Ocean Modelling最新文献

{"title":"Dynamic effects of different turbulence closure schemes on Lagrangian residual velocity in an elongated bay","authors":"Fangjing Deng ,&nbsp;Hao Chen ,&nbsp;Xudong Liu ,&nbsp;Fei Ji ,&nbsp;Shuwen Zhang ,&nbsp;Zhaoyun Chen","doi":"10.1016/j.ocemod.2025.102528","DOIUrl":"10.1016/j.ocemod.2025.102528","url":null,"abstract":"<div><div>Different turbulence closure schemes significantly influence the structure of Lagrangian residual velocity (LRV), yet the underlying mechanisms remain inadequately understood. Under constant eddy viscosity coefficient conditions, the different parameter <em>β</em> (the ratio of the eddy viscosity term to the local acceleration term) predominantly governs the LRV by modulating the Lagrangian mean barotropic pressure gradient, while the Lagrangian mean eddy viscosity term exerts a negative feedback effect. Under varying eddy viscosity conditions, dominant components of total Lagrangian mean eddy viscosity term vary across turbulence closure schemes and the governing mechanisms of LRV become increasingly complex. The pioneering research meticulously tracks particle motion from the zero-velocity initial phase, establishing an equivalence between LRV and the two-time Lagrangian integrals of the acceleration and other dynamic terms. The two-time Lagrangian integrals of local acceleration and horizontal advection terms play the dominant roles in the LRV, while the vertical advection terms provide supplementary effects. Under non-stratified conditions, the two-time Lagrangian integrals of barotropic component generally counterbalance the two-time Lagrangian integrals of eddy viscosity component. In stratified contexts, upper-layer LRV in the along-estuary direction is influenced primarily by the two-time Lagrangian integrals of barotropic component, contributing up to half of the total LRV, while the lower-layer inflow is significantly shaped by the combined interaction of two-time Lagrangian integrals of eddy viscosity and baroclinic components. In the cross-estuary direction, unlike the along-estuary direction, the two-time Lagrangian integrals of eddy viscosity component contribute negatively to LRV.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"195 ","pages":"Article 102528"},"PeriodicalIF":3.1,"publicationDate":"2025-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143549772","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}

{"title":"Transit time of deep and intermediate waters in the Gulf of St. Lawrence","authors":"Shani Rousseau ,&nbsp;Diane Lavoie ,&nbsp;Mathilde Jutras ,&nbsp;Joël Chassé","doi":"10.1016/j.ocemod.2025.102526","DOIUrl":"10.1016/j.ocemod.2025.102526","url":null,"abstract":"<div><div>The transit time of the subsurface waters in the hypoxic and acidified Gulf of St. Lawrence (GSL) is poorly understood, despite its strong influence on physical and biogeochemical water properties. Three estimates of the transit time of the deep waters between Cabot Strait and the head of the Laurentian Channel, a deep channel cutting through the GSL, have been published up to now. Here, using lagrangian tracking experiments in a regional ocean model, we provide a new estimate of the transit time in the deep layer (&gt; 225 m) of the GSL, as well as the first estimate of the transit time in the intermediate layer (50–175 m). Our estimate for the deep layer is 3.2 ± 0.7 years. The transit time in the intermediate layer (1.2 ± 0.5 years) is nearly three times faster than in the deep layer. The deep waters travel mainly up the Laurentian Channel, whereas most of the intermediate waters first transit through the Esquiman and /or Anticosti Channels. Our results also highlighted the impact of the seasonal changes in large-scale circulation on the transit times of the particles seeded at Cabot Strait. In summer and fall, the circulation is relaxed, and subsurface waters transit slowly but more directly upstream, leading to faster transit times. In winter and spring, the circulation is intensified but many particles get caught in large gyres prevalent during these seasons, leading to slower average transit times.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"195 ","pages":"Article 102526"},"PeriodicalIF":3.1,"publicationDate":"2025-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143549771","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Inferring diapycnal mixing using the internal wave continuum from the high resolution ocean model","authors":"Shumin Jiang ,&nbsp;Dejun Dai ,&nbsp;Dingqi Wang ,&nbsp;Shihong Wang ,&nbsp;Ying Li ,&nbsp;Jingsong Guo ,&nbsp;Fangli Qiao","doi":"10.1016/j.ocemod.2025.102525","DOIUrl":"10.1016/j.ocemod.2025.102525","url":null,"abstract":"<div><div>Internal wave (IW)-induced mixing plays a crucial role in maintaining the thermohaline circulation. However, as most ocean general circulation models (OGCMs) do not resolve the scales of IWs and turbulence, to appropriately express the IW-induced turbulent dissipation is a long-lasting issue for ocean model development. Here we report the estimates of IW-induced turbulent dissipation in the South China Sea (SCS), from a tide-included OGCM using an internal wave continuum parameterization scheme (IWCP). The estimation is based on the energy level of internal wave continuum (IWC), by converting the finescale parameterization (FSP) from wavenumber to the frequency domain. The estimated dissipation rates are elevated over the Luzon Strait (LS), northern SCS, and continental slopes, up to <em>O</em>(10⁻⁷) Wkg⁻¹ between 100 and 500 m depth, while those in the central basin of SCS are <em>O</em>(10⁻¹⁰) Wkg⁻¹. The performance of IWCP are evaluated against observational datasets (Argo, CTD based FSP estimates, and CTD based Thorpe scale method estimates) and tidal-mixing parameterization schemes. The errors in IWCP's results are comparable with the discrepancies among different observational datasets. The IWCP significantly outperforms the widely used tidal mixing parameterization that only considers local internal tides, and are comparable, if not superior, to the tidal mixing parameterization that consider both local and non-local generation, depending on the choice of benchmark observational datasets. This work illustrates that IWCP could be an alternative efficient mean to estimate the three-dimensional map of IW-induced mixing from the high resolution OGCM.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"195 ","pages":"Article 102525"},"PeriodicalIF":3.1,"publicationDate":"2025-02-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143549773","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}

{"title":"Upper ocean biophysical budget analysis during a cyclone using Regional Ocean Modeling System","authors":"Abhijit Shee ,&nbsp;Sourav Sil ,&nbsp;Avijit Gangopadhyay ,&nbsp;Neeraj Agarwal ,&nbsp;K.K. Sandeep","doi":"10.1016/j.ocemod.2025.102524","DOIUrl":"10.1016/j.ocemod.2025.102524","url":null,"abstract":"<div><div>The biophysical response of upper ocean during the passage of the very severe cyclonic storm ‘Titli’ over Bay of Bengal (BoB) during October 2018 is studied using the Regional Ocean Modeling System coupled with nutrient-phytoplankton-zooplankton-detritus (ROMS-NPZD) framework. Assessments with satellite and Argo observations show that the model forced by Scatsat-I winds and reanalysis fluxes simulates the changes in the upper ocean reasonably well. The biophysical variability is significantly realized in a cyclone-induced open-ocean upwelling region underlying the peak intensity of the cyclone. After the passage of the cyclone, surface chlorophyll-a concentration increases tenfold in the upwelling region. This increase of simulated surface phytoplankton further enhances the surface dissolved oxygen concentration by ∼10μ<em>M</em> from an initial value of 200μ<em>M</em>. Strong temporal correspondences between the depth of the thermocline (<span><math><msub><mi>D</mi><mrow><msup><mn>23</mn><mo>∘</mo></msup><mi>C</mi></mrow></msub></math></span>) and the oxycline (<em>r</em> = 0.97) and between nutricline and the mixed-layer depth (MLD) (<em>r</em> = 0.52) are also observed. A detailed term-by-term biophysical budget analysis is carried out with temporal tendency of temperature (and other biophysical parameters) subdivided into various components: (i) net surface heat (or buoyancy) flux; (ii) horizontal advection; (iii) vertical entrainment; (iv) vertical mixing; and (v) source and sink terms as necessary. Results show the cooling of upper layer temperature during cyclone is due to vertical processes (entrainment and mixing), which is recovered in the post-cyclone period due to MLD advection affecting entrainment and vertical mixing, followed by restoration of net surface heat flux. The increase in salinity during the cyclone is due to MLD tendency dominating vertical entrainment process, and the post-cyclone recovery is due to the horizontal advection and net surface freshwater flux. The shoaling of the nutricline allows for the transport of the nutrients from the deeper layers to the euphotic zone and enhancement of phytoplankton concentration at the surface about five days after the cyclone interaction. Budget analysis for biological processes showed the changes in the phytoplankton concentration are associated with (i) nitrate uptake by phytoplankton and (ii) zooplankton grazing on phytoplankton. This assessment of the state-of-the-art coupled ROMS-NPZD model emphasizes its applicability for various other applications on marine ecosystem modeling.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"195 ","pages":"Article 102524"},"PeriodicalIF":3.1,"publicationDate":"2025-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143549774","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}

{"title":"Mixed layer depth parameterization and ocean surface cooling induced by tropical cyclones","authors":"Vladimir N. Kudryavtsev ,&nbsp;Pavel D. Pivaev","doi":"10.1016/j.ocemod.2025.102514","DOIUrl":"10.1016/j.ocemod.2025.102514","url":null,"abstract":"<div><div>Cooling of the ocean surface, induced by the tropical cyclone (TC), provides strong negative feedback leading to TC weakening. Parameterization of the upper ocean mixed layer (ML) depth and the sea surface temperature (SST) anomalies induced by TC is a key issue which is addressed in this paper. Using the one-dimensional heat balance equation and pre-storm vertical profiles of temperature, ML depths are estimated from SST anomalies observed by satellites in wakes of 417 TCs in different regions the World Ocean. To analyze the data, a self-similar approach is used that relates the ML depth to the ML transport and the buoyancy drop at the base of the ML. Expressing the ML wind drift transport in terms of TC parameters (maximum wind speed, translation speed and radius), the Coriolis parameter and underlying stratification, and also taking into account the influence of TC-induced upwelling, a parameterization of the ML depth is proposed, which is quite general. This parameterization combined with the heat conservation equation constitutes parameterization of the SST anomaly induced by the TC. Comparison with observations demonstrates effectiveness of the suggested parameterization in reproducing the SST anomalies in a wide range of surface cooling values (up to −10 °C) produced by TCs of different categories moving across various basins of the World Ocean with different stratification.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"195 ","pages":"Article 102514"},"PeriodicalIF":3.1,"publicationDate":"2025-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143511300","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}

{"title":"Simulating ice–wave interactions in the Laurentian Great Lakes using a fully coupled hydrodynamic–ice–wave model","authors":"M. Javad Javaherian ,&nbsp;David Cannon ,&nbsp;Jia Wang ,&nbsp;Ayumi Fujisaki-Manome ,&nbsp;Peng Bai ,&nbsp;Lei Zuo","doi":"10.1016/j.ocemod.2025.102513","DOIUrl":"10.1016/j.ocemod.2025.102513","url":null,"abstract":"<div><div>Hydrodynamic modeling in cold climate regions necessitates more sophisticated approaches that accurately simulate ice–wave interactions. Traditional models often overlook the complex coupling mechanisms between ice and ocean waves, especially the two-way processes where ice attenuates wave energy and waves break ice floes. This oversight can also intensify modeling challenges in coastal areas, including large lakes, where ice–wave interactions influence storm surges, high waves, and coastal erosion. To address this gap, this paper introduces an enhanced modeling approach that integrates both ice-induced wave attenuation and wave-induced ice breakage. To implement these processes, the Finite-Volume Community Ocean Model (FVCOM) is coupled with an unstructured-grid wave model (SWAN) and the unstructured-grid version of the Los Alamos Sea Ice Model (UG-CICE) to form the FVCOM–SWAVE–UG-CICE framework. Using this fully coupled model, simulations were conducted for the Great Lakes. Results of the modeled ice concentration, ice thickness, and significant wave heights were reported and validated against observational data from the U.S. National Ice Center (NIC) and in-situ under-ice measurements. To further study the coupling effects, results of the proposed model were also compared with those from no coupling and one-way coupling (focusing only on ice-induced wave attenuation) models. Comparative analyses demonstrated significant improvements in the predicted ice concentration with the proposed fully coupled model. These findings reveal the importance of incorporating ice–wave interactions in accurately predicting ice cover dynamics in freshwater systems.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"195 ","pages":"Article 102513"},"PeriodicalIF":3.1,"publicationDate":"2025-02-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143509550","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}

{"title":"Impact of offshore wind farm monopiles on hydrodynamics interacting with wind-driven waves","authors":"Seyed Taleb Hosseini ,&nbsp;Johannes Pein ,&nbsp;Joanna Staneva ,&nbsp;Y. Joseph Zhang ,&nbsp;Emil Stanev","doi":"10.1016/j.ocemod.2025.102521","DOIUrl":"10.1016/j.ocemod.2025.102521","url":null,"abstract":"<div><div>This paper investigates the local and regional impact of offshore wind farm (OWF) foundations on hydrodynamics in interaction with wind-induced waves at the Meerwind-OWF site (German Bight, North Sea) on tidal and monthly time scales. For this purpose, a 3D high-resolution coupled circulation-wave model based on unstructured grids is employed, which enables an effective transition in resolution from ∼2 km in marine open boundaries to ∼2 m near the foundations. The OWF monopiles induce different local and regional changes of the monthly mean velocity at mid-depth: a decrease of ∼5 % near the piles and an increase of ∼1 % in a wider region surrounding the OWF. The latter can be attributed to the relevant regional reduction in water density of ∼0.02 %. Consequently, the monthly potential energy anomaly increases by ∼5 % outside the OWF, while it decreases by 40 % inside it. The monopiles reduce the monthly significant wave height (Hs) from ∼5 % within the OWF to &lt;1 % over distances of ∼20 km. The prevailing westerly waves can affect the tidal asymmetry, particularly on the eastern side of the piles. This results in an asymmetry in the intensity of turbulent wakes on either side of the piles, in both monthly and tidal timescales. However, wave intensification can disrupt the periodic tidal pattern of the wake. An extreme event with Hs&gt;4 m creates a peak wake during the slack water that is higher than those at times of maximum tidal currents during spring tides.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"195 ","pages":"Article 102521"},"PeriodicalIF":3.1,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143480685","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}

{"title":"Neural emulator based on physical fields for accelerating the simulation of surface chlorophyll in an Earth System Model","authors":"Bizhi Wu ,&nbsp;Shiyao Zheng ,&nbsp;Shasha Li,&nbsp;Shanlin Wang","doi":"10.1016/j.ocemod.2024.102491","DOIUrl":"10.1016/j.ocemod.2024.102491","url":null,"abstract":"<div><div>Simulating the ocean biogeochemical module (BGC-enabled) in the Community Earth System Model (CESM) is computationally expensive, often requiring significantly more resources than the physical climate component. In this study, we propose an alternative approach to generate biogeochemical data using a neural network emulator, BGC-UNet, which predicts ocean surface chlorophyll concentrations based on physical fields from CESM, such as solar short-wave heat flux (SHF-QSW), potential temperature (TEMP), and zonal and meridional velocity (UVEL, VVEL). BGC-UNet is designed as a UNet-like architecture and employs a patch-based methodology with dilated sampling to efficiently reconstruct biogeochemical data from physical inputs. This framework potentially enables high-resolution chlorophyll predictions without running full BGC-enabled simulations. Our evaluation demonstrates that BGC-UNet’s outputs closely align with CESM’s simulated surface chlorophyll, supported by both quantitative metrics and visual analysis. Additionally, the emulator achieves a simulation speed approximately 248 times faster than traditional BGC-enabled CESM simulations. Although the current focus is on surface chlorophyll, the model shows potential for future extension to other biogeochemical variables. By leveraging only 40 years of simulated data for training, BGC-UNet replicates the trends observed in CESM, making it a promising tool for accelerating Earth system modeling.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"195 ","pages":"Article 102491"},"PeriodicalIF":3.1,"publicationDate":"2025-02-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143480093","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}

{"title":"A hybrid model based on chaos particle swarm optimization for significant wave height prediction","authors":"Can Yang ,&nbsp;Qingchen Kong ,&nbsp;Zuohang Su ,&nbsp;Hailong Chen ,&nbsp;Lars Johanning","doi":"10.1016/j.ocemod.2025.102511","DOIUrl":"10.1016/j.ocemod.2025.102511","url":null,"abstract":"<div><div>Short-term prediction of significant wave height (SWH) has crucial impacts on operation safety of offshore structures and marine navigations. However, conventional intelligent models have limitations in predicting non-linear situations. This paper introduces a hybrid algorithm combining chaos particle swarm optimization (CPSO) with a support vector regression (SVR) model to enhance the generalization and nonlinear handling capabilities for SWH prediction. Additionally, Principal Component Analysis (PCA) is incorporated to reduce information redundancy. To validate the proposed model's predictive performance, several alternatives are tested, including the single SVR model, PCA-SVR, and PCA-GA (Genetic Algorithm)-SVR models. Additionally, the PCA-GWO (Grey Wolf Optimizer)-SVR and PCA-CPSO-SVR models are compared to assess the effects of GWO and CPSO techniques. Significant improvements were observed when comparing CPSO-SVR with other algorithms. Prediction efficiency was evaluated using mean absolute error (MAE), root mean square error (RMSE), and the correlation coefficient (R). Across different test set lengths, the PCA-CPSO-SVR model reduced RMSE by 54.12 % to 74.88 % compared to the benchmark. These results demonstrate the hybrid PCA-CPSO-SVR model's strong generalization ability and superior predictive capacity for non-stationary waves.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"195 ","pages":"Article 102511"},"PeriodicalIF":3.1,"publicationDate":"2025-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143453827","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}

{"title":"Correcting physics-based global tide and storm water level forecasts with the temporal fusion transformer","authors":"A.R. Cerrone ,&nbsp;L.G. Westerink ,&nbsp;G. Ling ,&nbsp;C.P. Blakely ,&nbsp;D. Wirasaet ,&nbsp;C. Dawson ,&nbsp;J.J. Westerink","doi":"10.1016/j.ocemod.2025.102509","DOIUrl":"10.1016/j.ocemod.2025.102509","url":null,"abstract":"<div><div>Global and coastal ocean surface water elevation prediction skill has advanced considerably with improved algorithms, more refined discretizations, and high-performance parallel computing. Model skill is tied to mesh resolution, the accuracy of specified bathymetry/topography, dissipation parameterizations, air-sea drag formulations, and the fidelity of forcing functions. Wind forcing skill can be particularly prone to errors, especially at the land-ocean interface. The resulting biases and errors can be addressed holistically with a machine-learning (ML) approach. Herein, we weakly couple the Temporal Fusion Transformer to the National Oceanic and Atmospheric Administration’s (NOAA) Storm and Tide Operational Forecast System (STOFS-2D-Global) to improve its forecasting skill throughout a 7-day horizon. We demonstrate the transformer’s ability to enrich the hydrodynamic model’s output at 228 observed water level stations operated by NOAA’s National Ocean Service. We conclude that the transformer is a rapid way to correct STOFS-2D-Global forecasted water levels provided that sufficient covariates are supplied. For stations in wind-dominant areas, we demonstrate that including past and future wind-speed covariates makes for a more skillful forecast. In general, while the transformer renders consistent corrections at both tidally and wind-dominant stations, it does so most aggressively at tidally-dominant stations. We show notable improvements in Alaska and the Atlantic and Pacific seaboards of the United States. We evaluate several transformers instantiated with different hyperparameters, covariates, and training data to provide guidance on how to enhance performance.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"195 ","pages":"Article 102509"},"PeriodicalIF":3.1,"publicationDate":"2025-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143428777","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}