Article
Computer Science, Interdisciplinary Applications
Rogerio G. Negri, Alejandro C. Frery
Summary: This study proposes a novel framework for imagery dataset simulation, specifically designed for evaluating and comparing change detection methods. The framework is versatile, allowing the use of both supervised and unsupervised techniques. Applying this framework, the performance of well-known algorithms was compared using simulated data resembling deforestation in a forest area observed by the Landsat 5 TM sensor. The results provide insights into method performance and demonstrate the usefulness of the proposed framework.
COMPUTERS & GEOSCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
David Atienza, Concha Bielza, Pedro Larranaga
Summary: PyBNesian package provides an implementation of Bayesian network models to represent uncertainty in data, and it can be extended and interoperated with other components. It also implements other related models, such as kernel density estimation with GPU acceleration using OpenCL.
Article
Chemistry, Analytical
Mauro Enrique de Souza Munoz, Matheus Chaves Menezes, Edison Pignaton de Freitas, Sen Cheng, Paulo Rogerio de Almeida Ribeiro, Areolino de Almeida Neto, Alexandre Cesar Muniz de Oliveira
Summary: This article introduces SLAM technology and the xRatSLAM framework based on RatSLAM, and conducts testing and validation. The results show that xRatSLAM can generate maps similar to OpenRatSLAM, and framework components can be easily changed.
Article
Geosciences, Multidisciplinary
Zhen Liu, Junhua Zhang, Youzhuang Sun, Zhengjun Yu, Ruijun Ren
Summary: This paper presents a novel Bayesian-based method for predicting brittleness by synthesizing petrophysical data and applying Bayesian facies classification to seismic data. The method combines non-shale facies data with Rickman brittleness data to obtain a new brittleness index. The results suggest the potential applicability of the method in enhancing the characterization and understanding of geological formations.
FRONTIERS IN EARTH SCIENCE
(2023)
Article
Computer Science, Hardware & Architecture
Jianwei Zhang, Chenwei Zhao
Summary: The paper proposes a Q-SR framework to fully explore the potential of segment routing, with high extensibility for various network topologies and traffic matrices. It introduces offline and online algorithms, proven through theoretical bounds and extensive simulations to validate computation efficiency and feasibility.
Article
Engineering, Mechanical
Xinyu Jia, Costas Papadimitriou
Summary: A hierarchical Bayesian learning framework is proposed for multi-level modeling in structural dynamics, which can effectively quantify uncertainties at different modeling levels and propagate them through the system.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Medicine, General & Internal
Xu-Jing Yao, Zi-Quan Zhu, Shui-Hua Wang, Yu-Dong Zhang
Summary: The COVID-19 virus has caused significant impact globally since late 2019, prompting the need for more accurate detection methods. This research developed an efficient deep learning framework named CSGBBNet to improve COVID-19 detection using lung CT scans. Results showed high accuracy and outperformance of previous methods, merging biomedical research and AI in the field of COVID-19 detection.
Article
Engineering, Civil
Xue Jiang, Rui Ma, Yanxin Wang, Wenlong Gu, Wenxi Lu, Jin Na
Summary: This study proposes a new two-stage surrogate-assisted Markov chain Monte Carlo-based Bayesian framework for identifying contaminant source parameters in groundwater. An adaptive update feedback process and a multiobjective feasibility-enhanced particle swarm optimization algorithm are utilized to enhance the accuracy and efficiency of the framework.
JOURNAL OF HYDROLOGY
(2021)
Article
Geochemistry & Geophysics
Jeremie Giraud, Mark Lindsay, Mark Jessell
Summary: A inversion method using a regularized least-squares framework has been developed to recover the geometry of multiple geological units. By inverting for signed distances between units, the method allows for minimum adjustments of interfaces between rock units to fit the data. It provides flexibility in design and regularization, and has been tested with synthetic and field data scenarios.
Article
Biochemical Research Methods
Zheng Ye, Shaohao Li, Xue Mi, Baoyi Shao, Zhu Dai, Bo Ding, Songwei Feng, Bo Sun, Yang Shen, Zhongdang Xiao
Summary: STMHCPan is an open-source package based on the Star-Transformer model, which can predict the binding affinity of MHC I peptides. It exhibits improved performance in receptor affinity training compared to classical deep learning algorithms and can handle peptides of arbitrary length with high scalability.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Computer Science, Interdisciplinary Applications
V Gimenez-Alventosa, V Gimenez Gomez, S. Oliver
Summary: Monte Carlo methods offer precise results for radiation transport simulations, but their high computational cost limits real-time applications. PenRed, a standalone C++ framework, addresses these limitations by providing parallel Monte Carlo simulations with DICOM image processing capabilities for medical applications. It has been verified against the original PENELOPE code and demonstrates improved simulation times without sacrificing precision.
COMPUTER PHYSICS COMMUNICATIONS
(2021)
Article
Geochemistry & Geophysics
James Atterholt, Zachary E. Ross
Summary: The article presents a fully Bayesian inverse scheme to determine the second moments of stress glut using teleseismic earthquake seismograms. This method provides probabilistic descriptions of source rupture characteristics efficiently.
JOURNAL OF GEOPHYSICAL RESEARCH-SOLID EARTH
(2022)
Article
Geochemistry & Geophysics
Dario Grana, Leandro de Figueiredo, Klaus Mosegaard
Summary: Stochastic petrophysical inversion is a method to predict reservoir properties from seismic data, and recent advances in stochastic optimization allow generating multiple realizations of rock and fluid properties. This paper presents a Bayesian approach based on an efficient implementation of the Markov chain Monte Carlo (MCMC) method for seismic data inversion. The approach includes vertical and lateral correlation models, and is tested on 1D and extended to 2D problems. The results show the advantage of integrating a spatial correlation model.
Article
Geochemistry & Geophysics
Wei Xiang, Xingyao Yin, Zhengqian Ma, Kun Li, Song Pei
Summary: This article investigates the monoclinic medium with vertical fractures using numerical simulation and inversion methods. The fracture weaknesses and Thomsen parameters are successfully estimated by the linearized approximate formula and Bayesian inference. The results demonstrate the stability and feasibility of the proposed method in analyzing the problems of vertical fractures and background anisotropy in seismic data.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Computer Science, Artificial Intelligence
Carl Poelking, Felix A. Faber, Bingqing Cheng
Summary: We have developed a machine-learning framework for high-throughput benchmarking of different representations of chemical systems. The framework evaluates raw descriptor performance using simple regression schemes, adheres to best ML practices, and assesses learning progress through learning curves. By comparing the training outcome of various representations, we gain insights into their relative merits and interrelatedness.
MACHINE LEARNING-SCIENCE AND TECHNOLOGY
(2022)
Article
Water Resources
Hojat Ghorbanidehno, Jonghyun Lee, Matthew Farthing, Tyler Hesser, Eric F. Darve, Peter K. Kitanidis
Summary: Riverine bathymetry is important for shipping and flood management, and indirect measurements with sensor technology can be used to estimate river bed topography. Physics-based techniques are computationally expensive, while deep learning offers a data-driven approach with potential for efficient training using limited data. The proposed method combines DNN with PCA to image river bed topography using flow velocity observations, showing satisfactory performance in bathymetry estimation with low computational cost and small number of training samples.
ADVANCES IN WATER RESOURCES
(2021)
Article
Geography, Physical
C. J. Roland, L. K. Zoet, J. E. Rawling, M. Cardiff
Summary: The study indicates that freeze-thaw environmental factors have a significant impact on the erosion of coastal bluffs at seasonal timescales, leading to increased pore pressures and frequent mass wasting events. Seasonal upslope erosion is primarily influenced by rising water levels and freeze-thaw processes, necessitating the inclusion of these transient conditions in landscape change models.
Article
Engineering, Environmental
Mojtaba Forghani, Yizhou Qian, Jonghyun Lee, Matthew W. Farthing, Tyler Hesser, Peter K. Kitanidis, Eric F. Darve
Summary: This study proposes a two-stage process utilizing principal component geostatistical approach to estimate bathymetry probability density function and multiple machine learning algorithms to solve shallow water equations (SWEs) efficiently. The first stage predicts flow velocities without direct bathymetry measurement, while the second stage incorporates additional bathymetry information for improved accuracy and generalization. Fast solvers are capable of accurately predicting flow velocities with variable bathymetry and BCs at a significantly lower computational cost compared to traditional methods.
STOCHASTIC ENVIRONMENTAL RESEARCH AND RISK ASSESSMENT
(2021)
Article
Environmental Sciences
Xueyuan Kang, Amalia Kokkinaki, Peter K. Kitanidis, Xiaoqing Shi, Jonghyun Lee, Shaoxing Mo, Jichun Wu
Summary: Characterizing the architecture of dense nonaqueous phase liquid (DNAPL) source zones is crucial for designing efficient remediation strategies, but traditional drilling investigations provide limited information and affect the accuracy of geostatistical methods. By parameterizing the DNAPL saturation field using a physics-based approach, improved prior descriptions and better resolution can be achieved in characterizing the source zones. Additionally, incorporating hydrogeological and geophysical datasets in the inversion framework can further enhance the performance of the method.
WATER RESOURCES RESEARCH
(2021)
Article
Geosciences, Multidisciplinary
Jeremy R. Patterson, Michael Cardiff
Summary: Characterizing aquifer properties and their associated uncertainty is a challenge in hydrogeology. Using oscillatory flow interference testing can help characterize aquifer flow properties. Studies show that multi-frequency testing improves inversion performance and decreases parameter uncertainty.
Article
Mathematics, Interdisciplinary Applications
Peter K. Kitanidis
Summary: This paper discusses the application of covariance and Fisher information matrix in inverse problems, as well as a reexamination within the Bayesian framework, proposing a lower bound for the covariance of the posterior probability density function.
GEM-INTERNATIONAL JOURNAL ON GEOMATHEMATICS
(2021)
Article
Geosciences, Multidisciplinary
Catherine Christenson, David J. Hart, Michael Cardiff, Susan Richmond, Dante Fratta
Summary: This article presents a method for improving the communication of hydrologic data to the public by connecting data to video representations. The authors collected water-quality and geophysical data using multiple instruments mounted on a canoe and recorded video using GoPro cameras. The data was georeferenced and logged using an Arduino microcontroller. The results show that the low-cost sensors performed well and the data-rich video provided context for the measurements. This method enhances spatial understanding of hydrogeologic systems and facilitates communication and management of sensitive habitats.
Article
Engineering, Civil
Xueyuan Kang, Amalia Kokkinaki, Christopher Power, Peter K. Kitanidis, Xiaoqing Shi, Limin Duan, Tingxi Liu, Jichun Wu
Summary: The proposed geostatistical data assimilation method, using deep learning techniques, significantly improves the monitoring of DNAPL remediation. Experimental results show a reduction of 51% in the estimation error of DNAPL mass remediation compared to the standard EnKF method, indicating better real-time monitoring of DNAPL remediation.
JOURNAL OF HYDROLOGY
(2021)
Article
Geosciences, Multidisciplinary
Michael Cardiff, Laura Schachter, Jake Krause, Madeline Gotkowitz, Brian Austin
Summary: Increased nitrate concentrations in groundwater and surface waters due to modern agriculture is a widespread and significant environmental issue. However, there is a lack of understanding regarding the specific contributions of individual agricultural fields and practices. In this study, a minimally invasive approach using edge-of-field monitoring and tracer application was developed to calculate annual nitrogen loss to groundwater. Results from a commercial field in Wisconsin showed that nitrogen losses were similar to previous studies, with more than 25% of applied nitrogen leaching to groundwater each year. This method provides a reliable estimation of nitrogen loss when using certain conditions, such as injecting the tracer directly at the water table and analyzing nitrate concentrations in the laboratory.
Article
Environmental Sciences
Lijing Wang, Peter K. Kitanidis, Jef Caers
Summary: Bayesian inversion is commonly used to quantify uncertainty of hydrological variables. This paper proposes a hierarchical Bayesian framework to quantify uncertainty of both global and spatial variables. The authors present a machine learning-based inversion method and a local dimension reduction method to efficiently estimate posterior probabilities and update spatial fields. Using three case studies, they demonstrate the importance of quantifying uncertainty of global variables for predictions and the acceleration effect of the local PCA approach.
WATER RESOURCES RESEARCH
(2022)
Article
Water Resources
Mojtaba Forghani, Yizhou Qian, Jonghyun Lee, Matthew Farthing, Tyler Hesser, Peter K. Kitanidis, Eric F. Darve
Summary: This article presents a reduced-order model (ROM) based approach that utilizes a variational autoencoder (VAE) to compress bathymetry and flow velocity information, allowing for fast solving of bathymetry inverse problems. By constructing ROMs on a nonlinear manifold and employing a Hierarchical Bayesian setting, variational inference and efficient uncertainty quantification can be achieved using a small number of ROM runs.
ADVANCES IN WATER RESOURCES
(2022)
Article
Environmental Sciences
Xueyuan Kang, Amalia Kokkinaki, Xiaoqing Shi, Hongkyu Yoon, Jonghyun Lee, Peter K. Kitanidis, Jichun Wu
Summary: This study presents a framework that combines a deep-learning-based inversion method with a process-based upscaled model to estimate source zone architecture (SZA) metrics and mass discharge from sparse data. By improving the estimation method, the upscaled model accurately reproduces the concentrations and uncertainties of multistage effluents, providing valuable input for decision making in remediation applications.
WATER RESOURCES RESEARCH
(2022)
Article
Environmental Sciences
Kan Bun Cheng, Gedeon Dagan, Warren Barrash, Michael Cardiff, Avinoam Rabinovich
Summary: Characterizing aquifer heterogeneity is crucial for accurate flow and transport modeling. This study presents a new approach for statistically analyzing hydraulic properties in continuous pumping tomography tests of phreatic aquifers. The method involves determining equivalent hydraulic conductivity, specific storage, and specific yield at multiple locations and calculating statistical moments assuming random space variables. The results show that the spatial averages of the equivalent properties decrease with distance from the pumping well and stabilize at larger distances, consistent with existing theory.
WATER RESOURCES RESEARCH
(2022)
Article
Geosciences, Multidisciplinary
Jeremy R. R. Patterson, Michael Cardiff
Summary: Fractured sedimentary bedrock aquifers are complex flow systems with fast fractures and slow porous media-dominated flow paths. Previous studies have used oscillatory flow testing to characterize single bedrock fractures but relied on an idealized analytical model. This study extends the testing to fractured sedimentary bedrock and suggests that other hydraulic processes are needed to accurately represent pressure propagation.
Article
Environmental Sciences
Simon Meunier, Peter K. Kitanidis, Amaury Cordier, Alan M. MacDonald
Summary: This study develops a numerical model to simulate the abstraction capacities of photovoltaic water pumping systems across Africa using openly available data. The model includes realistic geological constraints on pumping depth and sub-hourly irradiance time series. The simulation results show that for much of Africa, groundwater pumping using photovoltaic energy is limited by aquifer conditions rather than irradiance. These findings can help identify regions with high potential for photovoltaic pumping and guide large-scale investments.
COMMUNICATIONS EARTH & ENVIRONMENT
(2023)