Pub Date : 2025-02-11DOI: 10.1016/j.ocemod.2025.102509
A.R. Cerrone , L.G. Westerink , G. Ling , C.P. Blakely , D. Wirasaet , C. Dawson , J.J. Westerink
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.
{"title":"Correcting physics-based global tide and storm water level forecasts with the temporal fusion transformer","authors":"A.R. Cerrone , L.G. Westerink , G. Ling , C.P. Blakely , D. Wirasaet , C. Dawson , 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}
Pub Date : 2025-02-10DOI: 10.1016/j.ocemod.2025.102512
Y. Joseph Zhang , Joshua Anderson , Chin H. Wu , Dmitry Beletsky , Yuli Liu , Wei Huang , Eric J. Anderson , Saeed Moghimi , Edward Myers
In this paper, for the first time, all five Great Lakes are simulated using a 3D baroclinic model using a single, seamless unstructured mesh without nesting, including adjacent flood plains and watershed inflows to better connect the hydrodynamic model to the hydrologic model. The hydraulic controls at Sault St Marie and Niagara Falls are simulated using an internal flow boundary approach with the observed flow. The model is shown to exhibit good skills for total water level (TWL) and temperature, with RMSE of 9.5 cm for TWL and ∼1.6 °C for surface temperature and temperature profiles from a 60–day simulation. Sensitivity results reveal the importance of hydrologic forcing even for this short-term simulation. Results from a 210-day simulation indicate that the model is capable of capturing major lake-wide circulation patterns discussed in previous studies and providing further details in those patterns. The new model can potentially serve as a base to unify Great Lakes modeling while simultaneously providing flexibility for site specific studies in any areas of interest.
{"title":"Cross-scale prediction for the Laurentian Great Lakes","authors":"Y. Joseph Zhang , Joshua Anderson , Chin H. Wu , Dmitry Beletsky , Yuli Liu , Wei Huang , Eric J. Anderson , Saeed Moghimi , Edward Myers","doi":"10.1016/j.ocemod.2025.102512","DOIUrl":"10.1016/j.ocemod.2025.102512","url":null,"abstract":"<div><div>In this paper, for the first time, all five Great Lakes are simulated using a 3D baroclinic model using a single, seamless unstructured mesh without nesting, including adjacent flood plains and watershed inflows to better connect the hydrodynamic model to the hydrologic model. The hydraulic controls at Sault St Marie and Niagara Falls are simulated using an internal flow boundary approach with the observed flow. The model is shown to exhibit good skills for total water level (TWL) and temperature, with RMSE of 9.5 cm for TWL and ∼1.6 °C for surface temperature and temperature profiles from a 60–day simulation. Sensitivity results reveal the importance of hydrologic forcing even for this short-term simulation. Results from a 210-day simulation indicate that the model is capable of capturing major lake-wide circulation patterns discussed in previous studies and providing further details in those patterns. The new model can potentially serve as a base to unify Great Lakes modeling while simultaneously providing flexibility for site specific studies in any areas of interest.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"194 ","pages":"Article 102512"},"PeriodicalIF":3.1,"publicationDate":"2025-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143402694","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 : 2025-02-06DOI: 10.1016/j.ocemod.2025.102510
Michael McGlade , Nieves G. Valiente , Jennifer Brown , Christopher Stokes , Timothy Poate
Predicting coastal wave overtopping is a significant challenge, exacerbated by climate change, increasing the frequency of severe flooding and rising sea levels. Digital twin technologies, which utilise artificial intelligence to mimic coastal processes and dynamics, may offer new opportunities to predict coastal wave overtopping and flooding reliably and computationally efficiently. This study investigates the effectiveness of training various artificial intelligence models using wave buoy, meteorological, and recorded coastal wave overtopping observations to predict the occurrence and frequency of overtopping at 10-minute intervals. These models have the potential for future large-scale global applications in estimating wave overtopping and flood forecasting, particularly in response to climate warming. The model types selected include machine-learning random forests, extreme gradient boosting, support vector machines, and deep-learning neural networks. These models were trained and tested using recorded observational overtopping events, to estimate wave overtopping and flood forecasting in Dawlish and Penzance (Southwest England). The random forests performed exceptionally well by accurately and precisely estimating coastal wave overtopping and non-overtopping 97 % of the time within both locations, outperforming the other models. Moreover, the random forest model outperforms existing process-based and EurOtop-based models. This research has profound implications for increasing preparedness and resilience to future coastal wave overtopping and flooding events by using these random forest models to predict overtopping and flood forecasting on wider global and climate scales. These trained random forests are significantly less computationally demanding than existing process-based models and can incorporate the important effect of wind on overtopping, which was neglected in existing empirical approaches.
{"title":"Investigating appropriate artificial intelligence approaches to reliably predict coastal wave overtopping and identify process contributions","authors":"Michael McGlade , Nieves G. Valiente , Jennifer Brown , Christopher Stokes , Timothy Poate","doi":"10.1016/j.ocemod.2025.102510","DOIUrl":"10.1016/j.ocemod.2025.102510","url":null,"abstract":"<div><div>Predicting coastal wave overtopping is a significant challenge, exacerbated by climate change, increasing the frequency of severe flooding and rising sea levels. Digital twin technologies, which utilise artificial intelligence to mimic coastal processes and dynamics, may offer new opportunities to predict coastal wave overtopping and flooding reliably and computationally efficiently. This study investigates the effectiveness of training various artificial intelligence models using wave buoy, meteorological, and recorded coastal wave overtopping observations to predict the occurrence and frequency of overtopping at 10-minute intervals. These models have the potential for future large-scale global applications in estimating wave overtopping and flood forecasting, particularly in response to climate warming. The model types selected include machine-learning random forests, extreme gradient boosting, support vector machines, and deep-learning neural networks. These models were trained and tested using recorded observational overtopping events, to estimate wave overtopping and flood forecasting in Dawlish and Penzance (Southwest England). The random forests performed exceptionally well by accurately and precisely estimating coastal wave overtopping and non-overtopping 97 % of the time within both locations, outperforming the other models. Moreover, the random forest model outperforms existing process-based and EurOtop-based models. This research has profound implications for increasing preparedness and resilience to future coastal wave overtopping and flooding events by using these random forest models to predict overtopping and flood forecasting on wider global and climate scales. These trained random forests are significantly less computationally demanding than existing process-based models and can incorporate the important effect of wind on overtopping, which was neglected in existing empirical approaches.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"194 ","pages":"Article 102510"},"PeriodicalIF":3.1,"publicationDate":"2025-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143394419","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}
Pub Date : 2025-02-03DOI: 10.1016/j.ocemod.2025.102506
Hongyuan Zhang , Dongliang Shen , Len Pietrafesa , Paul Gayes , Shaowu Bao
Coastal flooding during hurricanes is a complex phenomenon involving the interaction of multiple drivers operating across different spatial scales, such as storm surge, rainfall, river discharge, and tides. Accurately assessing and predicting compound flooding requires considering the cross-scale nature of these processes and their interdependencies. This study investigates the compound effects and cross-scale interactions of flood drivers during Hurricane Matthew (2016) along the South Carolina coast using a coupled hydrology-hydrodynamic model. The model domain encompasses the land-ocean continuum, from rivers to the coastal ocean, allowing for the examination of compound flooding across the entire system. Controlled numerical experiments are conducted to quantify the individual, combined, and compound impacts of flood drivers across scales by selectively including or excluding riverine, storm surge, and tidal forcings. The coupled modeling approach reveals distinct zones of positive and negative compound effects, depending on the alignment of coastal water levels with river flood timing. River-surge interactions alter flooding, causing increases upstream and decreases downstream compared to isolated effects. Tide-surge and tide-river interactions induce oscillatory compound effects. The study demonstrates that the compound effect significantly influences hurricane coastal flooding beyond the linear superposition of flooding caused by individual drivers. The cross-scale modeling framework and analysis approach presented here can inform multi-hazard analysis, coastal flood risk management, and future studies of complex, multi-scale hydrologic systems.
{"title":"Quantify the compound effects caused by the interactions between inland river system and coastal processes in hurricane coastal flooding through controlled hydrodynamic modeling experiments","authors":"Hongyuan Zhang , Dongliang Shen , Len Pietrafesa , Paul Gayes , Shaowu Bao","doi":"10.1016/j.ocemod.2025.102506","DOIUrl":"10.1016/j.ocemod.2025.102506","url":null,"abstract":"<div><div>Coastal flooding during hurricanes is a complex phenomenon involving the interaction of multiple drivers operating across different spatial scales, such as storm surge, rainfall, river discharge, and tides. Accurately assessing and predicting compound flooding requires considering the cross-scale nature of these processes and their interdependencies. This study investigates the compound effects and cross-scale interactions of flood drivers during Hurricane Matthew (2016) along the South Carolina coast using a coupled hydrology-hydrodynamic model. The model domain encompasses the land-ocean continuum, from rivers to the coastal ocean, allowing for the examination of compound flooding across the entire system. Controlled numerical experiments are conducted to quantify the individual, combined, and compound impacts of flood drivers across scales by selectively including or excluding riverine, storm surge, and tidal forcings. The coupled modeling approach reveals distinct zones of positive and negative compound effects, depending on the alignment of coastal water levels with river flood timing. River-surge interactions alter flooding, causing increases upstream and decreases downstream compared to isolated effects. Tide-surge and tide-river interactions induce oscillatory compound effects. The study demonstrates that the compound effect significantly influences hurricane coastal flooding beyond the linear superposition of flooding caused by individual drivers. The cross-scale modeling framework and analysis approach presented here can inform multi-hazard analysis, coastal flood risk management, and future studies of complex, multi-scale hydrologic systems.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"194 ","pages":"Article 102506"},"PeriodicalIF":3.1,"publicationDate":"2025-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143387397","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 : 2025-02-03DOI: 10.1016/j.ocemod.2025.102505
Øyvind Breivik , Bente Moerman , Knut-Frode Dagestad , Tor Nordam , Gaute Hope , Lars Robert Hole , Arthur A. Allen , Lawrence D. Stone
Backtracking the drift of particles and substances is central to a range of studies in oceanography as well as in law enforcement, search and rescue and the mapping and investigation of marine pollution. Here we demonstrate how a Lagrangian particle model can be used in a forward mode with a Bayesian prior estimate on the release location of the object of interest. We show that for well-behaved drifters, forward and backward (reverse modelling) yield similar results over short periods, if the currents are only weakly divergent. However, for drifters undergoing discontinuous state changes, such as stranding, or objects abruptly and irreversibly changing their drift properties, or for buoyant drifters in strongly convergent flows, backward drift can yield wrongful search areas. We demonstrate this for a case where a liferaft is assigned a wind-speed dependent probability of capsizing, leading to an instantaneous change in drift properties. We also demonstrate the forward and backward methods for a drifter release experiment in the Agulhas current where we also assess the challenges of biases in the current fields. Finally, a method for incorporating multiple observations of debris with a forward model in the Bayesian posterior estimate of the initial location is outlined.
{"title":"The Bayesian backtracking problem in oceanic drift modelling","authors":"Øyvind Breivik , Bente Moerman , Knut-Frode Dagestad , Tor Nordam , Gaute Hope , Lars Robert Hole , Arthur A. Allen , Lawrence D. Stone","doi":"10.1016/j.ocemod.2025.102505","DOIUrl":"10.1016/j.ocemod.2025.102505","url":null,"abstract":"<div><div>Backtracking the drift of particles and substances is central to a range of studies in oceanography as well as in law enforcement, search and rescue and the mapping and investigation of marine pollution. Here we demonstrate how a Lagrangian particle model can be used in a forward mode with a Bayesian prior estimate on the release location of the object of interest. We show that for well-behaved drifters, forward and backward (reverse modelling) yield similar results over short periods, if the currents are only weakly divergent. However, for drifters undergoing discontinuous state changes, such as stranding, or objects abruptly and irreversibly changing their drift properties, or for buoyant drifters in strongly convergent flows, backward drift can yield wrongful search areas. We demonstrate this for a case where a liferaft is assigned a wind-speed dependent probability of capsizing, leading to an instantaneous change in drift properties. We also demonstrate the forward and backward methods for a drifter release experiment in the Agulhas current where we also assess the challenges of biases in the current fields. Finally, a method for incorporating multiple observations of debris with a forward model in the Bayesian posterior estimate of the initial location is outlined.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"194 ","pages":"Article 102505"},"PeriodicalIF":3.1,"publicationDate":"2025-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143147364","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 : 2025-02-03DOI: 10.1016/j.ocemod.2025.102508
Yueqi Zhao, Zhongjie He, Xiachuan Fu, Lihua Hou
To improve the performance of the multiscale assimilation algorithm, we propose a non-uniform downscaling (NUD) data assimilation algorithm in a variational framework based on the relationship between the space structure of the scale decomposition and the space distribution characteristics. The algorithm differs from the traditional uniform downscaling (UD) algorithm in that it enables the distribution of space grid points to be sparse in large scale regions and dense in small scale regions. The non-uniform scale decomposition can better control the propagation range of the observation information. Experiments show that the NUD can reduce the background error by about 5 % relative to the UD. The spatial distribution characteristics of the analysis field obtained by the NUD are also more similar to the true field. In addition, the forecast results show that the non-uniform scale decomposition assimilation algorithm with model integration can produce a stable positive impact and effectively improve the forecast capability for small and medium scale phenomena.
{"title":"Non-uniform downscaling data assimilation algorithm in variational framework","authors":"Yueqi Zhao, Zhongjie He, Xiachuan Fu, Lihua Hou","doi":"10.1016/j.ocemod.2025.102508","DOIUrl":"10.1016/j.ocemod.2025.102508","url":null,"abstract":"<div><div>To improve the performance of the multiscale assimilation algorithm, we propose a non-uniform downscaling (NUD) data assimilation algorithm in a variational framework based on the relationship between the space structure of the scale decomposition and the space distribution characteristics. The algorithm differs from the traditional uniform downscaling (UD) algorithm in that it enables the distribution of space grid points to be sparse in large scale regions and dense in small scale regions. The non-uniform scale decomposition can better control the propagation range of the observation information. Experiments show that the NUD can reduce the background error by about 5 % relative to the UD. The spatial distribution characteristics of the analysis field obtained by the NUD are also more similar to the true field. In addition, the forecast results show that the non-uniform scale decomposition assimilation algorithm with model integration can produce a stable positive impact and effectively improve the forecast capability for small and medium scale phenomena.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"194 ","pages":"Article 102508"},"PeriodicalIF":3.1,"publicationDate":"2025-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143232644","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 : 2025-02-03DOI: 10.1016/j.ocemod.2025.102498
Bijit K. Kalita, P.N. Vinayachandran
The response of a river plume to a tropical cyclone (TC) is a lesser-known aspect of tropical oceanography. We have investigated the interaction between a river plume and a category 5 TC, Phailin (8-14 October 2013), in the Bay of Bengal using high-resolution simulations of ROMS (Regional Ocean Modelling System). Two distinct experiments were performed: one including the river runoff into the ocean (Riv) and the other excluding it (NoRiv). From the east coast of India, a river plume advected offshore along the southern arm of a cyclonic eddy. The cyclone destroyed the river plume and scattered the low salinity water over a large region. The presence of a river plume made the pre-storm north bay a fresh, warm, and stably stratified basin, with a shallow mixed layer (ML) and thick barrier layer (BL). Upon the passage of Phailin, the stratification weakened, ML deepened, and BL thickness decreased. Maximum temperature drop and salinity rise were seen along the southern flank of the cyclonic eddy. The terms of salinity and temperature equations show that vertical mixing and advection caused these responses, aided by the cumulative effect of upwelling induced by the cyclonic eddy and storm, and the advection of redistributing river water.
{"title":"Bay of Bengal river plume response to a tropical cyclone in high-resolution numerical simulations","authors":"Bijit K. Kalita, P.N. Vinayachandran","doi":"10.1016/j.ocemod.2025.102498","DOIUrl":"10.1016/j.ocemod.2025.102498","url":null,"abstract":"<div><div>The response of a river plume to a tropical cyclone (TC) is a lesser-known aspect of tropical oceanography. We have investigated the interaction between a river plume and a category 5 TC, Phailin (8-14 October 2013), in the Bay of Bengal using high-resolution simulations of ROMS (Regional Ocean Modelling System). Two distinct experiments were performed: one including the river runoff into the ocean (Riv) and the other excluding it (NoRiv). From the east coast of India, a river plume advected offshore along the southern arm of a cyclonic eddy. The cyclone destroyed the river plume and scattered the low salinity water over a large region. The presence of a river plume made the pre-storm north bay a fresh, warm, and stably stratified basin, with a shallow mixed layer (ML) and thick barrier layer (BL). Upon the passage of Phailin, the stratification weakened, ML deepened, and BL thickness decreased. Maximum temperature drop and salinity rise were seen along the southern flank of the cyclonic eddy. The terms of salinity and temperature equations show that vertical mixing and advection caused these responses, aided by the cumulative effect of upwelling induced by the cyclonic eddy and storm, and the advection of redistributing river water.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"194 ","pages":"Article 102498"},"PeriodicalIF":3.1,"publicationDate":"2025-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143350432","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 : 2025-02-03DOI: 10.1016/j.ocemod.2025.102499
Haidong Pan , Dingqi Wang , Fei Teng , Xiaoqing Xu , Tengfei Xu , Zexun Wei
Accuracy assessments of global tide models are essential for practical applications such as navigation and tidal correction in satellite altimetry. Traditionally, model accuracy assessment is performed by comparing global tide models with point-based tide gauges which are spatially limited in terms of number and location. To address the issue of the scarcity of tide gauges in the sea areas of interest, previous studies have introduced tide-favourable Topex/Poseidon (T/P)-Jason series as supplements in the course of model evaluation. However, the overlap of T/P-Jason series in model construction and evaluation may destroy the reliability of their evaluation results. In this study, we evaluate the accuracy of five popular global tide models (EOT20, FES2014, TPXO9, HAMTIDE12, and DTU16) in the Bohai Sea via the combination of four tide gauges and ∼8-year Geosat Follow-On (GFO) satellite altimeter records. The wide coverage of GFO data and their independence from analyzed global tide models guarantee the reliability of our findings. The results from GFO data and tide gauges are highly consistent, both revealing the superiority of FES2014 and the poor performances of TPXO9 and HAMTIDE12 in the Bohai Sea. The performances of FES-referenced models (EOT20 and DTU16) are generally good but slightly worse than that of FES2014. Of note, tidal amplitudes of eight major constituents derived from TPXO9 are abnormally small in the Bohai Sea. As water depths decrease, the errors of all five global tide models elevate notably, indicting the challenges in modelling coastal tides which may necessitate further exploration and improvements.
{"title":"Accuracy evaluation of global tidal models in the Bohai Sea via the combination of tide gauges and GFO satellite altimeters","authors":"Haidong Pan , Dingqi Wang , Fei Teng , Xiaoqing Xu , Tengfei Xu , Zexun Wei","doi":"10.1016/j.ocemod.2025.102499","DOIUrl":"10.1016/j.ocemod.2025.102499","url":null,"abstract":"<div><div>Accuracy assessments of global tide models are essential for practical applications such as navigation and tidal correction in satellite altimetry. Traditionally, model accuracy assessment is performed by comparing global tide models with point-based tide gauges which are spatially limited in terms of number and location. To address the issue of the scarcity of tide gauges in the sea areas of interest, previous studies have introduced tide-favourable Topex/Poseidon (T/P)-Jason series as supplements in the course of model evaluation. However, the overlap of T/P-Jason series in model construction and evaluation may destroy the reliability of their evaluation results. In this study, we evaluate the accuracy of five popular global tide models (EOT20, FES2014, TPXO9, HAMTIDE12, and DTU16) in the Bohai Sea via the combination of four tide gauges and ∼8-year Geosat Follow-On (GFO) satellite altimeter records. The wide coverage of GFO data and their independence from analyzed global tide models guarantee the reliability of our findings. The results from GFO data and tide gauges are highly consistent, both revealing the superiority of FES2014 and the poor performances of TPXO9 and HAMTIDE12 in the Bohai Sea. The performances of FES-referenced models (EOT20 and DTU16) are generally good but slightly worse than that of FES2014. Of note, tidal amplitudes of eight major constituents derived from TPXO9 are abnormally small in the Bohai Sea. As water depths decrease, the errors of all five global tide models elevate notably, indicting the challenges in modelling coastal tides which may necessitate further exploration and improvements.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"194 ","pages":"Article 102499"},"PeriodicalIF":3.1,"publicationDate":"2025-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143147365","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 : 2025-01-28DOI: 10.1016/j.ocemod.2025.102500
Shuangying Du , Rong-Hua Zhang
In traditional ocean-atmosphere coupled modeling for El Niño-Southern Oscillation (ENSO) studies, statistical methods are typically used to represent the instantaneous linear relationship between monthly-averaged anomalies of sea surface temperature (SST) and wind stress (τ). Recently, deep learning (DL) techniques have presented promising prospects for ENSO modeling, and the integration of neural networks (NNs) with dynamical models is an active research area. This study incorporates the Residual convolution blocks and a Convolutional Block Attention Module (the Res-CBAM block) into the original UNet configuration to build a new RCUNet-based model, denoted as , which uses SST anomalies (SSTAs) during multi-day time intervals (TIs) to derive daily responses. Sensitivity tests to TIs are performed to illustrate how daily responses are dependent on the way multi-day SST forcings are used; the comparisons with different TIs show that when taking TI=3 days, the model can more precisely represent the relationship between SSTAs and anomalies. Next, daily anomalies obtained from the model are used to force an intermediate ocean model (IOM) in the ocean-only experiments, displaying coherent phase transitions and spatiotemporal evolutions of oceanic and atmospheric anomalies during typical ENSO events, which highlights the advantages of using the DL-based atmospheric model with multi-day SST time sequence information incorporated for ocean modeling. Furthermore, a new intermediate coupled model (ICM) is formed, named the ICM-RCUNet, in which the original atmospheric component represented by singular value decomposition (SVD) analyses is replaced by the model that is used as an atmospheric component, and a daily coupling is conducted with multi-day SST forcings. The ICM-RCUNet simulations exhibit interannual oscillations of atmospheric and oceanic states in the tropical Pacific, demonstrating the applicability of integrating physics-based dynamical ocean models with atmospheric NNs in ENSO-related studies. Further implications for ocean and coupled modelings using NNs are discussed.
{"title":"An RCUNet-based sea surface wind stress model with multi-day time sequence information incorporated and its applications to ENSO modeling","authors":"Shuangying Du , Rong-Hua Zhang","doi":"10.1016/j.ocemod.2025.102500","DOIUrl":"10.1016/j.ocemod.2025.102500","url":null,"abstract":"<div><div>In traditional ocean-atmosphere coupled modeling for El Niño-Southern Oscillation (ENSO) studies, statistical methods are typically used to represent the instantaneous linear relationship between monthly-averaged anomalies of sea surface temperature (SST) and wind stress (τ). Recently, deep learning (DL) techniques have presented promising prospects for ENSO modeling, and the integration of neural networks (NNs) with dynamical models is an active research area. This study incorporates the Residual convolution blocks and a Convolutional Block Attention Module (the Res-CBAM block) into the original UNet configuration to build a new RCUNet-based <span><math><mi>τ</mi></math></span> model, denoted as <span><math><msub><mi>τ</mi><mtext>RCUNet</mtext></msub></math></span>, which uses SST anomalies (SSTAs) during multi-day time intervals (TIs) to derive daily <span><math><mi>τ</mi></math></span> responses. Sensitivity tests to TIs are performed to illustrate how daily <span><math><mi>τ</mi></math></span> responses are dependent on the way multi-day SST forcings are used; the comparisons with different TIs show that when taking TI=3 days, the <span><math><msub><mi>τ</mi><mtext>RCUNet</mtext></msub></math></span> model can more precisely represent the relationship between SSTAs and <span><math><mi>τ</mi></math></span> anomalies. Next, daily <span><math><mi>τ</mi></math></span> anomalies obtained from the <span><math><msub><mi>τ</mi><mtext>RCUNet</mtext></msub></math></span> model are used to force an intermediate ocean model (IOM) in the ocean-only experiments, displaying coherent phase transitions and spatiotemporal evolutions of oceanic and atmospheric anomalies during typical ENSO events, which highlights the advantages of using the DL-based atmospheric <span><math><mi>τ</mi></math></span> model with multi-day SST time sequence information incorporated for ocean modeling. Furthermore, a new intermediate coupled model (ICM) is formed, named the ICM-RCUNet, in which the original atmospheric component represented by singular value decomposition (SVD) analyses is replaced by the <span><math><msub><mi>τ</mi><mtext>RCUNet</mtext></msub></math></span> model that is used as an atmospheric component, and a daily coupling is conducted with multi-day SST forcings. The ICM-RCUNet simulations exhibit interannual oscillations of atmospheric and oceanic states in the tropical Pacific, demonstrating the applicability of integrating physics-based dynamical ocean models with atmospheric NNs in ENSO-related studies. Further implications for ocean and coupled modelings using NNs are discussed.</div></div>","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"194 ","pages":"Article 102500"},"PeriodicalIF":3.1,"publicationDate":"2025-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143232643","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 : 2025-01-27DOI: 10.1016/j.ocemod.2025.102503
Oleksandr Nesterov
<div><div>The Arabian / Persian Gulf, hereinafter referred to as the Gulf, is one of the major sources of freshwater through desalination in Kuwait, Bahrain, Qatar, the United Arab Emirates, the eastern Kingdom of Saudi Arabia (KSA), as well as in the coastal regions of the Iranian provinces Bushehr and Hormozgan, and the Oman governorate Musandam. Over the past four decades, the seawater desalination capacities in these states have been steadily growing. By 2023 the total production capacity has reached 25.6 Mm<sup>3</sup>/day, more than double that in 2006 (∼12.1 Mm<sup>3</sup>/day), and it is expected to add at least 6.2 Mm<sup>3</sup>/day by 2030. However, not only the Gulf serves as a source of freshwater for domestic, agricultural and industrial use, but also as a sink for the reject brine. Being a semi-enclosed water body characterized by relatively high evaporation rates of 1.5 – 2 m/year on average, small and reportedly declining river inflows, and negligible precipitations, the Gulf may eventually become impacted by anthropogenic changes in salinities, raising concerns about acceptable freshwater production rates and the overall resilience of its ecosystem. In this study, to estimate a combined impact of the desalination facilities on salinities on global and local scales, a 3D high-resolution free-surface baroclinic model of the Gulf's hydrodynamics was established using the open-source unstructured-grid Semi-implicit Cross-scale Hydroscience Integrated System Model. The horizontal resolution of the numerical mesh varied from approximately 4 km in the Sea of Oman to higher than a hundred meters in some coastal areas of the Gulf, which is notably higher than in previous studies, with the vertical resolution of up to 64 local-sigma-coordinate levels. The desalination capacities of 738 coastal plants were considered based on the 35<sup>th</sup> Worldwide Desalting Inventory of 2023. Two simulations were carried out over 3 years to assess their impact: one without the effect of the plants (baseline), and the other one with the effect of the plants. The baseline model was validated against field measurements. The volume, near-surface and near-bed averaged salinity impacts in the whole Gulf were found converging to 0.03, 0.04, and 0.05 g/l, respectively. Higher salinity increments were modeled closer to the plants, in some cases at considerable distances. For example, the impact on the median salinities at the seabed was estimated to exceed 0.1 g/l as far as 50 km away from Jubail in KSA, and Al Abu Fontas and Umm Al Houl in Qatar. Even higher increments were modeled in poorly flushed zones, such as ∼0.3 g/l in the whole Gulf of Salwa, and in the range of 0.1 – 0.5 g/l in most of the Abu Dhabi canals. Although such salinity impacts are unlikely to pose environmental threats in the present, the effect of desalination may become more profound in the future, especially in view that salinities in the Gulf have remained quasi-equilibrium over the
{"title":"An assessment of seawater desalination impact on salinities in the Arabian / Persian Gulf using a 3D circulation model","authors":"Oleksandr Nesterov","doi":"10.1016/j.ocemod.2025.102503","DOIUrl":"10.1016/j.ocemod.2025.102503","url":null,"abstract":"<div><div>The Arabian / Persian Gulf, hereinafter referred to as the Gulf, is one of the major sources of freshwater through desalination in Kuwait, Bahrain, Qatar, the United Arab Emirates, the eastern Kingdom of Saudi Arabia (KSA), as well as in the coastal regions of the Iranian provinces Bushehr and Hormozgan, and the Oman governorate Musandam. Over the past four decades, the seawater desalination capacities in these states have been steadily growing. By 2023 the total production capacity has reached 25.6 Mm<sup>3</sup>/day, more than double that in 2006 (∼12.1 Mm<sup>3</sup>/day), and it is expected to add at least 6.2 Mm<sup>3</sup>/day by 2030. However, not only the Gulf serves as a source of freshwater for domestic, agricultural and industrial use, but also as a sink for the reject brine. Being a semi-enclosed water body characterized by relatively high evaporation rates of 1.5 – 2 m/year on average, small and reportedly declining river inflows, and negligible precipitations, the Gulf may eventually become impacted by anthropogenic changes in salinities, raising concerns about acceptable freshwater production rates and the overall resilience of its ecosystem. In this study, to estimate a combined impact of the desalination facilities on salinities on global and local scales, a 3D high-resolution free-surface baroclinic model of the Gulf's hydrodynamics was established using the open-source unstructured-grid Semi-implicit Cross-scale Hydroscience Integrated System Model. The horizontal resolution of the numerical mesh varied from approximately 4 km in the Sea of Oman to higher than a hundred meters in some coastal areas of the Gulf, which is notably higher than in previous studies, with the vertical resolution of up to 64 local-sigma-coordinate levels. The desalination capacities of 738 coastal plants were considered based on the 35<sup>th</sup> Worldwide Desalting Inventory of 2023. Two simulations were carried out over 3 years to assess their impact: one without the effect of the plants (baseline), and the other one with the effect of the plants. The baseline model was validated against field measurements. The volume, near-surface and near-bed averaged salinity impacts in the whole Gulf were found converging to 0.03, 0.04, and 0.05 g/l, respectively. Higher salinity increments were modeled closer to the plants, in some cases at considerable distances. For example, the impact on the median salinities at the seabed was estimated to exceed 0.1 g/l as far as 50 km away from Jubail in KSA, and Al Abu Fontas and Umm Al Houl in Qatar. Even higher increments were modeled in poorly flushed zones, such as ∼0.3 g/l in the whole Gulf of Salwa, and in the range of 0.1 – 0.5 g/l in most of the Abu Dhabi canals. Although such salinity impacts are unlikely to pose environmental threats in the present, the effect of desalination may become more profound in the future, especially in view that salinities in the Gulf have remained quasi-equilibrium over the","PeriodicalId":19457,"journal":{"name":"Ocean Modelling","volume":"194 ","pages":"Article 102503"},"PeriodicalIF":3.1,"publicationDate":"2025-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143372939","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