Sea Surface Temperature Prediction Research Articles

How do model biases affect large-scale teleconnections that control Southwest US precipitation? Part II: seasonal models

Abstract We explore the skill in predicting Southwest United States (SWUS) October to March precipitation and associated large-scale teleconnections in an ensemble of hindcasts from seasonal prediction systems. We identify key model biases that degrade the models’ capability to predict SWUS precipitation. The subtropical jet in the Pacific sector is generally too zonal and elongated. This is reflected in the models’ North Pacific ENSO teleconnections that are generally too weak with exaggerated northwest-southeast tilt, compared to observations. Also, the models are too dependent on tropical, El Niño-like, wave train anomalies for producing high seasonal SWUS precipitation, when in observations there is a larger influence of zonal Rossby wave trains such as the one observed in 2016/17. Overall, this is consistent with biases in the basic flow inducing errors in the propagation of zonal wave trains in the North Pacific, which affects SWUS precipitation downstream. Although higher skill may be gained from reducing mean flow biases in the models, a case study of the 2016/17 winter illustrates the great challenge behind skillful seasonal prediction of SWUS precipitation. Unsurprisingly, the almost record-breaking precipitation observed that year in the absence of ENSO is not predicted in the hindcasts, and model perturbation experiments suggest that even a perfect prediction of tropical sea surface temperature and tropical atmospheric variability would not have sufficed to produce a reasonable seasonal precipitation prediction. On a more positive note, our perturbation experiments suggest a potential role for Arctic variability that supports findings from prior studies and suggests re-examining high-latitude drivers of SWUS precipitation.

Sea Surface Temperature Prediction Using ConvLSTM-Based Model with Deformable Attention

Sea surface temperature (SST) prediction has received increasing attention in recent years due to its paramount importance in the various fields of oceanography. Existing studies have shown that neural networks are particularly effective in making accurate SST predictions by efficiently capturing spatiotemporal dependencies in SST data. Among various models, the ConvLSTM framework is notably prominent. This model skillfully combines convolutional neural networks (CNNs) with recurrent neural networks (RNNs), enabling it to simultaneously capture spatiotemporal dependencies within a single computational framework. To overcome the limitation that CNNs primarily capture local spatial information, in this paper we propose a novel model named DatLSTM that integrates a deformable attention transformer (DAT) module into the ConvLSTM framework, thereby enhancing its ability to process more complex spatial relationships effectively. Specifically, the DAT module adaptively focuses on salient features in space, while ConvLSTM further captures the temporal dependencies of spatial correlations in the SST data. In this way, DatLSTM can adaptively capture complex spatiotemporal dependencies between the preceding and current states within ConvLSTM. To evaluate the performance of the DatLSTM model, we conducted short-term SST forecasts in the Bohai Sea region with forecast lead times ranging from 1 to 10 days and compared its efficacy against several benchmark models, including ConvLSTM, PredRNN, TCTN, and SwinLSTM. Our experimental results show that the proposed model outperforms all of these models in terms of multiple evaluation metrics short-term SST prediction. The proposed model offers a new predictive learning method for improving the accuracy of spatiotemporal predictions in various domains, including meteorology, oceanography, and climate science.

Building an emulator for sea surface temperature

We are interested in the problem of predicting sea surface temperature with machine learning methods using only the relevant climate factors. We tried to resolve the cost inefficiency of the temperature predicting model using over a million of data[1] that is observed by about 6300[2] NASA weather stations, a typical model that prominent agencies like NASA uses. First of all, we minimized the required computing power by using only velocity and pressure variables. Moreover, we minimized the usage of data by using only 14GB of NASA-GISS data and supplemented the minimized data by comparing four different machine learning models to predict global sea surface temperature. In addition, beyond predicting the global sea surface temperature, we also wanted to delve into predicting each region’s temperature. Thus, we gridded the earth into 64,800 regions, based on latitudes and longitudes. However, the accuracy drastically decreased since the data was also divided into 64,800 parts. Therefore, we applied the Gaussian filter to our predicted outcomes, and it made a breakthrough in accuracy. This will have a huge impact on temperature prediction because it will astoundingly decrease the expense and the amount of data needed for predicting temperature.

Numerical Weather Prediction of Sea Surface Temperature in South China Sea Using Attention-Based Context Fusion Network

Numerical weather prediction of sea surface temperature (SST) is crucial for regional operational forecasts. Deep learning offers an alternative approach to traditional numerical general circulation models for numerical weather prediction. In our previous work, we developed a sophisticated deep learning model known as the Attention-based Context Fusion Network (ACFN). This model integrates an attention mechanism with a convolutional neural network framework. In this study, we applied the ACFN model to the South China Sea to evaluate its performance in predicting SST. The results indicate that for a 1-day lead time, the ACFN model achieves a Mean Absolute Error of 0.215 °C and a coefficient of determination (R2) of 0.972. In addition, in situ buoy data were utilized to validate the forecast results. The Mean Absolute Error for forecasts using these data increased to 0.500 °C for a 1-day lead time, with a corresponding R2 of 0.590. Comparative analyses show that the ACFN model surpasses traditional models such as ConvLSTM and PredRNN in terms of accuracy and reliability.

A coordination attention residual U-Net model for enhanced short and mid-term sea surface temperature prediction

Sea surface temperature (SST) is crucial for studying global oceans and evaluating ecosystems. Accurately predicting short and mid-term daily SST has been a significant challenge in oceanography. Traditional deep learning methods can handle temporal data and spatial features but often struggle with long-range spatiotemporal dependencies. To address this, we propose a coordination attention residual U-Net(CResU-Net) model designed to better capture the dynamic spatiotemporal correlations of high-resolution SST. The model integrates coordinate attention mechanisms, multiple residual modules, and depthwise separable convolutions to enhance prediction capabilities. The spatiotemporal variations of SST across different areas of the South China Sea are complex, making accurate predictions challenging. Experiments across various regions of the South China Sea show the model’s effectiveness and robust generalization in predicting high-resolution daily SST. For a 10-day forecast period, the model achieves approximately 0.3 °C in RMSE, outperforming several advanced models.

Seasonal forecasting of the European North-West shelf seas: limits of winter and summer sea surface temperature predictability

The European North-West shelf seas (NWS) support economic interests and provide environmental services to adjacent countries. Expansion of offshore activities, such as renewable energy infrastructure, aquaculture, and growth of international shipping, will place increasingly complex demands on the marine environment over the coming decades. Skilful forecasting of NWS properties on seasonal timescales will help to effectively manage these activities. Here we quantify the skill of an operational large-ensemble ocean-atmosphere coupled global forecasting system (GloSea), as well as benchmark persistence forecasts, for predictions of NWS sea surface temperature (SST) at 2–4 months lead time in winter and summer. We identify sources of and limits to SST predictability, considering what additional skill may be available in the future. We find that GloSea NWS SST skill is generally high in winter and low in summer. GloSea outperforms simple persistence forecasts by adding information about atmospheric variability, but only to a modest extent as persistence of anomalies in the initial conditions contributes substantially to predictability. Where persistence is low – for example in seasonally stratified regions – GloSea forecasts show lower skill. GloSea skill can be degraded by model deficiencies in the relatively coarse global ocean component, which lacks dynamic tides and subsequently fails to robustly represent local circulation and mixing. However, “atmospheric mode matched” tests show potential for improving prediction skill of currently low performing regions if atmospheric circulation forecasts can be improved. This underlines the importance of coupled atmosphere-ocean model development for NWS seasonal forecasting applications.

Predictability and prediction skill of summertime East/Japan Sea surface temperature events

The East/Japan Sea (EJS), a marginal sea of the Northwestern Pacific, is one of the ocean regions showing the most rapid warming and greatest increases in ocean heatwaves over the last several decades. Predictability and skillful prediction of the summer season EJS variability are crucial, given the increasing severity of ocean temperature events impacting fisheries and reinforcing climate conditions like the East Asian rainy season, which in turn affects adjacent high-population density areas over East Asia. We use observations and the Geophysical Fluid Dynamics Laboratory (GFDL) Seamless System for Prediction and Earth System Research (SPEAR) seasonal forecast system to investigate the summertime EJS Sea Surface Temperature (SST) predictability and prediction skill. The observations and seasonal prediction system show that the summer season EJS SST can be closely linked to the previous winter air-sea coupling and predictable 8–9 months in advance. The SPEAR seasonal prediction system demonstrates skillful forecast of EJS SST events from summer to late fall, with added skill for long-lead forecasts initialized in winter. We find that winter large-scale atmospheric circulations linked to Barents Sea variability can induce persistent surface wind anomalies and corresponding northward Ekman heat transport over the East China Sea. The ocean advection anomalies that enter the EJS in prior seasons appear to play a role in developing anomalous SST during summer, along with instantaneous atmospheric forcing, as the source of long-lead predictability. Our findings provide potential applications of large-scale ocean-atmosphere interactions in understanding and predicting seasonal variability of East Asian marginal seas.

A Physical‐Informed Neural Network for Improving Air‐Sea Turbulent Heat Flux Parameterization

AbstractThe parameterizations of air‐sea turbulent heat flux are one of the major bottlenecks in atmosphere‐ocean coupled model development, which play a crucial role in sea surface temperature (SST) prediction. Recently, neural networks start to be applied for the development of parameterizations of interface turbulent heat flux. However, these new parameterizations are primairily developed for specific regions and have not been tested in real atmosphere‐ocean coupled models. In this study, we propose a new air‐sea heat flux parameterization using a physical‐informed neural network (PINN) based on multiple observational data sets worldwide. Evaluated with an independent observation data set, it is shown that the PINN can significantly reduce the RMSE of latent heat flux by at least about 48.6% compared to three traditional bulk formulas. Moreover, the PINN can be flexibly updated with new observational data by transfer learning. To test the performance of the new parameterization in realistic application, we implement the PINN into a global ocean‐atmosphere coupled model and make seasonal forecasts for the first time. The PINN markedly reduce the errors of equatorial SST forecast, indicating a good performance of the PINN‐based air‐sea turbulent heat flux scheme. Noticeably, due to limited observational data, the NN‐based parameterizations tend to underestimate heat flux at high wind speeds compared with bulk formula‐based parameterizations. With more data available at extreme conditions, the PINN can be improved via transfer learning and need to be futher evaluated. This study suggests that PINN‐based air‐sea heat flux parameterization is promising to improve SST simulation.

Research on time characteristic sea surface temperature prediction based on decomposition

Abstract The effective prediction of sea surface temperature has become a hot topic of scientific research. The hidden characteristics and time characteristics of sea surface temperature are studied, and a new sea surface temperature prediction model SLA based on decomposition is proposed in this paper. Firstly, the trend term and periodic seasonal term of sea surface temperature were decomposed by STL decomposition algorithm, which improved the problem of single characteristics of sea surface. Secondly, on the basis of the calculation method of the improved LSTM model and the attention mechanism, the problem that the prediction of sea surface temperature cannot be remembered in the long term and the influence weight of historical information is ignored is solved. Finally, the sea surface temperature of the South China Sea (112.125E,10.125N) was taken as the research object, and the experimental verification was carried out in three data sets. The experimental results show that the SLA model proposed in this paper has good prediction accuracy.

A Non-Uniform Grid Graph Convolutional Network for Sea Surface Temperature Prediction

Sea surface temperature (SST) is an important factor in the marine environment and has significant impacts on climate, ecology, and maritime activities. Most existing SST prediction methods consider the ocean as a uniform field and use a uniform grid to predict SST. However, the marine environment is a complex system, and factors such as solar radiation, differences in land and sea thermal properties, and ocean circulation lead to uneven spatial distributions of SSTs. We propose a non-uniform grid construction method based on an SST spatial gradient to encode SST data, as well as a Non-uniform Grid Graph Convolutional Network (NGGCN) model. The NGGCN consists of two spatiotemporal modules, each of which extracts spatial features from the GCN module, captures temporal correlations through the GRU module, and performs feature restoration and output results through the fully connected module. We selected data from the Yellow Sea and Bohai Sea to validate the effectiveness of the NGGCN in predicting SST at different time scales and prediction steps. The results indicate that our model shows a significant improvement in prediction performance compared to other models.

Enhanced Atlantic Meridional Mode predictability in a high-resolution prediction system.

Accurate prediction of sea surface temperatures (SSTs) in the tropical North Atlantic on multiyear timescales is of paramount importance due to its notable impact on tropical cyclone activity. Recent advances in high-resolution climate predictions have demonstrated substantial improvements in the skill of multiyear SST prediction. This study reveals a notable enhancement in high-resolution tropical North Atlantic SST prediction that stems from a more realistic representation of the Atlantic Meridional Mode and the associated wind-evaporation-SST feedback. The key to this improvement lies in the enhanced surface wind response to changes in cross-equatorial SST gradients, resulting from Intertropical Convergence Zone bias reduction when atmospheric model resolution is increased, which, in turn, amplifies the positive feedback between latent and sensible surface heat fluxes and SST anomalies. These advances in high-resolution climate prediction hold promise for extending tropical cyclone forecasts at multiyear timescales.

TL-iTransformer: Revolutionizing sea surface temperature prediction through iTransformer and transfer learning

TL-iTransformer: Revolutionizing sea surface temperature prediction through iTransformer and transfer learning

STA-SST: Spatio-temporal time series prediction of Moroccan Sea surface temperature

Global Sea Surface Temperature (SST) trends have garnered significant attention in several ocean-related domains, including global warming, marine biodiversity, and environmental protection. This involves having an accurate and efficient forecast of future SST to ensure early detection and response in time to these events. Deep learning algorithms have become popular in SST prediction recently, although directly obtaining optimal prediction results from historical observation data is not simple. In this paper, we propose STA-SST, a new deep learning approach for forecasting SST, by combining the temporal dependencies of SST using the Bidirectional Long Short-Term Memory (BiLSTM) model, spatial features extracted from the convolution layer, and relevant information from the attention mechanism. To assess how well the Attention-BiLSTM with convolution layer predicts SST, we conducted a case study in the Moroccan Sea, concentrating on five different areas. The proposed model was compared against alternative forecasting models, including LSTM, XGBoost, Support Vector Regression (SVR), and Random Forest (RF). The experimental results show that STA-STT produces noticeably the best prediction results and is a solid choice for field implementation.

Prediction and analysis of sea surface temperature based on LSTM-transformer model

This paper introduces a novel method for predicting sea surface temperature (SST) using a hybrid LSTM-Transformer model. The study utilizes the ERA5 dataset for SST prediction at six specific locations near China. The LSTM-Transformer model combines the temporal processing capability of LSTM with the efficient data processing power of Transformer, showing superior performance in reducing Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE), as well as improving the R-squared (R²) value, compared to standalone LSTM, Transformer, and traditional Logistic Regression (LR) models. This is particularly evident in spring and autumn, indicating its robustness to seasonal changes. The model's performance varies across different geographic locations, with lower prediction errors in low-latitude and open sea areas, attributed to the less complex environmental dynamics compared to continental shelf areas. Overall, the LSTM-Transformer hybrid model presents a significant advancement in SST prediction, providing important implications for fisheries, meteorology, and climate change research.

Forecasting sea surface temperature during typhoon events in the Bohai Sea using spatiotemporal neural networks

Advanced neural network methods can effectively combine temporal and spatial information, allowing us to exploit complex relationships in the prediction of sea surface temperature (SST). Therefore, the authors have developed an attention-based context fusion network model recently, known as ACFN, to enhance the underlying spatiotemporal correlations in SST data. In a previous preliminary long-term evaluation, ACFN has demonstrated substantial improvements over several state-of-the-art baseline models, such as Convolutional Long-Short Term Memory (ConvLSTM) and Predictive Recurrent Neural Network (PredRNN). However, these methods have not been widely used for predicting SST during typhoons, and their performance has not been thoroughly evaluated. This study focuses on three specific typhoons, In-fa (2021), Lekima (2019), and Rumbia (2018), as case studies. The results show that ACFN consistently outperforms ConvLSTM and PredRNN, as demonstrated during Typhoon In-fa with a mean absolute error of 0.27 °C and a root-mean-square error of 0.33 °C for a 1-day forecast time. This study also examines the impact of forecast time on predictive skill, revealing that the performance of ACFN, as expected, gradually decreases with longer forecast times. This thorough evaluation of the ACFN model provides a valuable reference for using deep learning in predicting regional typhoon SST. Plain language summaryIn previous work, we developed an attention-based context fusion network model. This spatiotemporal neural network model has been successfully applied to short-term sea surface temperature (SST) prediction in the Bohai Sea, utilizing only current and previous 1- to 10-day SST data as input to forecast SST for the subsequent 1- to 10-day period. In this study, we further evaluated the capability of the model to predict SST during typhoons in the Bohai Sea, and the results were very promising. When predicting SST during Typhoon In-fa, the present spatiotemporal neural network model had a mean absolute error of only 0.27 °C for a 1-day forecast time, and a correlation coefficient of 0.74. On the other hand, comparison with baseline spatiotemporal neural network models shows that the present model is more effective in extracting input information. The current model has resolved the constraint that its predictive capability does not diminish as the forecast time increases, thereby enhancing the rationality of the prediction outcomes. In conclusion, the present spatiotemporal neural network model demonstrates its efficacy in forecasting typhoon-induced SST in the Bohai Sea, offering valuable insights for applications in meteorology and oceanography.

Investigating the seasonal SST Predictability in the Northern Tropical Atlantic Ocean in an ensemble prediction system

The present study comprehensively investigates the practical and intrinsic predictability of the sea surface temperature (SST) in the Northern Tropical Atlantic (NTA) based on the 138-year-long coupled hindcasts with a recently developed seasonal ensemble prediction system. This system can yield skillful deterministic predictions for the prominent warm and cold events at least 6 months ahead. Notably, it excels in providing probabilistic predictions for below- and above-normal events rather than for neutral events. The predictability of SST in the NTA undergoes remarkable seasonal variation with two peaks of predictability targeted at April and October regardless of the lead time. Various sources of predictability for these target months are revealed. For the target month of April, the preceding remote forcing from the El Niño–Southern Oscillation (ENSO) in the tropical Pacific Ocean combined with local signal results in the phase locking of the SST variation and seasonality of signal component over the NTA. This ultimately contributes to the high predictability targeted at April. However, From the perspective of potential predictability of the predictability targeted at October, which has been rarely mentioned in previous studies. It is also encouraging that, similar to the Indian Ocean Dipole, ENSO and the signal-to-noise ratio of the system mainly contribute to predictability beyond persistence at long lead times for the spring SST in the NTA. This indicates that potential future ENSO improvements may leave much room for improvement in the current SST prediction in the NTA.

Application of Fast MEEMD–ConvLSTM in Sea Surface Temperature Predictions

Application of Fast MEEMD–ConvLSTM in Sea Surface Temperature Predictions

Using multi-layer perceptron to predict sea surface temperature

Abstract This study employs a Multi-layer Perceptron (MLP) model to predict Sea Surface Temperature (SST) using Sea Surface Salinity (SSS) data collected by NASA over a period of 55 years. SSS is closely related to SST, as both are influenced by similar factors such as solar radiation, evaporation, and precipitation. The accuracy of these predictions is then evaluated through an error analysis, conducted on both annual and monthly scales. The results of this study indicate that the MLP model can effectively utilize SSS data to predict SST. However, it was observed that the model’s predictive performance varies across different seasons and regions. This study demonstrates that the MLP model is an effective tool for predicting SST based on SSS data. By employing a MLP to predict SST based on SSS data, this study contributes to the field of meteorology in several ways. However, further research and optimization of the model are needed to improve its predictive accuracy. Additionally, more data needs to be collected and the model’s performance needs to be validated across a wider temporal and spatial scale.

Spatiotemporal Fusion Prediction of Sea Surface Temperatures Based on the Graph Convolutional Neural and Long Short-Term Memory Networks

Sea surface temperature (SST) prediction plays an important role in scientific research, environmental protection, and other marine-related fields. However, most of the current prediction methods are not effective enough to utilize the spatial correlation of SSTs, which limits the improvement of SST prediction accuracy. Therefore, this paper first explores spatial correlation mining methods, including regular boundary division, convolutional sliding translation, and clustering neural networks. Then, spatial correlation mining through a graph convolutional neural network (GCN) is proposed, which solves the problem of the dependency on regular Euclidian space and the lack of spatial correlation around the boundary of groups for the above three methods. Based on that, this paper combines the spatial advantages of the GCN and the temporal advantages of the long short-term memory network (LSTM) and proposes a spatiotemporal fusion model (GCN-LSTM) for SST prediction. The proposed model can capture SST features in both the spatial and temporal dimensions more effectively and complete the SST prediction by spatiotemporal fusion. The experiments prove that the proposed model greatly improves the prediction accuracy and is an effective model for SST prediction.

Sea surface temperature prediction by stacked generalization ensemble of deep learning

Accurate sea surface temperature (SST) prediction is of great significance for fishery farming, marine ecological protection, and planning of maritime activities. In this paper, the stacked generalization ensemble is demonstrated for SST prediction to improve every single deep-learning model. Long-term high-resolution satellite-derived SST is used with the sub-regions of the Taiwan Strait and East China Sea taken as the study area. We select the Multilayer Perceptron, Long short-term memory (LSTM), Convolutional Neural Networks (CNN), CNN-LSTM as individual learners, and Convolutional LSTM as the meta-learner. The individual learners are trained and validated on the retained data subset I, while the meta-learner is trained and validated by constructing the samples with the predictions of validated individual learners on the retained data subset II. The two types of models are evaluated on the same test dataset with root-mean-square error and coefficient of efficiency as the scoring criteria. We find that the meta-model outperforms any individual model and other baselines for the one-day-/three-day-ahead forecasts in the Taiwan Strait and one-day-/three-day-/five-day-ahead predictions in the East China Sea. Furthermore, when the lead time is 1 day and 3 days, the meta-model has a better spatial distribution of prediction metrics across all grid points in the Taiwan Strait sub-area. For the East China Sea sub-region, the meta-model advantage is extended to the lead time of 5 days. Probably due to the higher quality of offshore satellite data, the prediction ability enhancement of the stacked ensemble applied in the East China Sea is better than that in the Taiwan Strait. The better-performing meta-model prediction suggests that the stacked generalization ensemble is encouraging and promising for improving the short-term prediction of the daily SST field.