It was recently established that self-healing slip pulses under uniform prestress $tau_b$ are unstable frictional rupture modes, i.e., they either slowly expand/decay with time t. Furthermore, their dynamics were shown to follow a reduced-dimensionality description corresponding to a special $L(c)$ line in a plane defined by the pulse propagation velocity $c(t)$ and size $L(t)$. Yet, uniform prestress is rather the exception than the rule in natural faults. We study the effects of a spatially-varying prestress $tau_b(x)$ on 2D slip pulses, initially generated under a uniform $tau_b$ along a rate-and-state friction fault. We consider periodic and constant-gradient prestress $tau_b(x)$ around the reference uniform $tau_b$. For a periodic $tau_b(x)$, pulses either sustain and form quasi-limit cycles in the $L-c$ plane or decay predominantly monotonically along the $L(c)$ line, depending on the instability index of the initial pulse and the properties of the periodic $tau_b(x)$. For a constant-gradient $tau_b(x)$, expanding/decaying pulses closely follow the $L(c)$ line, with systematic shifts determined by the sign and magnitude of the gradient. We also find that a spatially-varying $tau_b(x)$ can revert the expanding/decaying nature of the initial reference pulse. Finally, we show that a constant-gradient $tau_b(x)$, of sufficient magnitude and specific sign, can lead to the nucleation of a back-propagating rupture at the healing tail of the initial pulse, generating a bilateral crack-like rupture. This pulse-to-crack transition, along with the above-described effects, demonstrate that rich rupture dynamics merge from a simple, nonuniform prestress. Furthermore, we show that as long as pulses exist, their dynamics are related to the special $L(c)$ line, providing an effective, reduced-dimensionality description of unsteady slip pulses under spatially-varying prestress.
{"title":"Unsteady slip pulses under spatially-varying prestress","authors":"Anna Pomyalov, Eran Bouchbinder","doi":"arxiv-2407.21539","DOIUrl":"https://doi.org/arxiv-2407.21539","url":null,"abstract":"It was recently established that self-healing slip pulses under uniform\u0000prestress $tau_b$ are unstable frictional rupture modes, i.e., they either\u0000slowly expand/decay with time t. Furthermore, their dynamics were shown to\u0000follow a reduced-dimensionality description corresponding to a special $L(c)$\u0000line in a plane defined by the pulse propagation velocity $c(t)$ and size\u0000$L(t)$. Yet, uniform prestress is rather the exception than the rule in natural\u0000faults. We study the effects of a spatially-varying prestress $tau_b(x)$ on 2D\u0000slip pulses, initially generated under a uniform $tau_b$ along a\u0000rate-and-state friction fault. We consider periodic and constant-gradient\u0000prestress $tau_b(x)$ around the reference uniform $tau_b$. For a periodic\u0000$tau_b(x)$, pulses either sustain and form quasi-limit cycles in the $L-c$\u0000plane or decay predominantly monotonically along the $L(c)$ line, depending on\u0000the instability index of the initial pulse and the properties of the periodic\u0000$tau_b(x)$. For a constant-gradient $tau_b(x)$, expanding/decaying pulses\u0000closely follow the $L(c)$ line, with systematic shifts determined by the sign\u0000and magnitude of the gradient. We also find that a spatially-varying\u0000$tau_b(x)$ can revert the expanding/decaying nature of the initial reference\u0000pulse. Finally, we show that a constant-gradient $tau_b(x)$, of sufficient\u0000magnitude and specific sign, can lead to the nucleation of a back-propagating\u0000rupture at the healing tail of the initial pulse, generating a bilateral\u0000crack-like rupture. This pulse-to-crack transition, along with the\u0000above-described effects, demonstrate that rich rupture dynamics merge from a\u0000simple, nonuniform prestress. Furthermore, we show that as long as pulses\u0000exist, their dynamics are related to the special $L(c)$ line, providing an\u0000effective, reduced-dimensionality description of unsteady slip pulses under\u0000spatially-varying prestress.","PeriodicalId":501270,"journal":{"name":"arXiv - PHYS - Geophysics","volume":"34 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141870690","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}
Seismic imaging from sparsely acquired data faces challenges such as low image quality, discontinuities, and migration swing artifacts. Existing convolutional neural network (CNN)-based methods struggle with complex feature distributions and cannot effectively assess uncertainty, making it hard to evaluate the reliability of their processed results. To address these issues, we propose a new method using a generative diffusion model (GDM). Here, in the training phase, we use the imaging results from sparse data as conditional input, combined with noisy versions of dense data imaging results, for the network to predict the added noise. After training, the network can predict the imaging results for test images from sparse data acquisition, using the generative process with conditional control. This GDM not only improves image quality and removes artifacts caused by sparse data, but also naturally evaluates uncertainty by leveraging the probabilistic nature of the GDM. To overcome the decline in generation quality and the memory burden of large-scale images, we develop a patch fusion strategy that effectively addresses these issues. Synthetic and field data examples demonstrate that our method significantly enhances imaging quality and provides effective uncertainty quantification.
{"title":"Generative Diffusion Model for Seismic Imaging Improvement of Sparsely Acquired Data and Uncertainty Quantification","authors":"Xingchen Shi, Shijun Cheng, Weijian Mao, Wei Ouyang","doi":"arxiv-2407.21683","DOIUrl":"https://doi.org/arxiv-2407.21683","url":null,"abstract":"Seismic imaging from sparsely acquired data faces challenges such as low\u0000image quality, discontinuities, and migration swing artifacts. Existing\u0000convolutional neural network (CNN)-based methods struggle with complex feature\u0000distributions and cannot effectively assess uncertainty, making it hard to\u0000evaluate the reliability of their processed results. To address these issues,\u0000we propose a new method using a generative diffusion model (GDM). Here, in the\u0000training phase, we use the imaging results from sparse data as conditional\u0000input, combined with noisy versions of dense data imaging results, for the\u0000network to predict the added noise. After training, the network can predict the\u0000imaging results for test images from sparse data acquisition, using the\u0000generative process with conditional control. This GDM not only improves image\u0000quality and removes artifacts caused by sparse data, but also naturally\u0000evaluates uncertainty by leveraging the probabilistic nature of the GDM. To\u0000overcome the decline in generation quality and the memory burden of large-scale\u0000images, we develop a patch fusion strategy that effectively addresses these\u0000issues. Synthetic and field data examples demonstrate that our method\u0000significantly enhances imaging quality and provides effective uncertainty\u0000quantification.","PeriodicalId":501270,"journal":{"name":"arXiv - PHYS - Geophysics","volume":"74 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141870685","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}
Seismic tomography is a crucial technique used to image subsurface structures at various scales, accomplished by solving a nonlinear and nonunique inverse problem. It is therefore important to quantify velocity model uncertainties for accurate earthquake locations and geological interpretations. Monte Carlo sampling techniques are usually used for this purpose, but those methods are computationally intensive, especially for large datasets or high-dimensional parameter spaces. In comparison, Bayesian variational inference provides a more efficient alternative by delivering probabilistic solutions through optimization. The method has been proven to be efficient in 2D tomographic problems. In this study, we apply variational inference to solve 3D double-difference (DD) seismic tomographic system using both absolute and differential travel time data. Synthetic tests demonstrate that the new method can produce more accurate velocity models than the original DD tomography method by avoiding regularization constraints, and at the same time provides more reliable uncertainty estimates. Compared to traditional checkerboard resolution tests, the resulting uncertainty estimates measure more accurately the reliability of the solution. We further apply the new method to data recorded by a local dense seismic array around the San Andreas Fault Observatory at Depth (SAFOD) site along the San Andreas Fault (SAF) at Parkfield. Similar to other researches, the obtained velocity models show significant velocity contrasts across the fault. More importantly, the new method produces velocity uncertainties of less than 0.34 km/s for Vp and 0.23 km/s for Vs. We therefore conclude that variational inference provides a powerful and efficient tool for solving 3D seismic tomographic problems and quantifying model uncertainties.
{"title":"3D Variational Inference-Based Double-Difference Seismic Tomography Method and Application to the SAFOD Site, California","authors":"Hao Yang, Xin Zhang, Haijiang Zhang","doi":"arxiv-2407.21405","DOIUrl":"https://doi.org/arxiv-2407.21405","url":null,"abstract":"Seismic tomography is a crucial technique used to image subsurface structures\u0000at various scales, accomplished by solving a nonlinear and nonunique inverse\u0000problem. It is therefore important to quantify velocity model uncertainties for\u0000accurate earthquake locations and geological interpretations. Monte Carlo\u0000sampling techniques are usually used for this purpose, but those methods are\u0000computationally intensive, especially for large datasets or high-dimensional\u0000parameter spaces. In comparison, Bayesian variational inference provides a more\u0000efficient alternative by delivering probabilistic solutions through\u0000optimization. The method has been proven to be efficient in 2D tomographic\u0000problems. In this study, we apply variational inference to solve 3D\u0000double-difference (DD) seismic tomographic system using both absolute and\u0000differential travel time data. Synthetic tests demonstrate that the new method\u0000can produce more accurate velocity models than the original DD tomography\u0000method by avoiding regularization constraints, and at the same time provides\u0000more reliable uncertainty estimates. Compared to traditional checkerboard\u0000resolution tests, the resulting uncertainty estimates measure more accurately\u0000the reliability of the solution. We further apply the new method to data\u0000recorded by a local dense seismic array around the San Andreas Fault\u0000Observatory at Depth (SAFOD) site along the San Andreas Fault (SAF) at\u0000Parkfield. Similar to other researches, the obtained velocity models show\u0000significant velocity contrasts across the fault. More importantly, the new\u0000method produces velocity uncertainties of less than 0.34 km/s for Vp and 0.23\u0000km/s for Vs. We therefore conclude that variational inference provides a\u0000powerful and efficient tool for solving 3D seismic tomographic problems and\u0000quantifying model uncertainties.","PeriodicalId":501270,"journal":{"name":"arXiv - PHYS - Geophysics","volume":"21 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141880844","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}
Sergey N. Britvin, Oleg S. Vereshchagin, Natalia S. Vlasenko, Maria G. Krzhizhanovskaya, Marina A. Ivanova
The lack of benchmark data on the real minerals, native ammonium carriers in Solar System gives rise to controversial opinions on extraterrestrial ammonium reservoirs. We herein report on discovery of the first mineral carrier of meteoritic ammonium and show its relevance to the compositional and spectral characteristics of cometary and asteroidal bodies. Chemically distant from previously inferred volatile organics or ammoniated phyllosilicates, it is an aqueous metal-ammonium sulfate related to a family of so-called Tutton salts. Nickeloan boussingaultite, (NH4)2(Mg,Ni)(SO4)2 6H2O, occurs in Orgueil, a primitive carbonaceous chondrite closely related to (162173) Ryugu and (101955) Bennu, the C-type asteroids. The available spectroscopic, chemical and mineralogical data signify that natural Tutton salts perfectly fit into the role of ammonium reservoir under conditions of cometary nuclei and carbonaceous asteroids.
{"title":"Meteoritic Tutton salt, a naturally inspired reservoir of cometary and asteroidal ammonium","authors":"Sergey N. Britvin, Oleg S. Vereshchagin, Natalia S. Vlasenko, Maria G. Krzhizhanovskaya, Marina A. Ivanova","doi":"arxiv-2407.20997","DOIUrl":"https://doi.org/arxiv-2407.20997","url":null,"abstract":"The lack of benchmark data on the real minerals, native ammonium carriers in\u0000Solar System gives rise to controversial opinions on extraterrestrial ammonium\u0000reservoirs. We herein report on discovery of the first mineral carrier of\u0000meteoritic ammonium and show its relevance to the compositional and spectral\u0000characteristics of cometary and asteroidal bodies. Chemically distant from\u0000previously inferred volatile organics or ammoniated phyllosilicates, it is an\u0000aqueous metal-ammonium sulfate related to a family of so-called Tutton salts.\u0000Nickeloan boussingaultite, (NH4)2(Mg,Ni)(SO4)2 6H2O, occurs in Orgueil, a\u0000primitive carbonaceous chondrite closely related to (162173) Ryugu and (101955)\u0000Bennu, the C-type asteroids. The available spectroscopic, chemical and\u0000mineralogical data signify that natural Tutton salts perfectly fit into the\u0000role of ammonium reservoir under conditions of cometary nuclei and carbonaceous\u0000asteroids.","PeriodicalId":501270,"journal":{"name":"arXiv - PHYS - Geophysics","volume":"20 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141870691","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}
We develop a software package libEMMI_MGFD for 3D frequency-domain marine controlled-source electromagnetic (CSEM) modelling and inversion. It is the first open-source C program tailored for geometrical multigrid (GMG) CSEM simulation. An volumetric anisotropic averaging scheme has been employed to compute effective medium for modelling over uniform and nonuniform grid. The computing coordinate is aligned with acquisition geometry by rotation with the azimuth and dip angles, facilitating the injection of the source and the extraction of data with arbitrary orientations. Efficient nonlinear optimization is achieved using quasi-Newton scheme assisted with bisection backtracking line search. In constructing the modularized Maxwell solver and evaluating the misfit and gradient for 3D CSEM inversion, the reverse communication technique is the key to the compaction of the software while maintaining the computational performance. A number of numeric tests demonstrate the efficiency of the modelling while preserving the solution accuracy. A 3D marine CSEM inversion example been been examined for resistivity imaging.
我们开发了一个用于三维频域海洋可控源电磁（CSEM）建模和反演的软件包 libEMMI_MGFD 。这是第一个为几何多网格（GMG）CSEM 模拟定制的开源 C 程序。它采用了体积各向异性平均方案来计算有效介质，以便在均匀和非均匀网格上建模。计算坐标通过与方位角和倾角的旋转与采集几何对齐，便于注入源和提取任意方向的数据。采用准牛顿方案，并辅以分段回溯线搜索，实现了高效的非线性优化。在构建模块化麦克斯韦求解器和评估三维 CSEM 反演的失配和梯度时，反向通信技术是在保持计算性能的同时压缩软件的关键。大量数值测试证明了建模的效率，同时保持了求解精度。针对电阻率成像研究了三维海洋 CSEM 反演实例。
{"title":"libEMMI_MGFD: A program of marine controlled-source electromagnetic modelling and inversion using frequency-domain multigrid solver","authors":"Pengliang Yang, An Ping","doi":"arxiv-2407.20795","DOIUrl":"https://doi.org/arxiv-2407.20795","url":null,"abstract":"We develop a software package libEMMI_MGFD for 3D frequency-domain marine\u0000controlled-source electromagnetic (CSEM) modelling and inversion. It is the\u0000first open-source C program tailored for geometrical multigrid (GMG) CSEM\u0000simulation. An volumetric anisotropic averaging scheme has been employed to\u0000compute effective medium for modelling over uniform and nonuniform grid. The\u0000computing coordinate is aligned with acquisition geometry by rotation with the\u0000azimuth and dip angles, facilitating the injection of the source and the\u0000extraction of data with arbitrary orientations. Efficient nonlinear\u0000optimization is achieved using quasi-Newton scheme assisted with bisection\u0000backtracking line search. In constructing the modularized Maxwell solver and\u0000evaluating the misfit and gradient for 3D CSEM inversion, the reverse\u0000communication technique is the key to the compaction of the software while\u0000maintaining the computational performance. A number of numeric tests\u0000demonstrate the efficiency of the modelling while preserving the solution\u0000accuracy. A 3D marine CSEM inversion example been been examined for resistivity\u0000imaging.","PeriodicalId":501270,"journal":{"name":"arXiv - PHYS - Geophysics","volume":"41 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141870689","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}
Process-based hydrologic models are invaluable tools for understanding the terrestrial water cycle and addressing modern water resources problems. However, many hydrologic models are computationally expensive and, depending on the resolution and scale, simulations can take on the order of hours to days to complete. While techniques such as uncertainty quantification and optimization have become valuable tools for supporting management decisions, these analyses typically require hundreds of model simulations, which are too computationally expensive to perform with a process-based hydrologic model. To address this gap, we propose a hybrid modeling workflow in which a process-based model is used to generate an initial set of simulations and a machine learning (ML) surrogate model is then trained to perform the remaining simulations required for downstream analysis. As a case study, we apply this workflow to simulations of variably saturated groundwater flow at a prospective managed aquifer recharge (MAR) site. We compare the accuracy and computational efficiency of several ML architectures, including deep convolutional networks, recurrent neural networks, vision transformers, and networks with Fourier transforms. Our results demonstrate that ML surrogate models can achieve under 10% mean absolute percentage error and yield order-of-magnitude runtime savings over processed-based models. We also offer practical recommendations for training hydrologic surrogate models, including implementing data normalization to improve accuracy, using a normalized loss function to improve training stability and downsampling input features to decrease memory requirements.
基于过程的水文模型是了解陆地水循环和解决现代水资源问题的宝贵工具。然而，许多水文模型的计算成本很高，根据分辨率和规模的不同，模拟可能需要数小时至数天才能完成。虽然不确定性量化和优化等技术已成为支持管理决策的重要工具，但这些分析通常需要数百次模型模拟，而基于过程的水文模型的计算成本太高。为了弥补这一差距，我们提出了一种混合建模工作流程，即使用基于过程的模型生成初始模拟集，然后训练机器学习（ML）代理模型来执行下游分析所需的剩余模拟。作为案例研究，我们将这一工作流程应用于模拟一个潜在的有管理含水层补给（MAR）地点的可变饱和地下水流。我们比较了深度卷积网络、循环神经网络、视觉变换器和傅立叶变换网络等多种 ML 架构的准确性和计算效率。我们的研究结果表明，与基于处理的模型相比，ML 代用模型的平均绝对百分比误差低于 10%，并能节省数量级的运行时间。我们还为训练水文代用模型提供了实用建议，包括实施数据归一化以提高准确性，使用归一化损失函数以提高训练稳定性，以及降低输入特征采样以减少内存需求。
{"title":"Machine learning surrogates for efficient hydrologic modeling: Insights from stochastic simulations of managed aquifer recharge","authors":"Timothy Dai, Kate Maher, Zach Perzan","doi":"arxiv-2407.20902","DOIUrl":"https://doi.org/arxiv-2407.20902","url":null,"abstract":"Process-based hydrologic models are invaluable tools for understanding the\u0000terrestrial water cycle and addressing modern water resources problems.\u0000However, many hydrologic models are computationally expensive and, depending on\u0000the resolution and scale, simulations can take on the order of hours to days to\u0000complete. While techniques such as uncertainty quantification and optimization\u0000have become valuable tools for supporting management decisions, these analyses\u0000typically require hundreds of model simulations, which are too computationally\u0000expensive to perform with a process-based hydrologic model. To address this\u0000gap, we propose a hybrid modeling workflow in which a process-based model is\u0000used to generate an initial set of simulations and a machine learning (ML)\u0000surrogate model is then trained to perform the remaining simulations required\u0000for downstream analysis. As a case study, we apply this workflow to simulations\u0000of variably saturated groundwater flow at a prospective managed aquifer\u0000recharge (MAR) site. We compare the accuracy and computational efficiency of\u0000several ML architectures, including deep convolutional networks, recurrent\u0000neural networks, vision transformers, and networks with Fourier transforms. Our\u0000results demonstrate that ML surrogate models can achieve under 10% mean\u0000absolute percentage error and yield order-of-magnitude runtime savings over\u0000processed-based models. We also offer practical recommendations for training\u0000hydrologic surrogate models, including implementing data normalization to\u0000improve accuracy, using a normalized loss function to improve training\u0000stability and downsampling input features to decrease memory requirements.","PeriodicalId":501270,"journal":{"name":"arXiv - PHYS - Geophysics","volume":"213 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141870686","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}
We present a theoretical investigation into the dynamics of a viscous gravity current subjected to spatially-finite lubrication (i.e., a `slippery patch'). The work is motivated by grounded ice sheets flowing across patches of basal meltwater which reduce the ice-bed frictional coupling, causing perturbations enhancing ice motion, with implications for increased ice flux into the ocean and sea level rise. The flow is characterized by transitions between shear- and extension-dominated dynamics, which necessitates boundary-layer solutions at the transition points. We develop a depth-integrated analytical model of Newtonian flow which concisely reveals fundamental relationships between ice sheet geometry (thickness, surface slope, and slippery patch length) and the magnitude and spatial extent of resulting horizontal deviatoric stresses. This reduced-order analytical model shows good quantitative agreement with numerical simulations using 2-D Newtonian Stokes equations, which are further extended to the case of a non-Newtonian flow. From the reduced-order model, we rationalize that the slippery patch-induced stress perturbations are exponentially-decaying functions of distance upstream away from the patch onset. We also show that the amplitude of the perturbation scales linearly with the surface slope and patch length while the decay lengthscale scales linearly with ice thickness. These fundamental relationships have implications for the response of the Greenland Ice Sheet to the inland expansion of basal meltwater presence over the coming warming decades.
{"title":"Theoretical analysis of stress perturbations from a partially-lubricated viscous gravity current","authors":"Joshua H. Rines, Ching-Yao Lai, Yongji Wang","doi":"arxiv-2407.20565","DOIUrl":"https://doi.org/arxiv-2407.20565","url":null,"abstract":"We present a theoretical investigation into the dynamics of a viscous gravity\u0000current subjected to spatially-finite lubrication (i.e., a `slippery patch').\u0000The work is motivated by grounded ice sheets flowing across patches of basal\u0000meltwater which reduce the ice-bed frictional coupling, causing perturbations\u0000enhancing ice motion, with implications for increased ice flux into the ocean\u0000and sea level rise. The flow is characterized by transitions between shear- and\u0000extension-dominated dynamics, which necessitates boundary-layer solutions at\u0000the transition points. We develop a depth-integrated analytical model of\u0000Newtonian flow which concisely reveals fundamental relationships between ice\u0000sheet geometry (thickness, surface slope, and slippery patch length) and the\u0000magnitude and spatial extent of resulting horizontal deviatoric stresses. This\u0000reduced-order analytical model shows good quantitative agreement with numerical\u0000simulations using 2-D Newtonian Stokes equations, which are further extended to\u0000the case of a non-Newtonian flow. From the reduced-order model, we rationalize\u0000that the slippery patch-induced stress perturbations are exponentially-decaying\u0000functions of distance upstream away from the patch onset. We also show that the\u0000amplitude of the perturbation scales linearly with the surface slope and patch\u0000length while the decay lengthscale scales linearly with ice thickness. These\u0000fundamental relationships have implications for the response of the Greenland\u0000Ice Sheet to the inland expansion of basal meltwater presence over the coming\u0000warming decades.","PeriodicalId":501270,"journal":{"name":"arXiv - PHYS - Geophysics","volume":"126 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141870695","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}
Anthony Jourdon, Jorge Nicolas Hayek, Dave A. May, Alice-Agnes Gabriel
Tectonic deformation crucially shapes the Earth's surface, with strain localization resulting in the formation of shear zones and faults that accommodate significant tectonic displacement. Earthquake dynamic rupture models, which provide valuable insights into earthquake mechanics and seismic ground motions, rely on initial conditions such as pre-stress states and fault geometry. However, these are often inadequately constrained due to observational limitations. To address these challenges, we develop a new method that loosely couples 3D geodynamic models to 3D dynamic rupture simulations, providing a mechanically consistent framework for earthquake analysis. Our approach does not prescribe fault geometry but derives it from the underlying lithospheric rheology and tectonic velocities using the medial axis transform. We perform three long-term geodynamics models of a strike-slip geodynamic system, each involving different continental crust rheology. We link these with nine dynamic rupture models, in which we investigate the role of varying fracture energy and plastic strain energy dissipation in the dynamic rupture behavior. These simulations suggest that for our fault, long-term rheology, and geodynamic system, a plausible critical linear slip weakening distance falls within Dc in [0.6,1.5]. Our results indicate that the long-term 3D stress field favors slip on fault segments better aligned with the regional plate motion and that minor variations in the long-term 3D stress field can strongly affect rupture dynamics, providing a physical mechanism for arresting earthquake propagation. Our geodynamically informed earthquake models highlight the need for detailed 3D fault modeling across time scales for a comprehensive understanding of earthquake mechanics.
{"title":"Coupling 3D geodynamics and dynamic earthquake rupture: fault geometry, rheology and stresses across timescales","authors":"Anthony Jourdon, Jorge Nicolas Hayek, Dave A. May, Alice-Agnes Gabriel","doi":"arxiv-2407.20609","DOIUrl":"https://doi.org/arxiv-2407.20609","url":null,"abstract":"Tectonic deformation crucially shapes the Earth's surface, with strain\u0000localization resulting in the formation of shear zones and faults that\u0000accommodate significant tectonic displacement. Earthquake dynamic rupture\u0000models, which provide valuable insights into earthquake mechanics and seismic\u0000ground motions, rely on initial conditions such as pre-stress states and fault\u0000geometry. However, these are often inadequately constrained due to\u0000observational limitations. To address these challenges, we develop a new method\u0000that loosely couples 3D geodynamic models to 3D dynamic rupture simulations,\u0000providing a mechanically consistent framework for earthquake analysis. Our\u0000approach does not prescribe fault geometry but derives it from the underlying\u0000lithospheric rheology and tectonic velocities using the medial axis transform.\u0000We perform three long-term geodynamics models of a strike-slip geodynamic\u0000system, each involving different continental crust rheology. We link these with\u0000nine dynamic rupture models, in which we investigate the role of varying\u0000fracture energy and plastic strain energy dissipation in the dynamic rupture\u0000behavior. These simulations suggest that for our fault, long-term rheology, and\u0000geodynamic system, a plausible critical linear slip weakening distance falls\u0000within Dc in [0.6,1.5]. Our results indicate that the long-term 3D stress field\u0000favors slip on fault segments better aligned with the regional plate motion and\u0000that minor variations in the long-term 3D stress field can strongly affect\u0000rupture dynamics, providing a physical mechanism for arresting earthquake\u0000propagation. Our geodynamically informed earthquake models highlight the need\u0000for detailed 3D fault modeling across time scales for a comprehensive\u0000understanding of earthquake mechanics.","PeriodicalId":501270,"journal":{"name":"arXiv - PHYS - Geophysics","volume":"52 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141870687","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}
The Izu-Tobu region, on the eastern side of Izu Peninsula, in Japan, is volcanically and seismically active. In this region, earthquake swarms of ordinary earthquakes frequently occur at shallow depths, which is considered to be associated with magma intrusion. Beneath ordinary earthquakes, low-frequency earthquakes (LFEs) are infrequently observed. We conducted a timeseries analysis of both types of earthquakes during the time period 2005-2020, using a variant of the Epidemic-Type Aftershock Sequence model. For this analysis, we used the Japan Meteorological Agency catalog of ordinary earthquakes and the catalog of LFEs produced using the matched filter method. The observed result, which was common to both types of earthquakes, showed a significant change in seismicity, which became quiet, with the inflection point falling sometime between late 2009 and mid-2013, during which two out of three pronounced earthquake swarms occurred. We associated this seismic quiescence with changes in background rate to be low, where background rate, by removing the triggering effect of aftershocks, was interpreted as having been caused directly by the magma source, which can vary with time. We used surface displacement data obtained from the Geospatial Information Authority of Japan, and observed that the uplift due to magma intrusion was significant during the 1970s-1990s whereas it was in abatement or unobservable during the studied period (2005-2020). We also found that the seismic quiescence occurred without significant crustal movement during the studied period. Our implication from this finding is that magma source, which caused magma intrusion into the Izu-Tobu region, is in a transition phase, becoming less active, compared with the magma source during the 1970s-1990s.
{"title":"Changes in seismicity in a volcanically active region, on the eastern side of Izu Peninsula, Japan","authors":"K. Z. Nanjo, Y. Yukutake, T. Kumazawa","doi":"arxiv-2407.19648","DOIUrl":"https://doi.org/arxiv-2407.19648","url":null,"abstract":"The Izu-Tobu region, on the eastern side of Izu Peninsula, in Japan, is\u0000volcanically and seismically active. In this region, earthquake swarms of\u0000ordinary earthquakes frequently occur at shallow depths, which is considered to\u0000be associated with magma intrusion. Beneath ordinary earthquakes, low-frequency\u0000earthquakes (LFEs) are infrequently observed. We conducted a timeseries\u0000analysis of both types of earthquakes during the time period 2005-2020, using a\u0000variant of the Epidemic-Type Aftershock Sequence model. For this analysis, we\u0000used the Japan Meteorological Agency catalog of ordinary earthquakes and the\u0000catalog of LFEs produced using the matched filter method. The observed result,\u0000which was common to both types of earthquakes, showed a significant change in\u0000seismicity, which became quiet, with the inflection point falling sometime\u0000between late 2009 and mid-2013, during which two out of three pronounced\u0000earthquake swarms occurred. We associated this seismic quiescence with changes\u0000in background rate to be low, where background rate, by removing the triggering\u0000effect of aftershocks, was interpreted as having been caused directly by the\u0000magma source, which can vary with time. We used surface displacement data\u0000obtained from the Geospatial Information Authority of Japan, and observed that\u0000the uplift due to magma intrusion was significant during the 1970s-1990s\u0000whereas it was in abatement or unobservable during the studied period\u0000(2005-2020). We also found that the seismic quiescence occurred without\u0000significant crustal movement during the studied period. Our implication from\u0000this finding is that magma source, which caused magma intrusion into the\u0000Izu-Tobu region, is in a transition phase, becoming less active, compared with\u0000the magma source during the 1970s-1990s.","PeriodicalId":501270,"journal":{"name":"arXiv - PHYS - Geophysics","volume":"78 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141870692","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}
Whether it is oil and gas exploration or geological science research, it is necessary to accurately grasp the structural information of underground media. Full waveform inversion is currently the most popular seismic wave inversion method, but it is highly dependent on a high-quality initial model. Artificial intelligence algorithm deep learning is completely data-driven and can get rid of the dependence on the initial model. However, the prediction accuracy of deep learning algorithms depends on the scale and diversity of training data sets. How to improve the prediction accuracy of deep learning without increasing the size of the training set while also improving computing efficiency is a worthy issue to study. In this paper, an iterative deep learning algorithm in the sparse transform domain is proposed based on the characteristics of deep learning: first, based on the computational efficiency and the effect of sparse transform, the cosine transform is selected as the sparse transform method, and the seismic data and the corresponding velocity model are cosine transformed to obtain their corresponding sparse expressions, which are then used as the input data and corresponding label data for deep learning; then we give an iterative deep learning algorithm in the cosine transform domain, that is, after obtaining the seismic data residuals and velocity model residuals of the previous round of test results, they are used again as new input data and label data, and re-trained in the cosine domain to obtain a new network, and the prediction results of the previous round are corrected, and then the cycle is repeated until the termination condition is reached. The algorithm effect was verified on the SEG/EAGE salt model and the seabed sulfide physical model site data.
{"title":"Accurate background velocity model building method based on iterative deep learning in sparse transform domain","authors":"Guoxin Chen","doi":"arxiv-2407.19419","DOIUrl":"https://doi.org/arxiv-2407.19419","url":null,"abstract":"Whether it is oil and gas exploration or geological science research, it is\u0000necessary to accurately grasp the structural information of underground media.\u0000Full waveform inversion is currently the most popular seismic wave inversion\u0000method, but it is highly dependent on a high-quality initial model. Artificial\u0000intelligence algorithm deep learning is completely data-driven and can get rid\u0000of the dependence on the initial model. However, the prediction accuracy of\u0000deep learning algorithms depends on the scale and diversity of training data\u0000sets. How to improve the prediction accuracy of deep learning without\u0000increasing the size of the training set while also improving computing\u0000efficiency is a worthy issue to study. In this paper, an iterative deep\u0000learning algorithm in the sparse transform domain is proposed based on the\u0000characteristics of deep learning: first, based on the computational efficiency\u0000and the effect of sparse transform, the cosine transform is selected as the\u0000sparse transform method, and the seismic data and the corresponding velocity\u0000model are cosine transformed to obtain their corresponding sparse expressions,\u0000which are then used as the input data and corresponding label data for deep\u0000learning; then we give an iterative deep learning algorithm in the cosine\u0000transform domain, that is, after obtaining the seismic data residuals and\u0000velocity model residuals of the previous round of test results, they are used\u0000again as new input data and label data, and re-trained in the cosine domain to\u0000obtain a new network, and the prediction results of the previous round are\u0000corrected, and then the cycle is repeated until the termination condition is\u0000reached. The algorithm effect was verified on the SEG/EAGE salt model and the\u0000seabed sulfide physical model site data.","PeriodicalId":501270,"journal":{"name":"arXiv - PHYS - Geophysics","volume":"362 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141870693","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：