出更优越的性能。这项研究提供了一种可生成更精确重力地图的替代方法，从而提高了地质模型和相关应用的精度，有助于改善地球物理勘探。

{"title":"Generating land gravity anomalies from satellite gravity observations using PIX2PIX GAN image translation","authors":"Bisrat Teshome Weldemikael , Girma Woldetinsae , Girma Neshir","doi":"10.1016/j.acags.2024.100205","DOIUrl":"10.1016/j.acags.2024.100205","url":null,"abstract":"<div><div>Generative Adversarial Networks (GANs), specifically the Pix2Pix GAN, are used to effectively map gravity anomalies from satellite to ground, and adapt the Pix2Pix GAN model for large-scale data transformation. The impact of varying patch sizes on model performance is investigated using key metrics to ensure improved accuracy in gravity anomaly mapping. The model used 2728 satellite, and 2728 ground Bouguer gravity anomaly images from northern and northeast part of Ethiopia. 5456 images were used for training and 552 for testing. The findings indicate that Intermediate patch sizes, particularly 70 x 70 pixels, significantly enhanced model accuracy by capturing global features and contextual information. Additionally, models incorporating L2 loss with LcGAN demonstrated superior performance across qualitative metrics compared to those with L1 loss. The study will contribute to improve geophysical exploration by providing an alternative method that generates more accurate gravity maps, thereby enhancing the precision of geological models and related applications.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"24 ","pages":"Article 100205"},"PeriodicalIF":2.6,"publicationDate":"2024-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142652752","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Reconstruction of reservoir rock using attention-based convolutional recurrent neural network 利用基于注意力的卷积递归神经网络重建储层岩石

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2024-10-23 DOI: 10.1016/j.acags.2024.100202

Indrajeet Kumar, Anugrah Singh

The digital reconstruction of reservoir rock or porous media is important as it enables us to visualize and explore their real internal structures. The reservoir rocks (such as sandstone and carbonate) contain both spatial and temporal characteristics, which pose a big challenge in their characterization through routine core analysis or x-ray microcomputer tomography. While x-ray micro-computed tomography gives us three-dimensional images of the porous media, it is often impossible to quantify the variability of the pore, grains, structure, and orientation experimentally. Recently, machine learning has successfully demonstrated the reconstruction ability of reservoir rock images or any porous media. These reservoir rock images are crucial for the digital characterization of the reservoir. We propose a novel algorithm consisting of the convolutional neural network, an attention mechanism, and a recurrent neural network for the reconstruction of reservoir rock or porous media images. The attention-based convolutional recurrent neural network (ACRNN) can reconstruct a representative sample of reservoir rocks. The reconstructed image quality was checked by comparing them with the original Parker sandstone, Leopard sandstone, carbonate shale, Berea sandstone, and sandy medium images. We evaluated the reconstruction by measuring pore and throat properties, two-point probability function, and structural similarity index. Results show that ACRNN can reconstruct reservoir rock or porous media of different scales with approximately the same geometrical, statistical, and topological parameters of the reservoir rock images. This deep learning method is computationally efficient, fast, and reliable for synthetic image realizations. The model was trained and validated on real images, and the reconstructed images showed excellent concordance with the real images having almost the same pore and grain structures. The deep learning-based digital rock reconstruction of reservoir rock or porous media images can aid in rapid image generation to better understand reservoir rock or subsurface formation.

储层岩石或多孔介质的数字重建非常重要，因为它能使我们直观地了解和探索其真实的内部结构。储层岩石（如砂岩和碳酸盐岩）包含空间和时间特征，这给通过常规岩心分析或 X 射线微计算机断层扫描来描述其特征带来了巨大挑战。虽然 X 射线微计算机断层扫描可以提供多孔介质的三维图像，但通常无法通过实验量化孔隙、晶粒、结构和取向的变化。最近，机器学习已经成功证明了储层岩石图像或任何多孔介质的重建能力。这些储层岩石图像对于储层的数字化特征描述至关重要。我们提出了一种由卷积神经网络、注意力机制和递归神经网络组成的新算法，用于重建储层岩石或多孔介质图像。基于注意力的卷积递归神经网络（ACRNN）可以重建具有代表性的储层岩石样本。通过与原始的帕克砂岩、豹纹砂岩、碳酸盐页岩、贝里亚砂岩和砂质介质图像进行比较，检查了重建图像的质量。我们通过测量孔隙和喉管属性、两点概率函数和结构相似性指数来评估重建结果。结果表明，ACRNN 可以重建不同尺度的储层岩石或多孔介质，其几何、统计和拓扑参数与储层岩石图像大致相同。这种深度学习方法的计算效率高、速度快，而且对合成图像的实现非常可靠。该模型在真实图像上进行了训练和验证，重建后的图像与具有几乎相同孔隙和晶粒结构的真实图像非常吻合。基于深度学习的储层岩石或多孔介质图像的数字岩石重建有助于快速生成图像，从而更好地了解储层岩石或地下地层。

{"title":"Reconstruction of reservoir rock using attention-based convolutional recurrent neural network","authors":"Indrajeet Kumar, Anugrah Singh","doi":"10.1016/j.acags.2024.100202","DOIUrl":"10.1016/j.acags.2024.100202","url":null,"abstract":"<div><div>The digital reconstruction of reservoir rock or porous media is important as it enables us to visualize and explore their real internal structures. The reservoir rocks (such as sandstone and carbonate) contain both spatial and temporal characteristics, which pose a big challenge in their characterization through routine core analysis or x-ray microcomputer tomography. While x-ray micro-computed tomography gives us three-dimensional images of the porous media, it is often impossible to quantify the variability of the pore, grains, structure, and orientation experimentally. Recently, machine learning has successfully demonstrated the reconstruction ability of reservoir rock images or any porous media. These reservoir rock images are crucial for the digital characterization of the reservoir. We propose a novel algorithm consisting of the convolutional neural network, an attention mechanism, and a recurrent neural network for the reconstruction of reservoir rock or porous media images. The attention-based convolutional recurrent neural network (ACRNN) can reconstruct a representative sample of reservoir rocks. The reconstructed image quality was checked by comparing them with the original Parker sandstone, Leopard sandstone, carbonate shale, Berea sandstone, and sandy medium images. We evaluated the reconstruction by measuring pore and throat properties, two-point probability function, and structural similarity index. Results show that ACRNN can reconstruct reservoir rock or porous media of different scales with approximately the same geometrical, statistical, and topological parameters of the reservoir rock images. This deep learning method is computationally efficient, fast, and reliable for synthetic image realizations. The model was trained and validated on real images, and the reconstructed images showed excellent concordance with the real images having almost the same pore and grain structures. The deep learning-based digital rock reconstruction of reservoir rock or porous media images can aid in rapid image generation to better understand reservoir rock or subsurface formation.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"24 ","pages":"Article 100202"},"PeriodicalIF":2.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142534336","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Mapping landforms of a hilly landscape using machine learning and high-resolution LiDAR topographic data 利用机器学习和高分辨率激光雷达地形数据绘制丘陵地貌图

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2024-10-23 DOI: 10.1016/j.acags.2024.100203

Netra R. Regmi , Nina D.S. Webb , Jacob I. Walter , Joonghyeok Heo , Nicholas W. Hayman

Landform maps are important tools in assessment of soil- and eco-hydrogeomorphic processes and hazards, hydrological modeling, and natural resources and land management. Traditional techniques of mapping landforms based on field surveys or from aerial photographs can be time and labor intensive, highlighting the importance of remote sensing products based automatic or semi-automatic approaches. In addition, the time-intensive manual labeling can also be subjective rather than an objective identification of the landform. Here we implemented such an objective approach applying a random forest machine learning algorithm to a set of observed landform data and 1m horizontal resolution bare-earth digital elevation model (DEM) developed from airborne light detection and ranging (LiDAR) data to rapidly map various landforms of a hilly landscape. The landform classification includes upland plateaus, ridges, convex slopes, planar slopes, concave slopes, stream channels, and valley bottoms, across a 400 km² hilly landscape of the Ozark Mountains in northeastern Oklahoma. We used 4200 landform observations (600 per landform) and eight topographic indices derived from 2 m, 5 m and 10 m resolution LiDAR DEM in random forest algorithm to develop 2 m, 5 m and 10 m resolution landform models. We test the effectiveness of DEM resolution in mapping landforms via comparison of observed landforms with modeled landforms. Results showed that the approach mapped ∼84% of observed landforms when covariates were at 2 m resolution to ∼89% when they were at 10 m resolution. However, predicted maps showed that the 2 m resolution covariates performed better at capturing accurate landform boundaries and details of small-sized landforms such as stream channels and ridges. The approach presented here significantly reduces the time required for mapping landforms compared to traditional techniques using aerial imagery and field observations and allows incorporation of a wide variety of covariates. The landform map developed using this approach has several potential applications. It could be utilized in hydrological modeling, natural resource management, and characterizing soil-geomorphic processes and hazards in a hilly landscape.

地貌图是评估土壤和生态水文地质过程和危害、水文建模以及自然资源和土地管理的重要工具。基于实地勘测或航拍照片绘制地貌图的传统技术耗时耗力，因此基于遥感产品的自动或半自动方法显得尤为重要。此外，耗时耗力的人工标注也可能是主观的，而不是对地貌的客观识别。在此，我们将随机森林机器学习算法应用于一组观测到的地貌数据和由机载光探测与测距（LiDAR）数据开发的 1 米水平分辨率裸地数字高程模型（DEM），以快速绘制丘陵地形的各种地貌。地貌分类包括俄克拉荷马州东北部奥扎克山脉 400 平方公里丘陵地带的高地高原、山脊、凸坡、平面坡、凹坡、河道和谷底。我们使用 4200 个地貌观测点（每个地貌点 600 个观测点）以及从 2 m、5 m 和 10 m 分辨率的 LiDAR DEM 中提取的 8 个地形指数，通过随机森林算法开发出 2 m、5 m 和 10 m 分辨率的地貌模型。我们通过将观测到的地貌与建模的地貌进行比较，检验了 DEM 分辨率在绘制地貌图方面的有效性。结果表明，当协变量分辨率为 2 米时，该方法绘制了 84% 的观测地貌，而当协变量分辨率为 10 米时，则绘制了 89% 的观测地貌。然而，预测图显示，2 米分辨率的协变量在捕捉准确的地貌边界和小型地貌（如河道和山脊）细节方面表现更好。与使用航拍图像和实地观测的传统技术相比，本文介绍的方法大大缩短了绘制地貌图所需的时间，并可纳入多种协变量。使用这种方法绘制的地貌图有多种潜在应用。它可用于水文建模、自然资源管理以及描述丘陵地带的土壤地貌过程和危害。

{"title":"Mapping landforms of a hilly landscape using machine learning and high-resolution LiDAR topographic data","authors":"Netra R. Regmi , Nina D.S. Webb , Jacob I. Walter , Joonghyeok Heo , Nicholas W. Hayman","doi":"10.1016/j.acags.2024.100203","DOIUrl":"10.1016/j.acags.2024.100203","url":null,"abstract":"<div><div>Landform maps are important tools in assessment of soil- and eco-hydrogeomorphic processes and hazards, hydrological modeling, and natural resources and land management. Traditional techniques of mapping landforms based on field surveys or from aerial photographs can be time and labor intensive, highlighting the importance of remote sensing products based automatic or semi-automatic approaches. In addition, the time-intensive manual labeling can also be subjective rather than an objective identification of the landform. Here we implemented such an objective approach applying a random forest machine learning algorithm to a set of observed landform data and 1m horizontal resolution bare-earth digital elevation model (DEM) developed from airborne light detection and ranging (LiDAR) data to rapidly map various landforms of a hilly landscape. The landform classification includes upland plateaus, ridges, convex slopes, planar slopes, concave slopes, stream channels, and valley bottoms, across a 400 km<sup>2</sup> hilly landscape of the Ozark Mountains in northeastern Oklahoma. We used 4200 landform observations (600 per landform) and eight topographic indices derived from 2 m, 5 m and 10 m resolution LiDAR DEM in random forest algorithm to develop 2 m, 5 m and 10 m resolution landform models. We test the effectiveness of DEM resolution in mapping landforms via comparison of observed landforms with modeled landforms. Results showed that the approach mapped ∼84% of observed landforms when covariates were at 2 m resolution to ∼89% when they were at 10 m resolution. However, predicted maps showed that the 2 m resolution covariates performed better at capturing accurate landform boundaries and details of small-sized landforms such as stream channels and ridges. The approach presented here significantly reduces the time required for mapping landforms compared to traditional techniques using aerial imagery and field observations and allows incorporation of a wide variety of covariates. The landform map developed using this approach has several potential applications. It could be utilized in hydrological modeling, natural resource management, and characterizing soil-geomorphic processes and hazards in a hilly landscape.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"24 ","pages":"Article 100203"},"PeriodicalIF":2.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142534337","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Evaluating the performances of SVR and XGBoost for short-range forecasting of heatwaves across different temperature zones of India 评估 SVR 和 XGBoost 在印度不同温度带热浪短程预报中的性能

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2024-10-23 DOI: 10.1016/j.acags.2024.100204

Srikanth Bhoopathi, Nitish Kumar, Somesh, Manali Pal

This research aims to forecast maximum temperatures and the frequency of heatwave days across four different temperature zones (Zone 1, 2, 3 and 4) in India. These four zones are categorized based on the 30-year average maximum temperatures (T_30AMT) during the summer months of April, May, and June (AMJ). Two Machine Learning (ML) algorithms eXtreme Gradient Boosting (XGBoost) and Support Vector Regression (SVR) are employed to achieve this goal. The study utilizes nine key atmospheric variables namely air temperature, geopotential height, relative humidity, U-wind, V-wind, soil moisture, solar radiation, sea surface temperature, and mean sea level pressure at a daily scale spanning from 1991 to 2020 for the months of March, April, May, and June as predictors. The India Meteorological Department daily maximum temperature data spanning from 1991 to 2020 for the months of AMJ serves as the predictands. ML models are developed using spatially averaged atmospheric variables and daily maximum temperature across the grids falling within each temperature zone. Results indicate that for a 7-day lead time, SVR outperforms XGBoost in Zone-1 (T_30AMT > 38 °C) and Zone-2 (T_30AMT: 35.01 °C–38 °C) by more accurately capturing peak temperatures during training and testing. Conversely, for a 15-day lead time in Zone-1, XGBoost better predicts temperature peaks in both phases. In Zone-3 (T_30AMT: 30 °C–35 °C) and Zone-4 (T_30AMT < 30 °C) for both lead times, the performance of both models decline, indicating models and input variables are more effective in predicting higher temperatures typical of Zone-1 and 2 but less so in Zone-3 and 4. In a nutshell, the study attempts to highlight the capability of advanced ML techniques combined with spatial climate data to enhance the prediction of extreme heatwave events. These insights can aid in heatwave preparedness, climate management, and adaptation strategies for different Indian temperature zones.

本研究旨在预测印度四个不同温度区（1、2、3 和 4 区）的最高气温和热浪日频率。这四个区域是根据四月、五月和六月（AMJ）夏季的 30 年平均最高气温（T30AMT）划分的。为实现这一目标，采用了两种机器学习（ML）算法：梯度提升算法（XGBoost）和支持向量回归算法（SVR）。该研究采用了九个关键大气变量作为预测因子，即 1991 年至 2020 年 3 月、4 月、5 月和 6 月每日的气温、位势高度、相对湿度、U 风、V 风、土壤湿度、太阳辐射、海面温度和平均海平面气压。印度气象局 1991 年至 2020 年 AMJ 月份的日最高气温数据作为预测因子。利用空间平均大气变量和每个气温区域内网格的日最高气温开发了 ML 模型。结果表明，在 7 天的前导时间内，SVR 在 1 区（T30AMT > 38 ℃）和 2 区（T30AMT: 35.01 ℃-38 ℃）的表现优于 XGBoost，因为 SVR 能更准确地捕捉到训练和测试期间的峰值温度。相反，在前导时间为 15 天的 1 区，XGBoost 能更好地预测两个阶段的温度峰值。在第 3 区（T30AMT：30 °C-35 °C）和第 4 区（T30AMT < 30 °C），在两个提前期，两个模型的性能都有所下降，表明模型和输入变量在预测第 1 区和第 2 区典型的较高温度时更为有效，但在第 3 区和第 4 区则效果较差。总之，这项研究试图强调先进的 ML 技术与空间气候数据相结合的能力，以加强对极端热浪事件的预测。这些见解有助于印度不同温区的热浪防范、气候管理和适应战略。

{"title":"Evaluating the performances of SVR and XGBoost for short-range forecasting of heatwaves across different temperature zones of India","authors":"Srikanth Bhoopathi, Nitish Kumar, Somesh, Manali Pal","doi":"10.1016/j.acags.2024.100204","DOIUrl":"10.1016/j.acags.2024.100204","url":null,"abstract":"<div><div>This research aims to forecast maximum temperatures and the frequency of heatwave days across four different temperature zones (Zone 1, 2, 3 and 4) in India. These four zones are categorized based on the 30-year average maximum temperatures (T<sub>30AMT</sub>) during the summer months of April, May, and June (AMJ). Two Machine Learning (ML) algorithms eXtreme Gradient Boosting (XGBoost) and Support Vector Regression (SVR) are employed to achieve this goal. The study utilizes nine key atmospheric variables namely air temperature, geopotential height, relative humidity, U-wind, V-wind, soil moisture, solar radiation, sea surface temperature, and mean sea level pressure at a daily scale spanning from 1991 to 2020 for the months of March, April, May, and June as predictors. The India Meteorological Department daily maximum temperature data spanning from 1991 to 2020 for the months of AMJ serves as the predictands. ML models are developed using spatially averaged atmospheric variables and daily maximum temperature across the grids falling within each temperature zone. Results indicate that for a 7-day lead time, SVR outperforms XGBoost in Zone-1 (T<sub>30AMT</sub> > 38 °C) and Zone-2 (T<sub>30AMT</sub>: 35.01 °C–38 °C) by more accurately capturing peak temperatures during training and testing. Conversely, for a 15-day lead time in Zone-1, XGBoost better predicts temperature peaks in both phases. In Zone-3 (T<sub>30AMT</sub>: 30 °C–35 °C) and Zone-4 (T<sub>30AMT</sub> < 30 °C) for both lead times, the performance of both models decline, indicating models and input variables are more effective in predicting higher temperatures typical of Zone-1 and 2 but less so in Zone-3 and 4. In a nutshell, the study attempts to highlight the capability of advanced ML techniques combined with spatial climate data to enhance the prediction of extreme heatwave events. These insights can aid in heatwave preparedness, climate management, and adaptation strategies for different Indian temperature zones.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"24 ","pages":"Article 100204"},"PeriodicalIF":2.6,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142534341","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Current progress in subseasonal-to-decadal prediction based on machine learning 基于机器学习的副季节至十年期预测的最新进展

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2024-10-22 DOI: 10.1016/j.acags.2024.100201

Zixiong Shen , Qiming Sun , Xinyu Lu , Fenghua Ling , Yue Li , Jiye Wu , Jing-Jia Luo , Chaoxia Yuan

The application of machine learning (ML) techniques to climate science has received significant attention, particularly in the field of climate predictions, ranging from sub-seasonal to decadal time scales. This paper reviews recent progress of ML techniques employed in climate phenomena prediction and the enhancement of dynamic forecast models, which provide valuable insights into the great potentials of ML techniques to improve climate prediction capabilities with reduced computational time and resource consumption. This paper also discusses several major challenges in the application of ML to climate prediction, including the scarcity of datasets, physical inconsistency, and lack of model transparency and interpretability. Additionally, this paper sheds light on how climate change impacts ML model training and prediction, and explores three key areas with potential breakthroughs: large-scale climate models, knowledge discovery driven by ML, and hybrid dynamical-statistical models, underscoring the important role of the integration of “ML and dynamical models” in building a bridge between the artificial intelligence and climate science.

机器学习（ML）技术在气候科学中的应用，特别是在气候预测领域（从亚季节到十年时间尺度）的应用，已受到极大关注。本文回顾了机器学习技术在气候现象预测和动态预报模式增强方面的最新进展，这些进展提供了宝贵的见解，说明了机器学习技术在减少计算时间和资源消耗、提高气候预测能力方面的巨大潜力。本文还讨论了将 ML 应用于气候预测的几个主要挑战，包括数据集稀缺、物理不一致性以及模型缺乏透明度和可解释性。此外，本文还揭示了气候变化对 ML 模型训练和预测的影响，并探讨了可能取得突破的三个关键领域：大规模气候模型、ML 驱动的知识发现以及动态-统计混合模型，强调了 "ML 与动态模型 "的集成在搭建人工智能与气候科学之间桥梁的重要作用。

{"title":"Current progress in subseasonal-to-decadal prediction based on machine learning","authors":"Zixiong Shen , Qiming Sun , Xinyu Lu , Fenghua Ling , Yue Li , Jiye Wu , Jing-Jia Luo , Chaoxia Yuan","doi":"10.1016/j.acags.2024.100201","DOIUrl":"10.1016/j.acags.2024.100201","url":null,"abstract":"<div><div>The application of machine learning (ML) techniques to climate science has received significant attention, particularly in the field of climate predictions, ranging from sub-seasonal to decadal time scales. This paper reviews recent progress of ML techniques employed in climate phenomena prediction and the enhancement of dynamic forecast models, which provide valuable insights into the great potentials of ML techniques to improve climate prediction capabilities with reduced computational time and resource consumption. This paper also discusses several major challenges in the application of ML to climate prediction, including the scarcity of datasets, physical inconsistency, and lack of model transparency and interpretability. Additionally, this paper sheds light on how climate change impacts ML model training and prediction, and explores three key areas with potential breakthroughs: large-scale climate models, knowledge discovery driven by ML, and hybrid dynamical-statistical models, underscoring the important role of the integration of “ML and dynamical models” in building a bridge between the artificial intelligence and climate science.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"24 ","pages":"Article 100201"},"PeriodicalIF":2.6,"publicationDate":"2024-10-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142534340","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Novel application of unsupervised machine learning for characterization of subsurface seismicity, tectonic dynamics and stress distribution 无监督机器学习在表征地下地震活动性、构造动力学和应力分布方面的新应用

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2024-10-21 DOI: 10.1016/j.acags.2024.100200

Mohammad Salam, Muhammad Tahir Iqbal, Raja Adnan Habib, Amna Tahir, Aamir Sultan, Talat Iqbal

Our study pioneers an innovative use of unsupervised machine learning, a powerful tool for navigating unclassified data, to unravel the complexities of subsurface seismic activities and extract meaningful patterns. Our central objective is to comprehensively characterize seismicity within an active region by identifying distinct seismic clusters in spatial distribution, thereby gaining a deeper understanding of subsurface stress distribution and tectonic dynamics. Employing a diverse range of clustering algorithms, with particular emphasis on Fuzzy C-Means (FCM), our research meticulously dissects the intricate physical processes that govern a complex tectonic zone. This technique effectively delineates distinct tectonic zones, aligning seamlessly with established seismological knowledge and underscoring the transformative potential of Artificial Intelligence (AI) in analyzing regional subsurface phenomena, even under conditions of data scarcity. Moreover, associating earthquakes with specific seismogenic structures significantly enhances seismic hazard analyses, potentially paving the way for autonomous insights that inform engineering hazard assessments.

我们的研究开创性地使用了无监督机器学习这一用于浏览未分类数据的强大工具，以揭示地下地震活动的复杂性并提取有意义的模式。我们的核心目标是通过识别空间分布中不同的地震群，全面描述活跃区域内的地震活动特征，从而更深入地了解地下应力分布和构造动态。我们的研究采用了多种聚类算法，尤其侧重于模糊 C-Means (FCM)，细致地剖析了支配复杂构造带的错综复杂的物理过程。这项技术有效地划分了不同的构造带，与既有的地震学知识完美地结合在一起，并强调了人工智能（AI）在分析区域地下现象方面的变革潜力，即使在数据匮乏的条件下也是如此。此外，将地震与特定的成震结构联系起来，大大增强了地震灾害分析的效果，有可能为自主洞察力铺平道路，为工程灾害评估提供依据。

{"title":"Novel application of unsupervised machine learning for characterization of subsurface seismicity, tectonic dynamics and stress distribution","authors":"Mohammad Salam, Muhammad Tahir Iqbal, Raja Adnan Habib, Amna Tahir, Aamir Sultan, Talat Iqbal","doi":"10.1016/j.acags.2024.100200","DOIUrl":"10.1016/j.acags.2024.100200","url":null,"abstract":"<div><div>Our study pioneers an innovative use of unsupervised machine learning, a powerful tool for navigating unclassified data, to unravel the complexities of subsurface seismic activities and extract meaningful patterns. Our central objective is to comprehensively characterize seismicity within an active region by identifying distinct seismic clusters in spatial distribution, thereby gaining a deeper understanding of subsurface stress distribution and tectonic dynamics. Employing a diverse range of clustering algorithms, with particular emphasis on Fuzzy C-Means (FCM), our research meticulously dissects the intricate physical processes that govern a complex tectonic zone. This technique effectively delineates distinct tectonic zones, aligning seamlessly with established seismological knowledge and underscoring the transformative potential of Artificial Intelligence (AI) in analyzing regional subsurface phenomena, even under conditions of data scarcity. Moreover, associating earthquakes with specific seismogenic structures significantly enhances seismic hazard analyses, potentially paving the way for autonomous insights that inform engineering hazard assessments.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"24 ","pages":"Article 100200"},"PeriodicalIF":2.6,"publicationDate":"2024-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142534339","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A comparative study on machine learning approaches for rock mass classification using drilling data 利用钻探数据进行岩体分类的机器学习方法比较研究

IF 2.6 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Applied Computing and Geosciences

Pub Date : 2024-10-05 DOI: 10.1016/j.acags.2024.100199

Tom F. Hansen , Georg H. Erharter , Zhongqiang Liu , Jim Torresen

Current rock engineering design in drill and blast tunnelling primarily relies on engineers' observational assessments. Measure While Drilling (MWD) data, a high-resolution sensor dataset collected during tunnel excavation, is underutilised, mainly serving for geological visualisation. This study aims to automate the translation of MWD data into actionable metrics for rock engineering. It seeks to link data to specific engineering actions, thus providing critical decision support for geological challenges ahead of the tunnel face. Leveraging a large and geologically diverse dataset of ∼500,000 drillholes from 15 tunnels, the research introduces models for accurate rock mass quality classification in a real-world tunnelling context. Both conventional machine learning and image-based deep learning are explored to classify MWD data into Q-classes and Q-values—examples of metrics describing the stability of the rock mass—using both tabular- and image data. The results indicate that the K-nearest neighbours algorithm in an ensemble with tree-based models using tabular data effectively classifies rock mass quality. It achieves a cross-validated balanced accuracy of 0.86 in classifying rock mass into the Q-classes A, B, C, D, E1, E2, and 0.95 for a binary classification with E versus the rest. Classification using a CNN with MWD-images for each blasting round resulted in a balanced accuracy of 0.82 for binary classification. Regressing the Q-value from tabular MWD-data achieved cross-validated R² and MSE scores of 0.80 and 0.18 for a similar ensemble model as in classification. High performance in regression and classification boosts confidence in automated rock mass assessment. Applying advanced modelling on a unique dataset demonstrates MWD data's value in improving rock mass classification accuracy and advancing data-driven rock engineering design, reducing manual intervention.

目前钻爆法隧道的岩石工程设计主要依靠工程师的观察评估。边钻边测（MWD）数据是一种在隧道开挖过程中收集的高分辨率传感器数据集，但未得到充分利用，主要用于地质可视化。本研究旨在将 MWD 数据自动转化为岩石工程的可操作指标。它旨在将数据与具体的工程行动联系起来，从而为隧道工作面前方的地质挑战提供重要的决策支持。该研究利用来自 15 座隧道的 500,000 个钻孔组成的大型地质多样性数据集，引入了在现实世界隧道工程中对岩体质量进行准确分类的模型。研究探索了传统的机器学习和基于图像的深度学习，利用表格和图像数据将 MWD 数据划分为 Q 类和 Q 值（描述岩体稳定性的指标实例）。结果表明，在使用表格数据的树状模型的集合中，K-近邻算法能有效地对岩体质量进行分类。在将岩体划分为 Q 类 A、B、C、D、E1、E2 时，其交叉验证平衡准确率为 0.86，而将 E 与其他岩体进行二元分类的准确率为 0.95。使用带有每轮爆破的 MWD 图像的 CNN 进行分类，二元分类的均衡准确率为 0.82。通过对表格式 MWD 数据的 Q 值进行回归分析，在与分类类似的集合模型中，交叉验证的 R2 和 MSE 分别为 0.80 和 0.18。回归和分类的高性能增强了对岩体自动评估的信心。在一个独特的数据集上应用先进的建模方法，证明了 MWD 数据在提高岩体分类准确性、推进数据驱动的岩石工程设计、减少人工干预方面的价值。

{"title":"A comparative study on machine learning approaches for rock mass classification using drilling data","authors":"Tom F. Hansen , Georg H. Erharter , Zhongqiang Liu , Jim Torresen","doi":"10.1016/j.acags.2024.100199","DOIUrl":"10.1016/j.acags.2024.100199","url":null,"abstract":"<div><div>Current rock engineering design in drill and blast tunnelling primarily relies on engineers' observational assessments. Measure While Drilling (MWD) data, a high-resolution sensor dataset collected during tunnel excavation, is underutilised, mainly serving for geological visualisation. This study aims to automate the translation of MWD data into actionable metrics for rock engineering. It seeks to link data to specific engineering actions, thus providing critical decision support for geological challenges ahead of the tunnel face. Leveraging a large and geologically diverse dataset of ∼500,000 drillholes from 15 tunnels, the research introduces models for accurate rock mass quality classification in a real-world tunnelling context. Both conventional machine learning and image-based deep learning are explored to classify MWD data into Q-classes and Q-values—examples of metrics describing the stability of the rock mass—using both tabular- and image data. The results indicate that the K-nearest neighbours algorithm in an ensemble with tree-based models using tabular data effectively classifies rock mass quality. It achieves a cross-validated balanced accuracy of 0.86 in classifying rock mass into the Q-classes A, B, C, D, E1, E2, and 0.95 for a binary classification with E versus the rest. Classification using a CNN with MWD-images for each blasting round resulted in a balanced accuracy of 0.82 for binary classification. Regressing the Q-value from tabular MWD-data achieved cross-validated R<sup>2</sup> and MSE scores of 0.80 and 0.18 for a similar ensemble model as in classification. High performance in regression and classification boosts confidence in automated rock mass assessment. Applying advanced modelling on a unique dataset demonstrates MWD data's value in improving rock mass classification accuracy and advancing data-driven rock engineering design, reducing manual intervention.</div></div>","PeriodicalId":33804,"journal":{"name":"Applied Computing and Geosciences","volume":"24 ","pages":"Article 100199"},"PeriodicalIF":2.6,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142534338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0