Ocean Modelling最新文献_第2页

Integrating machine learning into a fully coupled current-wave-sediment model: Characterizing particle size in the settling process in estuaries of the great barrier reef, Australia 将机器学习集成到一个完全耦合的电流-波浪-沉积物模型中：表征澳大利亚大堡礁河口沉降过程中的粒度

IF 2.9 3区地球科学

Ocean Modelling Pub Date : 2025-08-13 DOI: 10.1016/j.ocemod.2025.102621

Ziyu Xiao , Daniel N. Livsey , Thomas Schroeder , David Blondeau-Patissier , Rodrigo Santa Cruz , Jiasheng Su , Dehai Song , Xiao Hua Wang , Geoffrey Carlin , Andrew D.L. Steven , Joseph R. Crosswell

{"title":"Integrating machine learning into a fully coupled current-wave-sediment model: Characterizing particle size in the settling process in estuaries of the great barrier reef, Australia","authors":"Ziyu Xiao , Daniel N. Livsey , Thomas Schroeder , David Blondeau-Patissier , Rodrigo Santa Cruz , Jiasheng Su , Dehai Song , Xiao Hua Wang , Geoffrey Carlin , Andrew D.L. Steven , Joseph R. Crosswell","doi":"10.1016/j.ocemod.2025.102621","DOIUrl":"10.1016/j.ocemod.2025.102621","url":null,"abstract":"<div><div>Accurate prediction of sediment settling is critical for management of coastal ecosystems, but complex estuarine processes that influence sediment deposition and erosion present a major modelling challenge. This study introduces a proof-of-concept framework that integrates machine learning (ML) into environmental simulations to improve accuracy and efficiency by modelling dynamic sediment flocculation processes and their influence on particle size, enabling a more precise determination of settling velocity. Environmental factors influencing in-situ sediment particle size were used to train a regression model based on coeval measurements of three key parameters: salinity, shear rate and suspended sediment concentration (SSC). This regression model was developed using ML and integrated into a fully coupled current-wave-sediment model to simulate the flocculation response to these three parameters. The integrated model framework demonstrates its reliability and accuracy when evaluated against the in-situ measurements, satellite-derived SSC for the Fitzroy Estuary (Great Barrier Reef), and a parametric flocculation model that only relates settling velocity to SSC. We present an example of the ML-based approach outperforming a parametric model by capturing nonlinear particle-hydrodynamic interactions while maintaining computational efficiency, enabling high-resolution SSC simulations. This work demonstrates an advancement for hybrid modelling using rapidly evolving ML applications, offering a scalable tool for sediment transport and water quality management.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"198 ","pages":"Article 102621"},"PeriodicalIF":2.9,"publicationDate":"2025-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144886821","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Evaluating effectiveness of round-off error compensation with three methods in shallow-water models 评估三种方法在浅水模型中舍入误差补偿的有效性

IF 2.9 3区地球科学

Ocean Modelling Pub Date : 2025-08-05 DOI: 10.1016/j.ocemod.2025.102617

Jiayi Lai , Lanning Wang , Yizhou Yang , Qizhong Wu , Mengxuan Chen

{"title":"Evaluating effectiveness of round-off error compensation with three methods in shallow-water models","authors":"Jiayi Lai , Lanning Wang , Yizhou Yang , Qizhong Wu , Mengxuan Chen","doi":"10.1016/j.ocemod.2025.102617","DOIUrl":"10.1016/j.ocemod.2025.102617","url":null,"abstract":"<div><div>High-performance computing (HPC) limitations remain a significant bottleneck in the development of numerical models. Mixed-precision techniques, which reduce arithmetic precision to improve speed and memory efficiency, offer a promising solution. However, these methods inevitably introduce increased round-off errors that may destabilize model integrations and require smaller integration steps. This study investigates whether round-off error compensation methods can mitigate such precision-reduced errors. Three widely used methods are evaluated including Gill, Kahan, and Quasi Double-Precision (QDP) within shallow-water models. The suitability of using the double-precision fourth-order Runge-Kutta (RK4-DBL) method as a benchmark is first validated through idealized 1D linear shallow-water model experiments with known analytical solutions. Subsequently, ten perturbed initial-condition experiments are conducted for 2D nonlinear shallow-water model to assess the robustness of each compensation method relative to the RK4-DBL benchmark. When applied to fourth-order Runge-Kutta (RK4) in single precision (RK4-SGL), the Gill, Kahan and QDP methods reduce surface-height root-mean-square (RMSE) errors by approximately one order, four orders, and half an order of magnitude, respectively. In terms of computational cost, runtimes increased by 53%, 4%, and 7% relative to the double-precision reference, respectively. Among these compensation methods, the Kahan method achieves the best performance in both error compensation and computational efficiency, followed by the Gill method. The QDP method, though less effective than the other two, still provides meaningful improvements. Overall, this study demonstrates that these three round-off error compensation methods can improve the accuracy of mixed-precision numerical models while maintaining a reasonable computational cost.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"198 ","pages":"Article 102617"},"PeriodicalIF":2.9,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144779315","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Assimilation effect of serial observation data in the East Asian Marginal Seas for long period 东亚边缘海长期连续观测资料的同化效应

IF 2.9 3区地球科学

Ocean Modelling Pub Date : 2025-08-05 DOI: 10.1016/j.ocemod.2025.102605

Jae-Ho Lee , You-Soon Chang , Yong Sun Kim , Yang-Ki Cho

{"title":"Assimilation effect of serial observation data in the East Asian Marginal Seas for long period","authors":"Jae-Ho Lee , You-Soon Chang , Yong Sun Kim , Yang-Ki Cho","doi":"10.1016/j.ocemod.2025.102605","DOIUrl":"10.1016/j.ocemod.2025.102605","url":null,"abstract":"<div><div>Despite the long-standing importance of the serial observation system since the 1960s in the East Asian Marginal Seas (EAMSs), research on the contribution of this valuable data to ocean analysis remains limited. In this study, an Observing System Simulation Experiment (OSSE) was conducted to assess the data assimilation effects of this serial observation system. The OSSE was applied to 22 serial observation lines, with different assimilation periods.</div><div>Results show that the best assimilation performance was achieved with the 2-month cycle, which matches the real observation system's period. In the surface layer, the 10-day and 1-month cycles exhibited poorer performance due to an increase in warm bias in the northern part of the East/Japan Sea. In contrast, for the deep layer below 500 m where no serial observation data is available, the 10-day and 1-month cycles showed better performance in short-term simulations for the first seven years for 2012–2018. This improvement is linked to the downward current generated in the northern East/Japan Sea.</div><div>In long-term simulations for 2019∼2041, the 2-month cycle demonstrated superior performance, likely due to signal propagation by the southward deep current, which is part of the meridional overturning circulation. These findings were also supported by results from the reverse bias experiment, although the physical mechanisms for interpreting the data assimilation process differ. This study provides valuable insights for long-term ocean prediction and highlights the significance of the serial observation system in enhancing ocean analysis.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"198 ","pages":"Article 102605"},"PeriodicalIF":2.9,"publicationDate":"2025-08-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144828144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Explainable artificial intelligence of machine and deep learning algorithms for multi-output prediction of wave characteristics 可解释的机器人工智能和深度学习算法，用于多输出波特性预测

IF 2.9 3区地球科学

Ocean Modelling Pub Date : 2025-07-31 DOI: 10.1016/j.ocemod.2025.102604

Zaid Allal , Hassan N. Noura , Ola Salman , Khaled Chahine

{"title":"Explainable artificial intelligence of machine and deep learning algorithms for multi-output prediction of wave characteristics","authors":"Zaid Allal , Hassan N. Noura , Ola Salman , Khaled Chahine","doi":"10.1016/j.ocemod.2025.102604","DOIUrl":"10.1016/j.ocemod.2025.102604","url":null,"abstract":"<div><div>Accurately predicting wave characteristics is essential for efficiently harnessing wave energy and ensuring safe maritime operations. This paper compares thirteen machine and deep learning algorithms to forecast wave characteristics using data from a buoy installation in Mooloolaba, Queensland, Australia. The approach diverges from tradition by making multi-output predictions across six wave characteristics, providing a more comprehensive understanding of wave behavior. In addition, it delves into the inner workings of the most effective models through explainable artificial intelligence, revealing the intricate mechanisms underlying their superior performance. The results showcase excellent model performance with minimal error values when dealing with multi-output regression challenges. The results underscore the remarkable potential of these algorithms to predict upcoming wave data on both short-term (30 min) and near-term (1-hour) horizons, allowing for timely intervention for nearshore device maintenance and activation of alert systems.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"198 ","pages":"Article 102604"},"PeriodicalIF":2.9,"publicationDate":"2025-07-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144771438","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Effectiveness of the Offline Fennel model for biogeochemical simulations in the Mediterranean Sea 离线Fennel模型在地中海生物地球化学模拟中的有效性

IF 2.9 3区地球科学

Ocean Modelling Pub Date : 2025-07-25 DOI: 10.1016/j.ocemod.2025.102596

Júlia Crespin , Morane Clavel-Henry , Miquel Canals , Kristen M. Thyng , Veronica Ruiz-Xomchuk , Jordi Solé

{"title":"Effectiveness of the Offline Fennel model for biogeochemical simulations in the Mediterranean Sea","authors":"Júlia Crespin , Morane Clavel-Henry , Miquel Canals , Kristen M. Thyng , Veronica Ruiz-Xomchuk , Jordi Solé","doi":"10.1016/j.ocemod.2025.102596","DOIUrl":"10.1016/j.ocemod.2025.102596","url":null,"abstract":"<div><div>Modeling the distribution of biogeochemical components in the ocean is essential for further understanding climate change impacts and assess the functioning of marine ecosystems. This requires robust and efficient physical-biological simulations of coupled ocean-ecosystem models, which are often hindered by limited data availability and computational resources. The option of running biological tracer fields offline, independently from the physical ocean simulation, is appealing due to increased computational efficiency. Here, we present an assessment and implementation of an offline biogeochemical model — the Offline Fennel model — within the Regional Ocean Modeling System (ROMS). Our methodology employs ROMS hydrodynamic outputs to run the biogeochemical model offline. This work also includes the first ground-truthing exercise of the referred offline biogeochemical model. We use a variety of skill metrics to compare the simulated surface chlorophyll to an ocean color dataset (Copernicus Marine Service Mediterranean Ocean Color) and BGC-Argo floats for the 2015–2020 period. The model is able to reproduce the temporal and spatial structures of the main chlorophyll fluctuation patterns in the study area, the Northwestern Mediterranean Sea. This area is of particular interest as it is one of the most productive regions in the entire Mediterranean Basin, with open-ocean upwellings and deep winter convection events occurring seasonally. The typical behavior of the region is likewise effectively represented in the implementation, including offshore primary production, nutrient supplies from the Rhone and Ebro rivers, and mesoscale hydrographic structures. This study provides a baseline for ROMS users in need of executing more biogeochemical simulations independently from more computationally demanding physical simulations.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"198 ","pages":"Article 102596"},"PeriodicalIF":2.9,"publicationDate":"2025-07-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144723023","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Quantification of multi-source contributions to volume transport in the Tsushima Strait 对马海峡体积运输多源贡献的量化

IF 2.9 3区地球科学

Ocean Modelling Pub Date : 2025-07-24 DOI: 10.1016/j.ocemod.2025.102597

Ziyin Meng , Qiyan Ji , Hui Chen , Guantong Lv

{"title":"Quantification of multi-source contributions to volume transport in the Tsushima Strait","authors":"Ziyin Meng , Qiyan Ji , Hui Chen , Guantong Lv","doi":"10.1016/j.ocemod.2025.102597","DOIUrl":"10.1016/j.ocemod.2025.102597","url":null,"abstract":"<div><div>As the sole channel connecting both the East China Sea and the Yellow Sea to the Sea of Japan, the contribution of specific sources of volume transport in the Tsushima Strait remains unknown. Using the Lagrangian trajectory model TRACMASS, this study identifies the sources of volume transport through the Tsushima Strait and quantifies the contributions from three major pathways: the East Taiwan Channel, the Taiwan Strait, and the Northern Yellow Sea. The model accurately reproduces the persistent quasi-unidirectional current that transports water through the Tsushima Strait into the Sea of Japan throughout the year. The East Taiwan Channel and Taiwan Strait serve as the principal contributors to the Tsushima Strait’s volume transport, accounting for 1.58 Sv (51.8 %) and 1.04 Sv (34.1 %), respectively. However, the Northern Yellow Sea makes a relatively minor contribution of 0.17 Sv (5.7 %). Volume transport through the Tsushima Strait is primarily sourced from the East Taiwan Channel for most of the year, exhibiting bimodal seasonal peaks in April and November driven by intensified Kuroshio shelf intrusion. In contrast, the Taiwan Strait becomes the dominant contributor during August and September, when strengthened transport occurs under the combined influence of the Taiwan Warm Current and monsoon transition. Volume transport from the Northern Yellow Sea to the Tsushima Strait is primarily driven by the Korean Coastal Current. In addition, the runoff and long-resident water masses retained within the East China Sea and the Yellow Sea also serve as supplementary contributors to the transport of Tsushima Strait.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"198 ","pages":"Article 102597"},"PeriodicalIF":2.9,"publicationDate":"2025-07-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144749737","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Real-time wave model error correction via coupled neural networks and WAM under extreme weather 极端天气下基于耦合神经网络和WAM的实时波模型误差校正

IF 2.9 3区地球科学

Ocean Modelling Pub Date : 2025-07-23 DOI: 10.1016/j.ocemod.2025.102600

Aiyue Liu , Xiaofeng Li , Dongliang Shen

{"title":"Real-time wave model error correction via coupled neural networks and WAM under extreme weather","authors":"Aiyue Liu , Xiaofeng Li , Dongliang Shen","doi":"10.1016/j.ocemod.2025.102600","DOIUrl":"10.1016/j.ocemod.2025.102600","url":null,"abstract":"<div><div>Accurate forecasts of wave parameters, especially significant wave height, are essential for maritime operations, yet predicting wave heights during extreme weather remains difficult due to rapid error growth in numerical models. This study presents a real-time error correction framework that couples a spatiotemporal attention-based neural network with the WAM wave model. The correction network is trained using CFOSAT satellite observations and dynamically coupled with WAM via a Fortran–Python interface. Applied to 114 typhoon events in the Northwest Pacific, the system reduces significant wave height (SWH) root mean square error (RMSE) by 24.6 % and increases the structural similarity index (SSIM) by 26.3 %, compared to WAM predictions made with default tuning parameters. Validation across 32 tropical cyclones with diverse intensities in the Gulf of Mexico shows strong generalization, achieving up to a 47 % reduction in RMSE and enhancing wave spectral accuracy by >30 %. These results highlight the robustness and scalability of this hybrid AI-physics framework, demonstrating its practical value for real-time wave forecasting during extreme weather events.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"198 ","pages":"Article 102600"},"PeriodicalIF":2.9,"publicationDate":"2025-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144723022","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Pre-trained physics-informed neural networks for one-dimensional wave propagation in coastal engineering 海岸工程中一维波浪传播的预训练物理信息神经网络

IF 3.1 3区地球科学

Ocean Modelling Pub Date : 2025-07-23 DOI: 10.1016/j.ocemod.2025.102601

Yunlong Yang , Feng Luo , Zhipeng Chen , Aifeng Tao , Hongping Zhao , Yongfu Dong , Peng Tian , Jinhai Zheng

{"title":"Pre-trained physics-informed neural networks for one-dimensional wave propagation in coastal engineering","authors":"Yunlong Yang , Feng Luo , Zhipeng Chen , Aifeng Tao , Hongping Zhao , Yongfu Dong , Peng Tian , Jinhai Zheng","doi":"10.1016/j.ocemod.2025.102601","DOIUrl":"10.1016/j.ocemod.2025.102601","url":null,"abstract":"<div><div>Modeling wave propagation over variable coastal topographies remains challenging due to the interplay of nonlinear shallow‑water dynamics and dispersive effects. Here, we introduce a Pre‑Trained Physics‑Informed Neural Network (PT‑PINN) framework that couples a physics‑guided pre‑training phase with rigorous cross‑validation to solve the Saint‑Venant and Boussinesq equations. During pre‑training, the network generates a physics‑informed initial approximation, constructs auxiliary supervised data, and yields optimized parameter seeds, all of which accelerate and stabilize subsequent formal training. Cross‑validation on the pre‑training–derived dataset then guides hyperparameter selection, ensuring an effective balance between physics‑driven and data‑driven loss components.</div><div>We demonstrate the PT‑PINN approach across four benchmark scenarios: (1) dam‑break flow in a wet domain, (2) non‑breaking wave propagation on inclined slopes, (3) periodic tidal waves in an inclined open channel, and (4) shoaling waves over a submerged breakwater. In each case, PT‑PINNs faithfully capture both bulk wave evolution and fine‑scale dispersive details. Comparative studies against analytical and finite‑difference solutions reveal that PT‑PINNs match their accuracy while offering enhanced stability in representing high‑frequency microscale features. These results underscore the promise of pre‑trained, physics‑informed networks as a versatile and robust tool for coastal wave modeling in complex bathymetric settings.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"198 ","pages":"Article 102601"},"PeriodicalIF":3.1,"publicationDate":"2025-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144713445","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Shelf-wide circulation impacts the flushing time of coastal bays 大陆架环流影响海岸海湾的冲刷时间

IF 3.1 3区地球科学

Ocean Modelling Pub Date : 2025-07-23 DOI: 10.1016/j.ocemod.2025.102603

Junwei Hua , Jiabi Du , Kyeong Park

{"title":"Shelf-wide circulation impacts the flushing time of coastal bays","authors":"Junwei Hua , Jiabi Du , Kyeong Park","doi":"10.1016/j.ocemod.2025.102603","DOIUrl":"10.1016/j.ocemod.2025.102603","url":null,"abstract":"<div><div>Once exiting an estuary into the shelf, transport and retention of materials are subject to the shelf dynamics, and sometimes the deep ocean dynamics. However, the impact of the deep ocean is rarely considered in previous coastal modeling studies, as coastal models typically have a fine resolution only for the coastal region and the domain rarely extends beyond the shelf (depth <200 m). This study demonstrates the role of deep and shelf ocean circulation on the flushing of estuarine bays. With a cross-scale and well-calibrated ocean model for the northwestern Gulf of Mexico (<em>Coarse Small Model</em>) and another one for the entire Gulf of Mexico (<em>Refined Large Model</em>), we examine the flushing time for Galveston Bay through Lagrangian particle-tracking simulations. Both models have similar results regarding salinity and currents near the coast, but <em>Coarse Small Model</em> persistently overestimates/underestimates the flushing time during winter/summer, respectively, compared to <em>Refined Large Model</em>. Analysis of sea surface height and geostrophic currents suggests that <em>Coarse Small Model</em>’s inability to capture the deep ocean synoptic circulations leads to the overestimations of estuarine materials’ retention on the inner shelf and unrealistic flushing time for coastal bays during winter. By increasing the resolution in the deep Gulf from 10 to 5 km, <em>Refined Small Model</em> produces results similar to <em>Refined Large Model</em>. This study highlights the role of shelf and deep ocean dynamics on exchange between estuarine bays and coastal ocean and emphasizes the importance of resolving the shelf-wide dynamics in models focusing on estuarine and coastal waters.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"198 ","pages":"Article 102603"},"PeriodicalIF":3.1,"publicationDate":"2025-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144713613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Machine learning-based intelligent parameterization of source functions in numerical wave model 数值波模型中基于机器学习的源函数智能参数化

IF 3.1 3区地球科学

Ocean Modelling Pub Date : 2025-07-22 DOI: 10.1016/j.ocemod.2025.102602

Fuhua Huang , Zeyu Wang , Longyu Jiang , Feng Hua

{"title":"Machine learning-based intelligent parameterization of source functions in numerical wave model","authors":"Fuhua Huang , Zeyu Wang , Longyu Jiang , Feng Hua","doi":"10.1016/j.ocemod.2025.102602","DOIUrl":"10.1016/j.ocemod.2025.102602","url":null,"abstract":"<div><div>In recent years, although the application of machine learning in parameterizing complex marine physical processes has gradually become widespread, most of the existing studies rely on statistically correlated parameter selection methods for neural network construction and lack physical support. This study proposed a physics-guided neural network parameterization method combining physical feature selection and data-driven modeling. By integrating source function parameterization equations (wind input, wave breaking dissipation, wave-wave nonlinear interactions) from the MASNUM-WAM physical framework into the feature engineering of a backpropagation neural network (BPNN), a physically guided parameterization model was developed. The experiments show that the three major source functions exhibit excellent prediction performance (R²>0.95, RMSE<0.09, BIAS between -0.02 and 0.05), with stable results across multi-test points. Then, a new directional wave spectra prediction model was developed using the prediction results. Directional wave spectra predictions show strong consistency with MASNUM-WAM (COR>0.92, RMSE<0.09 m²s, |BIAS|≤0.03 m²s). Spectral integration parameters achieve high accuracy: significant wave height (RMSE≤0.477 m), mean wave direction (RMSE≤1.010°), and mean wave period (0.203 s≤RMSE≤0.247 s). Feature importance analysis reveals that wave breaking dissipation contributes most substantially to directional wave spectra prediction accuracy, while initial conditions, wave-wave nonlinear interaction, wind field components exhibit variable influence, and wind input term maintains a minor but consistent role. This physics-guided approach retains data-driven advantages while enhancing model reliability and computational efficiency, offering a new pathway for parametric research in ocean wave simulation.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"198 ","pages":"Article 102602"},"PeriodicalIF":3.1,"publicationDate":"2025-07-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144713629","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0