67639 research outputs found
Sort by
Spatiotemporal Assessment of Urban Thermal Discomfort in Kolkata, India: Insights from Cloud-Based Remote Sensing
No full textUrbanization and climate change are intricately linked, significantly influencing local and regional thermal environments. Kolkata, a rapidly expanding metropolitan city in India, has witnessed substantial shifts in urban thermal dynamics due to increasing land surface temperatures (LST), the urban heat island (UHI) effect, and heightened thermal discomfort. This study integrates high-resolution remote sensing data and a cloud-based platform via Google Earth Engine (GEE) to assess spatiotemporal changes in thermal discomfort in Kolkata. The study employs the urban thermal field variance index (UTFVI) to estimate the urban thermal discomfort, while the Sen's Slope estimator and the modified Mann-Kendall tests are applied to assess long-term spatiotemporal trends in urban thermal conditions. Findings reveal that LST, UHI, and UTFVI are significantly increasing at the rates of 0.149°C/year, 0.041°C/year, and 0.0041°C/year, respectively. The results further reveal that those areas with low vegetation cover experience extreme thermal stress, highlighting the critical role of urban greenery in mitigating heat-related discomfort. The study offers data-driven insights into Kolkata's urban thermal landscape, with guidance for policymakers in developing sustainable urban planning and climate adaptation strategies, such as expanding green spaces, implementing cool roof technologies, and enhancing urban ventilation. By leveraging cloud-based remote sensing, this study provides a scalable framework for assessing and addressing urban thermal discomfort in other rapidly urbanizing cities worldwide.The authors would like to express their sincere gratitude to Dr. Fulena Rajak, Head of the Department of Architecture and Planning at the for his invaluable support and assistance throughout this research. We also extend our thanks to the authorities of the NIT, Patna, and King Abdullah University of Science and Technology (KAUST), Saudi Arabia, for providing the necessary support and facilitating this work. The authors are thankful to Dr. Shailendra K. Mandal for his efforts in facilitating the collaboration between KAUST and the NIT, Patna. Furthermore, the authors gratefully acknowledge the United States Geological Survey (USGS) for making Landsat satellite data available, Google for granting access to the Google Earth Engine (GEE) platform, which was instrumental in the computational analysis, and the relevant Indian ministries for the provision of essential geospatial data and support. This research would not have been possible without the generous contribution of resources and data from these esteemed organizations
Detecting Faulty Steel Plates Using Machine Learning
No full textEfficiently detecting faults in steel plates is essential for maintaining the safety and dependability of structures and industrial machinery. Timely identification of faults mitigates further damage and averts exorbitant repair costs. This study delves into the efficacy of employing ensemble machine-learning classifiers for fault detection in steel plate manufacturing processes. Specifically, five powerful machine learning models—Random Forest (RF), AdaBoost, Decision Tree, Support Vector Machines (SVM) and Naive Bayes —are investigated in this study. The ensemble models (i.e., RF and AdaBoost) harness the collective power of multiple weak learners to enhance discrimination capacity. Evaluation is conducted using a publicly available dataset comprising seven distinct fault types: Pastry, Z_Scratch, K_Scratch, Stains, Dirtiness, Bumps, and Other_Faults. Results demonstrate Random Forest achieving the highest AUC of 0.942, with an accuracy of 0.771 and balanced F1 score, compared to the other models. This comprehensive investigation enhances fault detection efficacy, fostering informed decision-making in steel plate manufacturing processes
Slope assisted Physics Informed Neural Networks for seismic signal separation with applications on ground roll removal and interpolation
No full textThe knowledge of the local slope field of prestack seismic data is essential in several seismic signal processing tasks. Building on our previous slope-assisted, physics-informed seismic interpolation framework, dubbed PINNslope, we introduce a series of enhancements that elevate the framework's versatility. This ultimately enables its application to different signal separation problems, with a specific focus on ground roll removal. To begin with, the local slope estimated using our physics-informed neural networks framework is compared against the analytical local slope and those obtained from several conventional slope estimation algorithms. This comparison showcases that our prediction better approximates the analytical one. Second, we use the derived relation between the analytical slope and the physics-informed neural networks estimated slope to constrain the slope estimation network in the ground roll removal problem, predicting only the clean reflections and avoiding the prediction of the ground roll. To address the large difference in the frequency content of the field data, we utilize a time derivative term in the loss function to emphasize the amplitude of the comparatively higher frequency reflection arrivals. Furthermore, we modify the framework loss function and architecture to demonstrate the possibility of predicting two separate components of the seismic data according to the estimate of two local slopes that can span opposite or different ranges of values between each other. The effectiveness of the double slope framework is demonstrated on a proof of concept of the deblending problem and for the interpolation of complex aliased data characterized by conflicting dips, two tasks that were not achievable using our single slope prediction network implementation.We also thank Pham and Li (2022) to make the data
open source.
Open access publishing facilitated by King Abdullah University of Science and Technology, as part of the Wiley King
Abdullah University of Science and Technology (KAUST)
agreement
Unveiling the Impact of Corticosterone on Metabolic Crosstalk Between Astrocytes and Neurons
No full textChronic stress and its associated hormone, corticosterone (CORT), profoundly affect astrocytic energy metabolism and neuronal activity in the brain. However, it remains unclear how prolonged CORT exposure alters astrocyte metabolism and how these changes influence neuronal excitability via the astrocyte-neuron lactate shuttle (ANLS). This thesis investigates the metabolic reprogramming of astrocytes under sustained CORT exposure and the consequent impact on astrocyte-neuron interactions using transcriptomics, metabolomics, imaging, and electrophysiological approaches.
I found that sustained CORT exposure shifts astrocytes from a protoplasmic to a reactive phenotype. Transcriptomic and metabolomic analyses revealed enhanced glucose uptake, reduced glycolytic flux, and a redirection of glucose flux into the pentose phosphate pathway (PPP), thereby enhancing antioxidant capacity at the expense of ATP production. This reprogramming was supported by the upregulation of glucose-6-phosphate dehydrogenase (G6pd), the key regulatory enzyme of the PPP, increased NADPH levels, decreased redox glutathione ratio (GSH/GSSG), reduced glycogen stores, elevated extracellular lactate release, and impaired mitochondrial pyruvate oxidation. Changes in glucose transporter expression further indicated an adaptive response to prolonged metabolic stress.
I next examined how these astrocytic metabolic changes affect neuronal physiology. CORT-treated astrocytes showed persistent elevations and altered dynamics of intracellular calcium, consistent with disrupted calcium-dependent signaling and gliotransmission. Using both mixed astrocyte-neuron co-cultures and transwell systems, I assessed neuronal activity with multielectrode array (MEA) and patch-clamp recordings. In mixed cultures, CORT treatment consistently reduced neuronal network excitability, an effect reversed with 10 mM lactate. In contrast, in neuron-only cultures or in neurons co-cultured with CORT-treated astrocytes (without direct contact), network-level excitability remained unchanged; however, patch-clamp recordings revealed increased intrinsic excitability of individual neurons. When both astrocytes and neurons were exposed to CORT together, neuronal excitability was suppressed, mirroring the findings in mixed cultures.
Collectively, this work demonstrates that prolonged CORT exposure induces metabolic reprogramming in astrocytes that favor antioxidant defense over energy supply. These adaptations alter astrocyte-neuron metabolic coupling and neuronal excitability, highlighting the pivotal role of astrocytic metabolism in stress responses. My results suggest novel avenues for understanding, and potentially targeting, astrocyte-driven mechanisms in stress-related brain disorders
The fractional Laplacian uncovered
No full textNonlocal operators extend the concept of classical differential operators. These lecture notes offer a comprehensive introduction to the fractional Laplacian – a primary example of a nonlocal operator. Over the past few decades, the study of nonlocal equations has gained significant attention due to their wide range of applications in various fields, including financial mathematics, biotechnology, data science, machine learning, optimal design problems, and chemical engineering, to mention a few.
These notes are intended for students and young researchers already familiar with the classical Laplace operator and eager to uncover its nonlocal counterpart. We aimed to offer a clear, well-structured, and self-contained presentation of the material with attention to core ideas and details.
The first three chapters cover a series of properties that the fractional Laplacian shares with the classical Laplacian, such as the maximum principle and the Harnack inequality. Evidently, the nonlocal nature of the operator reshapes familiar properties in subtle ways – changes that begin to feel natural as the reader steps behind the curtain and sees the machinery driving the theory. The striking differences between these two operators are also highlighted. The last three chapters develop tools for establishing regularity results
for nonlocal equations. More precisely, they cover the nonlocal KrylovSafonov regularity theory, the approximation method of Caffarelli & Silvestre, as well as Serra’s approach to improving regularity. In an effort to introduce the core ideas clearly, these tools are presented for the model operator – the fractional Laplacian.
The appendix gathers the notations used throughout the notes and includes several well-known results, omitting proofs that are purely technical.
This work has grown out of our earlier survey and research papers, with the writing also influenced by the books [7, 40], and we do not claim
any originality. We hope it serves as a useful resource for those beginning their study of nonlocal equations and inspires further exploration into this fascinating area of mathematical analysis.We are deeply grateful to our mentors and colleagues who have enriched our understanding of the topic through many fruitful discussions over the years. We would like to mention, in particular, Luis Caffarelli, Eduardo Teixeira, José Miguel Urbano, Fernando Charro, Damião Araújo, Patrício Felmer, Erwin Topp, Gabrielle Nornberg, Alexander Quaas, Uberlândio Severo, José Anderson Cardoso, and Diego Marcon
Towards Usable and Useful Explainable Artificial Intelligence
No full textExplainable AI (XAI) plays a pivotal role in enabling understanding and trust in AI systems, especially as their complexity continues to grow. In the era of classical deep learning, research efforts have focused on providing human-understandable explanations for model outputs and behaviors, often delivered through explanation-as-a-service, i.e., usable XAI. More recently, with the advent of large language models, researchers have paid additional attention to understanding the internal representations and mechanisms of these models, leveraging this knowledge to address broader challenges such as safety issues, forming the foundation of useful XAI.
Despite many interpretation techniques having been developed to enhance the explainability of deep learning models. A central obstacle to deploying XAI in practice is faithfulness: do we truly trust these explainers? In other words, are these interpretable methods faithful to the underlying models? This thesis addresses faithfulness by improving three key properties of interpretation methods: stability, controllability, and consistency. Our contributions span theoretical definitions, novel algorithms, and empirical validation, and outline open problems for future research
SkinningGS: Editable Dynamic Human Scene Reconstruction Using Gaussian Splatting Based on a Skinning Model
No full textReconstructing an interactive human avatar and the background from a monocular video of a dynamic human scene is highly challenging. In this work we adopt a strategy of point cloud decoupling and joint optimization to achieve the decoupled reconstruction of backgrounds and human bodies while preserving the interactivity of human motion. We introduce a position texture to subdivide the Skinned Multi-Person Linear (SMPL) body model's surface and grow the human point cloud. To capture fine details of human dynamics and deformations, we incorporate a convolutional neural network structure to predict human body point cloud features based on texture. This strategy makes our approach free of hyperparameter tuning for densification and efficiently represents human points with half the point cloud of HUGS. This approach ensures high-quality human reconstruction and reduces GPU resource consumption during training. As a result, our method surpasses the previous state-of-the-art HUGS in reconstruction metrics while maintaining the ability to generalize to novel poses and views. Furthermore, our technique achieves real-time rendering at over 100 FPS, 6 the HUGS speed using only Linear Blend Skinning (LBS) weights for human transformation. Additionally, this work demonstrates that this framework can be extended to animal scene reconstruction when an accurately-posed model of an animal is available.We extend our gratitude to the developers of the outstanding open-source projects COLMAP, ROMP, Gaussian Splatting, Gaussian Avatar, Neuman, and HUGS, whose contributions have significantly supported the development of our work. We also sincerely thank everyone who provided valuable feedback and assistance throughout this project
Design and Testing of a Laminar Flow Windows High-Temperature Optical Cell
No full textThis work presents the development of a high-temperature, atmospheric-pressure, optical cell equipped with laminar flow windows, designed for measuring high-temperature cross-sections in the long-wave infrared range. Previous high-temperature optical cell designs often lack access to the long-wave infrared range due to thermal limitations of optical materials and viewport designs. The developed cell incorporates two laminar flow windows that confine the test gas within a uniform-temperature central zone of the cell. This configuration ensures that the zinc selenide viewports – which have extended transmissivity – remain within their rated temperature limits, while maintaining a uniform temperature along the path of the test sample.
The concept of laminar flow windows has previously been applied to high-temperature spectroscopy. However, a detailed evaluation of their performance and limitations – critical for assessing the accuracy of measurements based on this approach – remains absent from the literature. In this study, we conduct rigorous experimental tests to quantify leak rates between adjacent regions of the optical cell, inferred from flow composition measurements at the exhaust of each region. Additionally, computational fluid dynamics simulations are performed to investigate fluid flow, heat transfer, and species transport within the cell. These investigations provide insights into the cell’s performance, focusing on how flow rates affect pressure gradients across the laminar flow windows and how molecular diffusion influences mole fraction profiles.
Results indicate that the error in the effective optical depth can be limited to within 3%, provided that flow rates are sufficiently high to mitigate the effects of molecular diffusion. Experimental observations underscore the influence of manufacturing tolerances and thermal cycling on pressure distribution within the optical cell. As a demonstration, the cell was used to measure methane (CH₄) absorption over 1331.2 - 1333.5 cm⁻¹ at temperatures up to 950 K, and nitrous oxide (N₂O) absorption over 1313.2 - 1317 cm⁻¹ at temperatures up to 570 K. Comparisons between the measured spectra and HITEMP simulations revealed residuals within the combined uncertainties of the measurements and the simulated spectra, validating the accuracy of the laminar flow windows cell for high-temperature spectral analysis.This work was funded by King Abdullah University of Science and Technology (KAUST)
Physics-based, data-driven production forecasting in the Utica and Point Pleasant Formation
No full textWe introduce a robust, data-driven and physics-based method to forecast the estimated ultimate recovery (EUR) of gas and condensates from the Utica-Point Pleasant Formation. By categorizing 3,410 horizontal wells into eight static cohorts based on the initial gas-oil ratios (GOR) and completion dates, we construct generalized extreme value (GEV) distributions of annual reservoir fluid mass production for each cohort. We use the expected values (means) to build historical, statistical well prototypes for each cohort. These prototypes account for hydraulic fracture deterioration, pressure interference between neighboring fractures, well interference, and advancements in completion technology. We extrapolate mass production of the reservoir fluids for several decades by fitting the scaling parameters of our physics-based model to the statistical well prototypes. A key innovation in this approach is using GEV statistics to generate a time series for each cohort’s GOR and condensate-gas ratio (CGR). We replace the individual production rates from all existing wells with their corresponding extended well prototypes and convert mass to volumetric rates. We sum up the latter rates and provide a base forecast of total reservoir fluid rates and cumulative production, closely matching the historical field production. Our base forecasts predict approximately 23 trillion scf of gas and 200 million barrels of condensate by 2035.
The Utica-Point Pleasant Formation is divided into core and noncore areas, each further subdivided into regions based on fluid types. We analyze the future infill potential per square mile in each of these regions, proposing two drilling schedules and four potential future drilling scenarios. We identify approximately 35,847 potential future wells: 12,476 in the core area and 23,371 in the noncore area. The core wells can contribute about 157 trillion scf of natural gas and 0.9 billion barrels of condensate.
This study marks the first integrated evaluation of the future production from the Utica-Point Pleasant Formation by combining GEV statistics, physics-based modeling, big-data analysis, geology, proposed drilling programs, and economics. It provides a comprehensive understanding of the Utica-Point Pleasant Formation’s significance in future U.S. shale production.The authors thank King Abdullah University of Science and Technology (KAUST) for supporting this research through baseline research funding allocated to Professor Tadeusz Patzek. We thank Dr. Wardana Sapura of UT Austin for numerous discussions and valuable help in developing this paper
Lactate promotes longevity through redox-driven lipid remodeling in <i>Caenorhabditis elegans</i>.
No full textLactate has emerged as a key metabolite involved in multiple physiological processes, including memory formation, immune response regulation, and muscle biogenesis. However, its role in aging and cellular protection remains unclear. Here, we show that lactate promotes longevity in C. elegans through a mechanism that requires early-life intervention, indicating a hormetic priming effect. This pro-longevity action depends on its metabolic conversion via LDH-1 and NADH, which drives redox-dependent metabolic reprogramming. Multi-omics approaches revealed that lactate induces early-stage metabolic adaptations, with a strong modulation of lipid metabolism, followed by late-life transcriptional remodeling. These shifts are characterized by enhanced stress response pathways and suppression of energy-associated metabolic processes. Our genetic screening identified sir-2.1/SIRT1 and rict-1/RICTOR as essential for lactate-mediated lifespan extension. Our findings establish lactate as a pro-longevity metabolite that couples redox signaling with lipid remodeling and nutrient-sensing pathways. This work advances our understanding of lactate's dual role as a metabolic intermediary and geroprotector signaling molecule, offering insights into therapeutic strategies for age-related metabolic disorders.We thank Dr. Sylvia Lee and Dr. Bennett Fox for their technical assistance and constructive comments. Thanks to the C. elegans Genetic Center and National Bioresource Project for the different strains used in this study. Support for this study was provided by King Abdullah University of Science and Technology (PJM), Canadian Institutes of Health Research (AGE-477545) (AT), NIH R00GM140217(JDM) and NIH R35GM131877 (FCS)