Hasil untuk "Ocean engineering"

Xiaobing Zhou, Frank Colberg, Debra Hudson et al.

Operational ocean forecasting systems conventionally employ dynamical ocean models driven by atmospheric forcing derived from numerical weather prediction (NWP) models. Recent advancements in artificial intelligence and machine learning (ML) have led to the development of ML-based atmospheric weather models, which have competitive, if not better, medium range forecast accuracy compared to traditional NWP systems. This study evaluates the impact of ML-based atmospheric forcing on ocean forecast skill through two sets of 10-day forecasts using the UK Met Office GOSI9 configuration of the NEMO dynamical ocean model. Both experiments share identical ocean initial conditions; but differ in atmospheric forcing: one uses ECMWF's ML-based AIFS model, while the other uses the Australian Bureau of Meteorology's physics-based NWP model, ACCESS-G3. Forecasts were initialized on the first day of each month over the period 2023-2024. The quality of the atmospheric forcing was assessed by comparing AIFS and ACCESS-G3 forecast skill against both ECMWF reanalysis v5 (ERA5) and ACCESS-G3 analyses. Results indicate that AIFS consistently outperforms ACCESS-G3, either from the initial forecast time or after the first few days. Oceanic forecast skill was evaluated against both the GOSI9 reanalysis and observations, focusing on key surface variables including sea surface temperature, salinity, sea level, and ocean currents. The ocean forecasts forced with AIFS atmospheric data exhibit comparable or enhanced predictive skill compared to those forced with ACCESS-G3 data. These findings underscore the potential of ML-based atmospheric models to replace traditional NWP forcing in operational ocean forecasting systems, offering improved accuracy and computational efficiency.

Yanjie Zhang, Yuqi Ji, Yingpeng Zhang

High-entropy alloys (HEAs) have rapidly evolved from a seminal concept in 2004 to a mainstream materials science frontier, witnessing exponential growth since 2010. To date, the preparation and research methods for HEAs have undergone substantial diversification, the systems have been optimized, and their application scope has widely broadened. Herein, we provide a systematic review of various synthesis methodologies, including mechanical alloying, vacuum smelting, magnetron sputtering, and additive manufacturing. This paper meticulously summarizes a series of findings on the crucial properties of HEAs, such as mechanical properties, wear resistance, and corrosion resistance, as well as functional properties, including irradiation resistance, hydrogen storage capacity, and biocompatibility. In addition, this review explores the promising applications of HEAs in fields such as aerospace and ocean engineering. Modeling techniques applicable to HEAs, namely ab initio molecular dynamics simulations and CALPHAD modeling, are introduced and discussed. Finally, despite significant successes, the current shortcomings of HEAs, as well as future opportunities and challenges, are outlined. In summary, this review aims to offer both theoretical references and practical guidelines for the rapid evolution of HEAs.

Yingjing Jiang, Shaoqing Zhang, Yang Gao et al.

The global ocean meridional overturning circulation (GMOC) is central for ocean transport and climate variations. However, a comprehensive picture of its historical mean state and variability remains vague due to limitations in modelling and observing systems. Incorporating observations into models offers a viable approach to reconstructing climate history, yet achieving coherent estimates of GMOC has proven challenging due to difficulties in harmonizing ocean stratification. Here, we demonstrate that applying multiscale data assimilation scheme that integrates atmospheric and oceanic observations into multiple coupled models in a dynamically consistent way, the global ocean currents and GMOC over the past 80 years are retrieved. While the major historic events are printed in variability of the rebuilt GMOC, the timeseries of multisphere 3-dimensional physical variables representing the realistic historical evolution enable us to advance understanding of mechanisms of climate signal propagation cross spheres and give birth to Artificial Intelligence coupled big models, thus advancing the Earth science.

Gennady A. Chepurin, James A. Carton, Luyu Sun et al.

This paper describes the new Simple Ocean Data Assimilation version 4 (SODA4) global eddy-resolving ocean/sea ice reanalysis that spans the 45-year period 1980-2024. The reanalysis is constructed using GFDL MOM5/SIS1 numerics and ECMWF ERA5 forcings with surface and subsurface temperature and salinity observations as constraints within an optimal interpolation data assimilation algorithm. The method of construction and resulting output files are described. Comparison of the SODA4 temperature and salinity fields to observations and to the UK Met Office EN4 temperature and salinity analyses in the upper ocean shows SODA4 has marginal bias and exhibits more regional variability, with less of an imprint of the sparse and inhomogeneous distribution of observations. Comparison of transports across major ocean sections and passages are generally consistent with independent moored observations.

Andre N. Souza, Simone Silvestri, Katherine Deck et al.

Understanding subsurface ocean dynamics is essential for quantifying oceanic heat and mass transport, but direct observations at depth remain sparse due to logistical and technological constraints. In contrast, satellite missions provide rich surface datasets-such as sea surface height, temperature, and salinity-that offer indirect but potentially powerful constraints on the ocean interior. Here, we present a probabilistic framework based on score-based diffusion models to reconstruct three-dimensional subsurface velocity and buoyancy fields, including the energetic ocean eddy field, from surface observations. Using a 15-level primitive equation simulation of an idealized double-gyre system, we evaluate the skill of the model in inferring the mean circulation and the mesoscale variability at depth under varying levels of surface information. We find that the generative model successfully recovers key dynamical structures and provides physically meaningful uncertainty estimates, with predictive skill diminishing systematically as the surface resolution decreases or the inference depth increases. These results demonstrate the potential of generative approaches for ocean state estimation and uncertainty quantification, particularly in regimes where traditional deterministic methods are underconstrained or ill-posed.

Florentin Daniel, Ludovic Petitdemange, Christophe Gissinger

Jupiter's icy moons are believed to host subsurface liquid oceans, and among them, Europa stands out as one of the most promising candidates for extraterrestrial life. Yet, the processes driving oceanic flows beneath its ice shell, as well as the factors controlling the thickness of this ice, remain incompletely understood. One especially distinctive feature of Europa is that its salty ocean is electrically conducting and thus influenced by Jupiter's time-varying magnetic field, which is believed to drive a large-scale zonal flow. Here, we examine hos this magnetically-induced jet affects both the heat flux and the dynamics of the convective flow within Europa's ocean. We first show that the magnetically-driven jet efficiently transports heat in stably stratified regions near the top of the ocean, and may alter the expected convective scaling laws in deeper layers. Second, by analysing the latitudinal distribution of heat flux and relating it to ice-thickness variations, we make predictions that can be compared with current observations. In anticipation of the upcoming JUICE and Europa Clipper missions, we discuss how improved measurement precision could help further constrain the ocean's properties and refine our model-based forecasts.

Muhammad Tayyab Khan, Zane Yong, Lequn Chen et al.

Engineering drawings are fundamental to manufacturing communication, serving as the primary medium for conveying design intent, tolerances, and production details. However, interpreting complex multi-view drawings with dense annotations remains challenging using manual methods, generic optical character recognition (OCR) systems, or traditional deep learning approaches, due to varied layouts, orientations, and mixed symbolic-textual content. To address these challenges, this paper proposes a three-stage hybrid framework for the automated interpretation of 2D multi-view engineering drawings using modern detection and vision language models (VLMs). In the first stage, YOLOv11-det performs layout segmentation to localize key regions such as views, title blocks, and notes. The second stage uses YOLOv11-obb for orientation-aware, fine-grained detection of annotations, including measures, GD&T symbols, and surface roughness indicators. The third stage employs two Donut-based, OCR-free VLMs for semantic content parsing: the Alphabetical VLM extracts textual and categorical information from title blocks and notes, while the Numerical VLM interprets quantitative data such as measures, GD&T frames, and surface roughness. Two specialized datasets were developed to ensure robustness and generalization: 1,000 drawings for layout detection and 1,406 for annotation-level training. The Alphabetical VLM achieved an overall F1 score of 0.672, while the Numerical VLM reached 0.963, demonstrating strong performance in textual and quantitative interpretation, respectively. The unified JSON output enables seamless integration with CAD and manufacturing databases, providing a scalable solution for intelligent engineering drawing analysis.

Beibei Zhou, Yingshuang Li, Feng Xu et al.

As a key vehicle for ecological conservation, National Parks (NPs) require accurate and timely land cover (LC) information for refined management. However, complex terrain and frequent human activities pose challenges to efficient LC extraction. This study focuses on the Chengdu section of the Giant Panda National Park and proposes a framework of multisource data integration, dynamic feature selection, algorithm performance evaluation, temporal sample migration, and LC change analysis. Sentinel-1, Sentinel-2 A, and SRTM data are integrated to construct 67 multidimensional features. Recursive feature elimination combined with Bayesian optimization is used for feature selection, and the classification performance of random forest (RF), support vector machine, and classification and regression tree are compared. Spectral angle mapper and spectral Euclidean distance are introduced for temporal sample migration. Results show that the RF classifier with optimized features yields the best performance, achieving an overall accuracy of 0.9330 and a Kappa coefficient of 0.9196, significantly outperforming GLC_FCS30D, Esri Land Cover, and China Land Cover Dataset. Accuracy after sample migration remains above 0.8800 annually. The framework effectively identified bamboo forests critical to panda habitats. For example, using only water indexes, bamboo forest producer accuracy was 0.0555, but increased to 0.8571 with added spectral and vegetation features. From 2018 to 2023, woodland increased by 64.77 km<inline-formula><tex-math notation="LaTeX">$^{2}$</tex-math></inline-formula> and bamboo forest by 22.26 km<inline-formula><tex-math notation="LaTeX">$^{2}$</tex-math></inline-formula>, while barren land and construction land increased by 116.04km and 174.28 km<inline-formula><tex-math notation="LaTeX">$^{2}$</tex-math></inline-formula>. The proposed framework effectively enhances LC monitoring in mountainous environments and provides technical support for conservation planning and ecological supervision in NPs.

Huiyu Ding, Jun Liu, Zhihui Wang et al.

The core of hyperspectral image (HSI) classification lies in the effective fusion of spatial-spectral features. However, traditional methods are limited by the capacity of handcrafted feature representation, while deep learning methods face challenges such as overfitting with small sample sizes and high computational complexity. This article proposes a Mamba-driven multiscale spatial-spectral fusion network (M²S²F-Net). This network extracts spatial-spectral features at different granularities through the spatial-spectral multigranularity feature extraction module, adaptively enhances the spatial-spectral correlation through the spatial-spectral fusion attention module, optimizes feature fusion by combining local and global streams with the feature fusion enhanced vision transformer, and establishes long-sequence dependencies using the dual-path feature fusion mamba. The M²S²F-Net employs a multistage feature fusion strategy of &#x201C;coarse fusion-fine optimization-strong screening&#x201D; to achieve efficient classification with few samples. The network was validated on three publicly available HSI datasets to demonstrate its superiority in few-shot scenarios, with significant improvements in classification accuracy. It also exhibited remarkable classification performance across different numbers of training samples.

Zih-Wei Tang, Liang-Chi Wang, Huei-Fen Chen et al.

Abstract To address the environmental changes in the South China coastal region and to investigate the interplay among sea-level fluctuations, monsoon variability, and sediment dynamics, a sediment core from the Pinqing Lagoon was extracted, covering the last 8.5 ka. Furthermore, multiple proxies were analyzed in the core, including grain size end-members (EM1, EM2, and EM3), magnetic susceptibility and S-ratio, the carbon (C) isotopic composition of organic matter, its carbon and nitrogen (N) contents, the resulting C/N ratio, and Itrax XRF-derived elemental ratios such as Mn/Ti, Si/Ti, K/Ti, and Fe/Ti. The results reveal that changes in sea level play a primary role in shaping the lagoon sedimentary and geochemical evolution, with EASM-driven runoff acting as a secondary control on terrestrial sediment supply, especially during low sea-level phases. During the 8.5–6.8 ka, low water levels, strong EASM-driven runoff, and dominant terrestrial C₄ plant input resulted in coarse detrital sedimentation (high EM2 and low S-ratio) and poor bottom water oxygenation (low Mn/Ti). Between 6.8 and 5.8 ka, despite already high sea levels, the lagoon underwent rapid deepening, with a shift toward in-situ aquatic productivity, improved oxygenation, and finer sedimentation (EM1 dominance and high S-ratio), likely reflecting local geomorphological changes. From 5.8 to 4.2 ka, as sea level stabilized, the lagoon became stratified and marine-influenced, with low oxygenation, minimal terrestrial input, and background fine-grained sedimentation. After 4.2 ka, stable high water levels and low runoff persisted. A prominent EM3 peak between 0.4 and 0.2 ka, coinciding with the Late Little Ice Age (LIA), reflects frequent typhoon-induced high-energy deposition, supported by coarse grain size, elevated MS, and increased Si/Ti, K/Ti, and Fe/Ti ratios. Overall, the results highlight that long-term sea-level fluctuations primarily controlled lagoonal sedimentation and oxygenation, while EASM variability shaped runoff-driven detrital input.

Libin Zhu, Jie Wang, Yiwei Xu et al.

This study investigates a typical ozone pollution episode in Hong Kong from May 29 to 31, 2023. Based on the observations of a Differential Absorption Lidar (DIAL) system, both ozone and aerosols accumulated below 1.5 km during the pollution episode. Ozone exhibited distinct formation and accumulation characteristics, with concentrations exceeding 200 μg m⁻³. Aerosols presented evident features of regional transport and local coupling, with extinction coefficients surpassing 1.1 km⁻¹. During late spring to early summer, the northward extension of the Western Pacific Subtropical High (WPSH) established favorable conditions for ozone production. This background was amplified by Typhoon Mawar, whose peripheral circulation channeled pollutants from the Pearl River Delta into Hong Kong through horizontal and vertical pathways, significantly worsening near-surface air quality. The episode was eventually mitigated, as enhanced vertical mixing facilitated the dispersion and removal of accumulated pollutants. These results highlight the critical role of meteorological–chemical interactions in shaping this ozone pollution episode.

Jiyong Kim, Seoyeon Cho, Minkyung Chung et al.

Stereo matching algorithms are considered one of the most important subtasks in 3-D reconstruction, as 3-D coordinates are derived from the disparity values of pixels obtained through stereo matching. Recently, deep learning-based satellite stereo matching algorithms have been widely investigated, as they can capture both deep and shallow features of complex satellite scenes. However, several problems in satellite stereo matching, due to the unique properties of satellite images, remain unsolved, particularly in textureless and repetitive regions. In these regions, a single object in a satellite image is likely to be matched with similar objects, causing multiple disparity probabilities and shifts in the disparity estimation. To address the problem of disparity shifts, we propose a novel cost volume refinement strategy (CVRS). CVRS introduces both left-right and left-left cost volumes, which work together to refine disparities and eliminate false matches in textureless or repetitive regions, while preserving the original disparity values. With CVRS, we propose a new model for satellite stereo matching, the self-refined cost volume network (SRCV-Net). We evaluated CVRS and SRCV-Net on the US3D and WHU-Stereo datasets, comparing it using the EPE and D1 metrics. The application of CVRS demonstrated performance improvements in all models, and SRCV-Net achieved superior accuracy in satellite stereo matching. Furthermore, CVRS can be easily applied to various models with minimal structural changes and a small increase in parameters. SRCV-Net, with its innovative CVRS, provides an effective solution to the challenges of satellite stereo matching, offering enhanced accuracy, efficiency, and adaptability.

Khaled Mohsen Helal, Jacopo Fragasso, Lorenzo Moro

Everton Cavalcante, Thais Batista, Flavio Oquendo

Modern systems are increasingly connected and more integrated with other existing systems, giving rise to \textit{systems-of-systems} (SoS). An SoS consists of a set of independent, heterogeneous systems that interact to provide new functionalities and accomplish global missions through emergent behavior manifested at runtime. The distinctive characteristics of SoS, when contrasted to traditional systems, pose significant research challenges within Software Engineering. These challenges motivate the need for a paradigm shift and the exploration of novel approaches for designing, developing, deploying, and evolving these systems. The \textit{International Workshop on Software Engineering for Systems-of-Systems} (SESoS) series started in 2013 to fill a gap in scientific forums addressing SoS from the Software Engineering perspective, becoming the first venue for this purpose. This article presents a study aimed at outlining the evolution and future trajectory of Software Engineering for SoS based on the examination of 57 papers spanning the 11 editions of the SESoS workshop (2013-2023). The study combined scoping review and scientometric analysis methods to categorize and analyze the research contributions concerning temporal and geographic distribution, topics of interest, research methodologies employed, application domains, and research impact. Based on such a comprehensive overview, this article discusses current and future directions in Software Engineering for SoS.

Siddhant Dutta, Iago Leal de Freitas, Pedro Maciel Xavier et al.

Federated Learning (FL) is a decentralized machine learning approach that has gained attention for its potential to enable collaborative model training across clients while protecting data privacy, making it an attractive solution for the chemical industry. This work aims to provide the chemical engineering community with an accessible introduction to the discipline. Supported by a hands-on tutorial and a comprehensive collection of examples, it explores the application of FL in tasks such as manufacturing optimization, multimodal data integration, and drug discovery while addressing the unique challenges of protecting proprietary information and managing distributed datasets. The tutorial was built using key frameworks such as $\texttt{Flower}$ and $\texttt{TensorFlow Federated}$ and was designed to provide chemical engineers with the right tools to adopt FL in their specific needs. We compare the performance of FL against centralized learning across three different datasets relevant to chemical engineering applications, demonstrating that FL will often maintain or improve classification performance, particularly for complex and heterogeneous data. We conclude with an outlook on the open challenges in federated learning to be tackled and current approaches designed to remediate and improve this framework.

Pengfei Gao, Yuanyuan Song, Jian Wang et al.

The chloride ion permeability resistance of recycled aggregate concrete (RAC) is influenced by multiple factors, and the prediction model for this resistance based on machine learning is still limited. In the paper, six impact factors (IFs), including the carbonation of recycled coarse aggregates (<i>YN</i>), the replacement ratio of recycled coarse aggregates (<i>r</i>), the bending load level (<i>L</i>), the carbonation time (<i>t</i>) and temperature (<i>T</i>) of RAC, and the replacement ratio of carbonated recycled fine aggregates (<i>f</i>), were considered to conduct a chloride penetration test on RAC. Based on the experimental data, four algorithms, including artificial neural network (ANN), support vector machine (SVM), random forest (RF) and extreme gradient boosting (XGBoost), were adopted to establish the machine learning prediction models and study the relationships between the electric flux of RAC and the IFs. The results showed that the predicted values of all four models were in good agreement with the experimental values, and the XGBoost model had the best prediction performance on the testing set. Based on the XGBoost model, the LIME method was adopted to solve the interpretability problem in the prediction process. The importance ranking of IFs on the electric flux was <i>r</i> > <i>t</i> > <i>f</i> > <i>T</i> > <i>L</i> > <i>YN</i>. A graphical user interface (GUI) was developed based on Python 3.8 software to facilitate the use of machine learning models for the chloride ion permeability resistance of RAC. The research results can provide an accurate prediction of the electric flux of RAC.

H. Zuo, H. Zuo, Y. Liu et al.

<p>Silicate weathering, which is of great importance in regulating the global carbon cycle, has been found to be affected by complicated factors, including climate, tectonics and vegetation. However, the exact transfer function between these factors and the silicate weathering rate is still unclear, leading to large model–data discrepancies in the CO<span class="inline-formula">₂</span> consumption associated with silicate weathering. Here we propose a simple parameterization for the influence of vegetation cover on erosion rate to improve the model–data comparison based on a state-of-the-art silicate weathering model. We found out that the current weathering model tends to overestimate the silicate weathering fluxes in the tropical region, which can hardly be explained by either the uncertainties in climate and geomorphological conditions or the optimization of model parameters. We show that such an overestimation of the tropical weathering rate can be rectified significantly by parameterizing the shielding effect of vegetation cover on soil erosion using the leaf area index (LAI), the high values of which are coincident with the distribution of leached soils. We propose that the heavy vegetation in the tropical region likely slows down the erosion rate, much more so than thought before, by reducing extreme streamflow in response to precipitation. The silicate weathering model thus revised gives a smaller global weathering flux which is arguably more consistent with the observed value and the recently reconstructed global outgassing, both of which are subject to uncertainties. The model is also easily applicable to the deep-time Earth to investigate the influence of land plants on the global biogeochemical cycle.</p>

Yoann Arhant, Olga Lopera Tellez, Xavier Neyt et al.

Seabed characterization consists in the study of the physical and biological properties of the of ocean floors. Sonar is commonly employed to capture the acoustic backscatter reflected from the seabed. It has been extensively used for automatic target recognition (ATR) within mine countermeasures (MCM) operations in shallow waters. However, conventional machine learning (ML) and deep learning approaches face challenges in automatically mapping the seabed due to noise and limited labels. Thus, this article introduces the Deep Supervised Semantic Segmentation model for Seabed Characterization (D4SC), tailored for addressing challenges associated with sonar data. D4SC employs convolutional neural networks, specific high-resolution (HR) synthetic aperture sonar (SAS) data preprocessing and data augmentation methods, including the novel boundary pixel label rejection, and moves from the low-label regime. Performance comparisons against standard methods in the literature are conducted, demonstrating D4SC&#x0027;s superiority on challenging HR SAS survey datasets from real-world MCM exercises at sea. In addition, this work thoroughly explores the effect of the quality of the datasets, the robustness of training models on Out-of-Distribution data, and the estimation of epistemic uncertainty to refine predictions at large scale.

Hirotaka Ijima, Marko Jusup

Tuna and billfish are charismatic pelagic fishes attracting considerable scientific attention due to their ecophysiological and socioeconomic importance. However, the knowledge of their basin-wide spawning and larval habitats, especially in a warming ocean, is limited. This knowledge gap undermines effective fishery management by introducing recruitment uncertainty, which makes population dynamics unpredictable. We fill the gap with a parsimonious geostatistical species-distribution model trained on the largest available dataset on tuna and billfish larvae in the Pacific Ocean. The model reveals (i) the basin-wide seasonal larval distributions over the reference period 1960-85, (ii) the expected impact of ongoing ocean warming on these distributions, and (iii) the biogeochemical factors, such as pH, phosphate concentration, and sea-surface height, that shape the larval habitat. Our findings make a quantum leap in understanding the ecophysiology of tuna and billfish, providing valuable information for future conservation efforts.

Jun Korenaga

The early evolution of the Earth-Moon system prescribes the tidal environment of the Hadean Earth and holds the key to the formation mechanism of the Moon and its thermal evolution. Estimating its early state by backtracking from the present, however, suffers from substantial uncertainties associated with ocean tides. Tidal evolution during the solidification of Earth's magma ocean, on the other hand, has the potential to provide robust constraints on the Earth-Moon system before the appearance of a water ocean. Here we show that energy dissipation in a solidifying magma ocean results in considerably more limited lunar recession than previously thought, and that the Moon was probably still at the distance of $\sim$7-9 Earth radii at the end of solidification. This limited early recession aggravates the often overlooked difficulty of modeling tidal dissipation in Earth's first billion years, but it also offers a new possibility of resolving the lunar inclination problem by allowing the operation of multiple excitation mechanisms.