Pub Date : 2026-01-27DOI: 10.1016/j.ocemod.2026.102692
Dandan Li , Changjiang Xiao , Min Huang , Qiquan Yang , Xiong Xu , Yingbing Liu
Accurate long-lead daily Sea Surface Salinity prediction remains a significant challenge, primarily attributed to complex oceanic dynamics and the cumulative propagation of prediction errors over extended time frames. Existing methodologies, encompassing classical machine learning approaches and recurrent deep learning architectures, struggle to balance computational efficiency with the modeling of long-range temporal dependencies in SSS time series. This study introduces a temporal convolutional network (TCN)-based model to address these challenges, leveraging dilated causal convolutions to model multi-scale SSS dynamics while mitigating error accumulation in long-lead forecasting. The proposed deep learning model enables the automatic capture and modeling of SSS temporal dependencies using only historical time-series SSS data from satellite remote sensing. Comprehensive experiments conducted at eight geographically dispersed sites in the Indian Ocean, utilizing European Space Agency (ESA) Climate Change Initiative (CCI) Level 4 (L4) SSS satellite data, demonstrate that the proposed model outperforms both baseline machine learning and deep learning models, demonstrating its superior capability for long-lead daily SSS prediction.
{"title":"Long-lead daily sea surface salinity prediction using time-series CCI L4 SSS satellite data and a temporal convolutional deep learning model","authors":"Dandan Li , Changjiang Xiao , Min Huang , Qiquan Yang , Xiong Xu , Yingbing Liu","doi":"10.1016/j.ocemod.2026.102692","DOIUrl":"10.1016/j.ocemod.2026.102692","url":null,"abstract":"<div><div>Accurate long-lead daily Sea Surface Salinity prediction remains a significant challenge, primarily attributed to complex oceanic dynamics and the cumulative propagation of prediction errors over extended time frames. Existing methodologies, encompassing classical machine learning approaches and recurrent deep learning architectures, struggle to balance computational efficiency with the modeling of long-range temporal dependencies in SSS time series. This study introduces a temporal convolutional network (TCN)-based model to address these challenges, leveraging dilated causal convolutions to model multi-scale SSS dynamics while mitigating error accumulation in long-lead forecasting. The proposed deep learning model enables the automatic capture and modeling of SSS temporal dependencies using only historical time-series SSS data from satellite remote sensing. Comprehensive experiments conducted at eight geographically dispersed sites in the Indian Ocean, utilizing European Space Agency (ESA) Climate Change Initiative (CCI) Level 4 (L4) SSS satellite data, demonstrate that the proposed model outperforms both baseline machine learning and deep learning models, demonstrating its superior capability for long-lead daily SSS prediction.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"201 ","pages":"Article 102692"},"PeriodicalIF":2.9,"publicationDate":"2026-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146081501","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}
Pub Date : 2026-01-24DOI: 10.1016/j.ocemod.2026.102691
Zhe Guo , Zhiqiang Liu , Zhongya Cai
Surface circulation in the South China Sea (SCS), primarily driven by water exchange through the Luzon Strait and regional wind forcing, exhibits a strong seasonal cycle, typically intensifying in winter and weakening in summer. However, this seasonality varies significantly across the basin, reflecting complex interactions between local dynamics and external forcing. Using satellite altimetry and numerical simulations, this study identifies a latitudinal dependence in the timing of surface circulation transitions. From south to north, the decay phase, when mean kinetic energy declines from its seasonal peak, becomes progressively longer, while the growth phase shortens. Energy budget analysis reveals that in the northern SCS, mean kinetic energy is sustained longer due to joint contributions from local wind power and external kinematic energy (KE) input. In contrast, the southern SCS experiences a rapid drop in KE, driven primarily by a sharp decline in wind power. This spatial pattern also varies interannually, modulated by the El Niño–Southern Oscillation (ENSO). In the south, decay phase duration is positively correlated with ENSO strength, largely due to ENSO-driven variations in wind stress. In the north, ENSO influences wind stress and Kuroshio intrusion in opposite ways, resulting in a negative correlation between ENSO and decay time. These findings enhance our understanding of how large-scale climate variability modulates marginal sea circulation and offer new insights for improving regional ocean modeling.
{"title":"Latitudinal dependence of circulation seasonality in the South China Sea and its response to ENSO","authors":"Zhe Guo , Zhiqiang Liu , Zhongya Cai","doi":"10.1016/j.ocemod.2026.102691","DOIUrl":"10.1016/j.ocemod.2026.102691","url":null,"abstract":"<div><div>Surface circulation in the South China Sea (SCS), primarily driven by water exchange through the Luzon Strait and regional wind forcing, exhibits a strong seasonal cycle, typically intensifying in winter and weakening in summer. However, this seasonality varies significantly across the basin, reflecting complex interactions between local dynamics and external forcing. Using satellite altimetry and numerical simulations, this study identifies a latitudinal dependence in the timing of surface circulation transitions. From south to north, the decay phase, when mean kinetic energy declines from its seasonal peak, becomes progressively longer, while the growth phase shortens. Energy budget analysis reveals that in the northern SCS, mean kinetic energy is sustained longer due to joint contributions from local wind power and external kinematic energy (KE) input. In contrast, the southern SCS experiences a rapid drop in KE, driven primarily by a sharp decline in wind power. This spatial pattern also varies interannually, modulated by the El Niño–Southern Oscillation (ENSO). In the south, decay phase duration is positively correlated with ENSO strength, largely due to ENSO-driven variations in wind stress. In the north, ENSO influences wind stress and Kuroshio intrusion in opposite ways, resulting in a negative correlation between ENSO and decay time. These findings enhance our understanding of how large-scale climate variability modulates marginal sea circulation and offer new insights for improving regional ocean modeling.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"201 ","pages":"Article 102691"},"PeriodicalIF":2.9,"publicationDate":"2026-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146081499","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}
Pub Date : 2026-01-19DOI: 10.1016/j.ocemod.2026.102684
Zhiyong Peng , Jiehua Wu , Peng Wang
Storm surge represents a significant coastal hazard in Fujian Province, where intense wave activity can exacerbate inundation by imparting additional momentum and mass flux to coastal waters. Accurate simulation of storm surge events requires explicit consideration of wave-current interactions, particularly wave-enhanced bottom stress. In this study, the effects of wave-enhanced bottom stress on storm surge induced by Typhoon Doksuri (2023) are investigated using numerical simulations. Results demonstrate that wave-enhanced bottom stress contributes to a 4-8 % increase in nearshore storm surge by modifying volume transport through the strait. Moreover, alternative parameterizations of bottom stress produce substantial differences in simulated surge magnitude. These findings highlight the critical importance of appropriately representing wave-enhanced bottom stress in numerical models to improve the reliability of storm surge and coastal inundation forecasts.
{"title":"How does wave-enhanced bottom stress affect typhoon-induced storm surge","authors":"Zhiyong Peng , Jiehua Wu , Peng Wang","doi":"10.1016/j.ocemod.2026.102684","DOIUrl":"10.1016/j.ocemod.2026.102684","url":null,"abstract":"<div><div>Storm surge represents a significant coastal hazard in Fujian Province, where intense wave activity can exacerbate inundation by imparting additional momentum and mass flux to coastal waters. Accurate simulation of storm surge events requires explicit consideration of wave-current interactions, particularly wave-enhanced bottom stress. In this study, the effects of wave-enhanced bottom stress on storm surge induced by Typhoon Doksuri (2023) are investigated using numerical simulations. Results demonstrate that wave-enhanced bottom stress contributes to a 4-8 % increase in nearshore storm surge by modifying volume transport through the strait. Moreover, alternative parameterizations of bottom stress produce substantial differences in simulated surge magnitude. These findings highlight the critical importance of appropriately representing wave-enhanced bottom stress in numerical models to improve the reliability of storm surge and coastal inundation forecasts.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"200 ","pages":"Article 102684"},"PeriodicalIF":2.9,"publicationDate":"2026-01-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146034765","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}
Pub Date : 2026-01-12DOI: 10.1016/j.ocemod.2026.102683
Hui Xu , Kai Yin , Sudong Xu , Pengju Han , Shangpeng Gong
Global warming has resulted in rising sea levels and an increased frequency of extreme weather events, intensifying the need for effective coastal defences against marine disasters. Nature-based solutions, which offer benefits such as wave dissipation, sediment deposition, ecological compatibility, and environmental adaptability, have emerged as a crucial strategy for coastal protection. Previous studies have predominantly modelled vegetation as simplified cylinders to examine its role in wave attenuation, often ignoring the potential influence of vegetation’s branching structure. Based on these, this research constructs reasonable flume experiments to analyse the discrepancy in wave-attenuation performance between typical branched flexible vegetation and non-branched flexible vegetation. The experiments involve varying water depths, wave heights, and wave periods to comprehensively evaluate the impact of vegetation structure on wave propagation. Comparative analysis results indicate that the branching structure significantly enhances the wave-absorbing capacity of flexible vegetation. Within the tested simulation parameters, the average wave attenuation effect in flexible branched vegetation zones (FBV) was approximately 47 % higher than in flexible unbranched vegetation zones (FUV), with this quantitative difference varying considerably under different hydrodynamic conditions. Furthermore, the findings suggest that the impact of vegetation on wave attenuation is more pronounced under conditions of greater wave heights, shallower water depths, and longer wave periods. On the other hand, the function of vegetation's branching structure in wave reduction becomes more significant as the water depth increases or the wave height decreases. Physical modelling experiments revealing the specific influence of branching structure on wave attenuation can provide a scientific foundation and practical guidance for designing more effective ecological coastal protection measures.
{"title":"Effects of monopodial branching flexible vegetation on the wave attenuation by vegetation","authors":"Hui Xu , Kai Yin , Sudong Xu , Pengju Han , Shangpeng Gong","doi":"10.1016/j.ocemod.2026.102683","DOIUrl":"10.1016/j.ocemod.2026.102683","url":null,"abstract":"<div><div>Global warming has resulted in rising sea levels and an increased frequency of extreme weather events, intensifying the need for effective coastal defences against marine disasters. Nature-based solutions, which offer benefits such as wave dissipation, sediment deposition, ecological compatibility, and environmental adaptability, have emerged as a crucial strategy for coastal protection. Previous studies have predominantly modelled vegetation as simplified cylinders to examine its role in wave attenuation, often ignoring the potential influence of vegetation’s branching structure. Based on these, this research constructs reasonable flume experiments to analyse the discrepancy in wave-attenuation performance between typical branched flexible vegetation and non-branched flexible vegetation. The experiments involve varying water depths, wave heights, and wave periods to comprehensively evaluate the impact of vegetation structure on wave propagation. Comparative analysis results indicate that the branching structure significantly enhances the wave-absorbing capacity of flexible vegetation. Within the tested simulation parameters, the average wave attenuation effect in flexible branched vegetation zones (FBV) was approximately 47 % higher than in flexible unbranched vegetation zones (FUV), with this quantitative difference varying considerably under different hydrodynamic conditions. Furthermore, the findings suggest that the impact of vegetation on wave attenuation is more pronounced under conditions of greater wave heights, shallower water depths, and longer wave periods. On the other hand, the function of vegetation's branching structure in wave reduction becomes more significant as the water depth increases or the wave height decreases. Physical modelling experiments revealing the specific influence of branching structure on wave attenuation can provide a scientific foundation and practical guidance for designing more effective ecological coastal protection measures.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"200 ","pages":"Article 102683"},"PeriodicalIF":2.9,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146034766","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}
Pub Date : 2026-01-06DOI: 10.1016/j.ocemod.2026.102682
Zhor Benhafid , Elise Mayrand , Sid Ahmed Selouani
Ocean acidification poses a growing threat to marine ecosystems and aquaculture productivity, particularly in under-monitored coastal regions such as Eastern Canada. Existing pH prediction frameworks typically rely on multi-year records combining extensive carbonate chemistry, physical, and biological parameters. While these models can achieve high accuracy, their data requirements make them costly, complex, and challenging to implement for local, site-specific acidification forecasting in aquaculture contexts. To address this limitation, this study benchmarks several machine learning models for coastal pHSWS prediction using only three routinely measured environmental variables (temperature, salinity, sea level), from which we derived moving-average descriptors, local gradients, and two temporal indicators, resulting in a compact set of 11 input features. Six different models and a multivariate linear regression baseline were trained on one of the most complete and extended high-frequency datasets available (BSSS2018) and evaluated across four independent datasets: one from the same site but six months earlier (BSSS2017), and three from nearby bays in northeastern New Brunswick collected between 2017 and 2019. Among all tested models, XGBoost emerged as the most reliable and interpretable, achieving the best trade-off between sensitivity and precision at the operational acidification threshold (). Its performance remained acceptable within-site but declined across bays due to environmental and seasonal discrepancies, underscoring the importance of training data representativeness. SHAP-based explainability confirmed that Julian day was the dominant predictor, integrating the composite effects of seasonal environmental variability. Overall, this study demonstrates that using only low-cost, routinely measured features provides a promising foundation for short-term coastal pH forecasting, particularly for aquaculture monitoring needs. Despite limited inter-bay generalization, the proposed framework shows that interpretable machine learning models can deliver actionable early-warning insights under realistic data constraints. It constitutes one of the first data-driven benchmarks explicitly tested at aquaculture-relevant thresholds, highlighting a scalable and transparent approach toward operational acidification forecasting.
{"title":"Explainable machine learning models for coastal pH forecasting at aquaculture-relevant thresholds in Eastern Canada","authors":"Zhor Benhafid , Elise Mayrand , Sid Ahmed Selouani","doi":"10.1016/j.ocemod.2026.102682","DOIUrl":"10.1016/j.ocemod.2026.102682","url":null,"abstract":"<div><div>Ocean acidification poses a growing threat to marine ecosystems and aquaculture productivity, particularly in under-monitored coastal regions such as Eastern Canada. Existing pH prediction frameworks typically rely on multi-year records combining extensive carbonate chemistry, physical, and biological parameters. While these models can achieve high accuracy, their data requirements make them costly, complex, and challenging to implement for local, site-specific acidification forecasting in aquaculture contexts. To address this limitation, this study benchmarks several machine learning models for coastal pH<sub>SWS</sub> prediction using only three routinely measured environmental variables (temperature, salinity, sea level), from which we derived moving-average descriptors, local gradients, and two temporal indicators, resulting in a compact set of 11 input features. Six different models and a multivariate linear regression baseline were trained on one of the most complete and extended high-frequency datasets available (BSSS2018) and evaluated across four independent datasets: one from the same site but six months earlier (BSSS2017), and three from nearby bays in northeastern New Brunswick collected between 2017 and 2019. Among all tested models, XGBoost emerged as the most reliable and interpretable, achieving the best trade-off between sensitivity and precision at the operational acidification threshold (<span><math><mrow><mi>p</mi><msub><mrow><mi>H</mi></mrow><mrow><mi>S</mi><mi>W</mi><mi>S</mi></mrow></msub><mo><</mo><mn>7</mn><mo>.</mo><mn>75</mn></mrow></math></span>). Its performance remained acceptable within-site but declined across bays due to environmental and seasonal discrepancies, underscoring the importance of training data representativeness. SHAP-based explainability confirmed that Julian day was the dominant predictor, integrating the composite effects of seasonal environmental variability. Overall, this study demonstrates that using only low-cost, routinely measured features provides a promising foundation for short-term coastal pH forecasting, particularly for aquaculture monitoring needs. Despite limited inter-bay generalization, the proposed framework shows that interpretable machine learning models can deliver actionable early-warning insights under realistic data constraints. It constitutes one of the first data-driven benchmarks explicitly tested at aquaculture-relevant thresholds, highlighting a scalable and transparent approach toward operational acidification forecasting.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"200 ","pages":"Article 102682"},"PeriodicalIF":2.9,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145938698","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}
Pub Date : 2026-01-06DOI: 10.1016/j.ocemod.2026.102681
Jie Yu , David D. Flagg , Tommy G. Jensen , Tim J. Campbell , Qing Wang , Denny P. Alappattu
We present a recent study to implement and test schemes for diagnostic calculations of skin sea surface temperature in the Navy’s Coastal Ocean Model (NCOM). This includes three schemes for estimating the cool anomaly in the viscous sublayer (i.e., the ocean skin), and a fourth scheme that adds an estimate of a warm anomaly in the solar radiation-driven, thermally stratified diurnal layer at near-surface depths. Applications of these schemes are made, and their performances are evaluated against field measurements from the Coupled Air-Sea Processes and Electromagnetic Ducting Research East campaign (CASPER-East), showing overall good agreements. The statistics of the model-observation comparisons are similar and do not indicate any systematic bias towards any scheme, but differences in the model performances are noticeable and vary depending on the surface wind and solar conditions. To understand the discrepancies among the schemes, inter-model comparisons are analyzed based on the conditions of surface wind stress and solar radiation flux. The issues associated with making the warm-layer correction are discussed, in particular, including the sensitivity of the diagnostic warm-layer anomaly to the layer thickness specified a priori, and the risk of double-counting the effect of solar radiation penetration when using the high-resolution NCOM temperature fields.
{"title":"Skin sea surface temperature diagnostics in a regional ocean model","authors":"Jie Yu , David D. Flagg , Tommy G. Jensen , Tim J. Campbell , Qing Wang , Denny P. Alappattu","doi":"10.1016/j.ocemod.2026.102681","DOIUrl":"10.1016/j.ocemod.2026.102681","url":null,"abstract":"<div><div>We present a recent study to implement and test schemes for diagnostic calculations of skin sea surface temperature in the Navy’s Coastal Ocean Model (NCOM). This includes three schemes for estimating the cool anomaly in the viscous sublayer (i.e., the ocean skin), and a fourth scheme that adds an estimate of a warm anomaly in the solar radiation-driven, thermally stratified diurnal layer at near-surface depths. Applications of these schemes are made, and their performances are evaluated against field measurements from the Coupled Air-Sea Processes and Electromagnetic Ducting Research East campaign (CASPER-East), showing overall good agreements. The statistics of the model-observation comparisons are similar and do not indicate any systematic bias towards any scheme, but differences in the model performances are noticeable and vary depending on the surface wind and solar conditions. To understand the discrepancies among the schemes, inter-model comparisons are analyzed based on the conditions of surface wind stress and solar radiation flux. The issues associated with making the warm-layer correction are discussed, in particular, including the sensitivity of the diagnostic warm-layer anomaly to the layer thickness specified <em>a priori</em>, and the risk of double-counting the effect of solar radiation penetration when using the high-resolution NCOM temperature fields.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"200 ","pages":"Article 102681"},"PeriodicalIF":2.9,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145938699","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}
Pub Date : 2026-01-03DOI: 10.1016/j.ocemod.2026.102678
Davide Grande , Roberto Buizza , Andrea Storto
This review examines recent advances in the application of machine learning to ocean data assimilation, covering contributions published between 2020 and 2025. We identify emerging trends, recurring limitations, and critical open questions, structuring the discussion around four scientific challenges: observation integration, boundary treatment, fine-scale process representation, and physical consistency. While convolutional neural networks remain widely used, particularly in bias correction and super-resolution tasks, recent studies increasingly employ multilayer perceptrons, long short-term memories, transformers and neural operators for error estimation, sequential bias correction, and latent-space assimilation. Despite this architectural diversity, most contributions remain confined to idealized configurations or offline modules, with limited evidence of generalization and integration into operational pipelines. We conclude that hybrid systems combining embedded physical knowledge with systematic validation across different oceanic regimes will be essential to unlock the full potential of machine learning-enhanced ocean data assimilation.
{"title":"Machine learning in ocean data assimilation: Advances, gaps and the road to operations","authors":"Davide Grande , Roberto Buizza , Andrea Storto","doi":"10.1016/j.ocemod.2026.102678","DOIUrl":"10.1016/j.ocemod.2026.102678","url":null,"abstract":"<div><div>This review examines recent advances in the application of machine learning to ocean data assimilation, covering contributions published between 2020 and 2025. We identify emerging trends, recurring limitations, and critical open questions, structuring the discussion around four scientific challenges: observation integration, boundary treatment, fine-scale process representation, and physical consistency. While convolutional neural networks remain widely used, particularly in bias correction and super-resolution tasks, recent studies increasingly employ multilayer perceptrons, long short-term memories, transformers and neural operators for error estimation, sequential bias correction, and latent-space assimilation. Despite this architectural diversity, most contributions remain confined to idealized configurations or offline modules, with limited evidence of generalization and integration into operational pipelines. We conclude that hybrid systems combining embedded physical knowledge with systematic validation across different oceanic regimes will be essential to unlock the full potential of machine learning-enhanced ocean data assimilation.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"200 ","pages":"Article 102678"},"PeriodicalIF":2.9,"publicationDate":"2026-01-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145938700","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}
Pub Date : 2026-01-03DOI: 10.1016/j.ocemod.2026.102680
Wenfan Wu , Zhengui Wang , Y Joseph Zhang , Jian Shen , Richard Tian , Lewis C Linker , Carl F Cerco
Estuaries, as transitional zones between land and ocean, exhibit highly nonlinear, cross-scale hydrodynamic processes that present substantial challenges for numerical modeling. Using Chesapeake Bay as an example, we demonstrate a physically based calibration procedure with observation-derived parametrizations, together with a high-resolution unstructured model without bathymetry smoothing. The results indicate that highly turbid water greatly affects the downward penetration of solar radiation, particularly in the upper Bay and tributaries. By incorporating the spatially varying Jerlov water types derived from satellite-based Kd490 data, we systematically improve water temperature simulations across the Bay, reducing the average RMSE to 0.484 °C (0.775 °C) for surface (bottom) temperature at 121 long-term monitoring stations maintained by EPA's Chesapeake Bay Program. Moreover, the presence of mud layers is found to facilitate tidal propagation in tributaries, thereby enhancing saltwater intrusion there. By applying spatially varying bottom drag coefficients calculated from the observed sediment types, we achieve significant improvements in salinity simulations, with an average RMSE of 0.809 PSU (1.331 PSU) for surface (bottom) salinity. In general, the present study reduces temperature and salinity errors by ∼60 % compared to previous modeling studies in the Bay. This study underscores the advantages of physically based calibration procedures that help make the model results more defensible.
{"title":"Improving cross-scale hydrodynamic simulations in the Chesapeake Bay with physically based calibration","authors":"Wenfan Wu , Zhengui Wang , Y Joseph Zhang , Jian Shen , Richard Tian , Lewis C Linker , Carl F Cerco","doi":"10.1016/j.ocemod.2026.102680","DOIUrl":"10.1016/j.ocemod.2026.102680","url":null,"abstract":"<div><div>Estuaries, as transitional zones between land and ocean, exhibit highly nonlinear, cross-scale hydrodynamic processes that present substantial challenges for numerical modeling. Using Chesapeake Bay as an example, we demonstrate a physically based calibration procedure with observation-derived parametrizations, together with a high-resolution unstructured model without bathymetry smoothing. The results indicate that highly turbid water greatly affects the downward penetration of solar radiation, particularly in the upper Bay and tributaries. By incorporating the spatially varying Jerlov water types derived from satellite-based Kd490 data, we systematically improve water temperature simulations across the Bay, reducing the average RMSE to 0.484 °C (0.775 °C) for surface (bottom) temperature at 121 long-term monitoring stations maintained by EPA's Chesapeake Bay Program. Moreover, the presence of mud layers is found to facilitate tidal propagation in tributaries, thereby enhancing saltwater intrusion there. By applying spatially varying bottom drag coefficients calculated from the observed sediment types, we achieve significant improvements in salinity simulations, with an average RMSE of 0.809 PSU (1.331 PSU) for surface (bottom) salinity. In general, the present study reduces temperature and salinity errors by ∼60 % compared to previous modeling studies in the Bay. This study underscores the advantages of physically based calibration procedures that help make the model results more defensible.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"200 ","pages":"Article 102680"},"PeriodicalIF":2.9,"publicationDate":"2026-01-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145976936","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}
Pub Date : 2026-01-02DOI: 10.1016/j.ocemod.2026.102677
Qingyu Zheng , Qi Shao , Guijun Han , Wei Li , Hong Li , Xuan Wang
Advances in ocean observation technology have significantly enhanced the accuracy of Earth system forecasting. Reconstructing missing information of nonlinear evolution processes from observational data is essential for investigating rapid changes in the marine environment and climate. However, traditional methods often struggle to extract unseen nonlinear processes from data. In fact, a large amount of dynamic evolution information hidden in historical data has not been effectively mined. To address this issue, we propose DeepDA, a latent space data assimilation approach based on deep learning. DeepDA employs a generative deep learning model to capture complex spatiotemporal multiscale features and nonlinear evolution processes in observations. By incorporating an attention mechanism, DeepDA effectively assimilates rich historical information of sea surface temperature. The results show that DeepDA remains highly stable in generating nonlinear evolution even with extensive data gaps and high noise levels. Notably, when only 10% (sparse sampling) of observation is available, the increase in error for DeepDA is limited to 40% compared to the case with complete data. Furthermore, DeepDA demonstrates effectiveness in multiscale reconstruction and analysis of climate variability, generating nonlinear patterns that are more physically consistent than linear methods. The nonlinear features extracted from the latent space exhibit multiscale structures, which may provide new insights into enhancing ocean data assimilation.
{"title":"Generating unseen nonlinear evolution in the ocean using deep learning-based latent space data assimilation model","authors":"Qingyu Zheng , Qi Shao , Guijun Han , Wei Li , Hong Li , Xuan Wang","doi":"10.1016/j.ocemod.2026.102677","DOIUrl":"10.1016/j.ocemod.2026.102677","url":null,"abstract":"<div><div>Advances in ocean observation technology have significantly enhanced the accuracy of Earth system forecasting. Reconstructing missing information of nonlinear evolution processes from observational data is essential for investigating rapid changes in the marine environment and climate. However, traditional methods often struggle to extract unseen nonlinear processes from data. In fact, a large amount of dynamic evolution information hidden in historical data has not been effectively mined. To address this issue, we propose DeepDA, a latent space data assimilation approach based on deep learning. DeepDA employs a generative deep learning model to capture complex spatiotemporal multiscale features and nonlinear evolution processes in observations. By incorporating an attention mechanism, DeepDA effectively assimilates rich historical information of sea surface temperature. The results show that DeepDA remains highly stable in generating nonlinear evolution even with extensive data gaps and high noise levels. Notably, when only 10% (sparse sampling) of observation is available, the increase in error for DeepDA is limited to 40% compared to the case with complete data. Furthermore, DeepDA demonstrates effectiveness in multiscale reconstruction and analysis of climate variability, generating nonlinear patterns that are more physically consistent than linear methods. The nonlinear features extracted from the latent space exhibit multiscale structures, which may provide new insights into enhancing ocean data assimilation.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"200 ","pages":"Article 102677"},"PeriodicalIF":2.9,"publicationDate":"2026-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145976937","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}
Pub Date : 2025-12-29DOI: 10.1016/j.ocemod.2025.102676
Silje Christine Iversen , Ann Kristin Sperrevik , Kai Håkon Christensen
Infrared satellite sea surface temperature (SST) observations capture the strong warming signal at the surface of the ocean during diurnal warming events. Assimilating these observations into ocean forecast models might introduce forecast errors, as the models do not fully resolve the strong diurnal cycle captured by SST observations. A standard approach is to remove affected SST observations from the assimilated dataset using a simplistic filter. However, this approach also removes observations not affected by diurnal warming. For high northern latitudes, where diurnal warming events are less frequent than at lower latitudes, it is unclear whether observations affected by strong warming negatively impact the forecasts, which would justify applying the filter. Here, we explore the consequences of both assimilating diurnal warming-affected SSTs and applying the filter using an ocean forecast model covering the seas off Norway. Through data assimilation experiments assimilating synthetic SSTs, we find that the strong warming signal in observations impacts forecasts in undesirable ways. The surface warming spreads below the mixed layer, decreasing its depth, and changes made to the model below the mixed layer persist into subsequent forecasts. Applying the filter reduces assimilated observations by 46 % without degrading forecasts, suggesting a redundancy in satellite SSTs.
{"title":"How does assimilating satellite SSTs affected by strong diurnal warming impact higher latitude ocean forecasts?","authors":"Silje Christine Iversen , Ann Kristin Sperrevik , Kai Håkon Christensen","doi":"10.1016/j.ocemod.2025.102676","DOIUrl":"10.1016/j.ocemod.2025.102676","url":null,"abstract":"<div><div>Infrared satellite sea surface temperature (SST) observations capture the strong warming signal at the surface of the ocean during diurnal warming events. Assimilating these observations into ocean forecast models might introduce forecast errors, as the models do not fully resolve the strong diurnal cycle captured by SST observations. A standard approach is to remove affected SST observations from the assimilated dataset using a simplistic filter. However, this approach also removes observations not affected by diurnal warming. For high northern latitudes, where diurnal warming events are less frequent than at lower latitudes, it is unclear whether observations affected by strong warming negatively impact the forecasts, which would justify applying the filter. Here, we explore the consequences of both assimilating diurnal warming-affected SSTs and applying the filter using an ocean forecast model covering the seas off Norway. Through data assimilation experiments assimilating synthetic SSTs, we find that the strong warming signal in observations impacts forecasts in undesirable ways. The surface warming spreads below the mixed layer, decreasing its depth, and changes made to the model below the mixed layer persist into subsequent forecasts. Applying the filter reduces assimilated observations by 46 % without degrading forecasts, suggesting a redundancy in satellite SSTs.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"200 ","pages":"Article 102676"},"PeriodicalIF":2.9,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145938697","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号