Neural Network Driven by Space-time Partial Differential Equation for Predicting Sea Surface Temperature

Yuan, Taikang; Zhu, Junxing; Ren, Kaijun; Wang, Wuxin; Wang, Xiang; Li, Xiaoyong

doi:10.1109/icdm54844.2022.00076

Cited by 5 publications

(4 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At present, real-time SST predictions mainly rely on physics-based dynamical models and data-driven methods, but there are still significant biases and uncertainties that greatly hinder medium-to-long-term SST forecasting [34]. Recent advancements in DL models based on physics-informed neural networks (PINNs) provide a promising approach for nonlinear system modeling, with potential applications in SST prediction.…”

Section: Discussionmentioning

confidence: 99%

“…Recently, a physics-informed deep learning architecture called PINNs and its variants have played a vital role in addressing the poor interpretability and robustness, and strong data dependency, of traditional deep learning models and have demonstrated certain advantages in materials science, chemistry, astrophysics, hydrology, and fluid mechanics [31][32][33][34]. This architecture incorporates physical laws and boundary conditions into neural networks, making the network outputs conform to the requirements of PDEs [35][36][37].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Space-Time Partial Differential Equation Based Physics-Guided Neural Network for Sea Surface Temperature Prediction

et al. 2023

Self Cite

View full text Add to dashboard Cite

Sea surface temperature (SST) prediction has attracted increasing attention, due to its crucial role in understanding the Earth’s climate and ocean system. Existing SST prediction methods are typically based on either physics-based numerical methods or data-driven methods. Physics-based numerical methods rely on marine physics equations and have stable and explicable outputs, while data-driven methods are flexible in adapting to data and are capable of detecting unexpected patterns. We believe that these two types of method are complementary to each other, and their combination can potentially achieve better performances. In this paper, a space-time partial differential equation (PDE) is employed to form a novel physics-based deep learning framework, named the space-time PDE-guided neural network (STPDE-Net), to predict daily SST. Comprehensive experiments for SST prediction were conducted, and the results proved that our method could outperform the traditional finite-difference forecast method and several state-of-the-art deep learning and physics-guided deep learning methods.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Space-Time Partial Differential Equation Based Physics-Guided Neural Network for Sea Surface Temperature Prediction

et al. 2023

Self Cite

View full text Add to dashboard Cite

show abstract

“…PINNs have recently been investigated in industry informatics settings such as modeling flow equations for ocean models [24], modeling crack propagation [27][28], modeling leakage [29], modeling faults [30], and modeling electric loads [31]. Forecasting SST is commonly found as a full-coverage modeling problem combining either generative models [32] [33] or convolutional neural networks [34] with various PDEs. Continual discussion on PINNs and the types of equations usually solved can be reviewed in [4] and [35].…”

Section: Related Workmentioning

confidence: 99%

“…This formulation is useful in our methodology where we want to train a neural network on the observations themselves while regularizing with numerical model data. This differs to similar PINNs that provide full-coverage modeling of ocean and climate features, where the training data is limited to full-coverage reanalysis and the regularizing PDEs are formulated from simpler equations as seen in [32] [33] [34].…”

Section: A Numerical Models Overviewmentioning

confidence: 99%

Forecasting Buoy Observations Using Physics-Informed Neural Networks

Schmidt,

Pokhrel,

Abdelguerfi

et al. 2024

Preprint

View full text Add to dashboard Cite

Methodologies inspired by physics-informed neural networks (PINNs) were used to forecast observations recorded by stationary ocean buoys. We combined buoy observations with numerical models to train surrogate deep learning networks that performed better than with either data alone. Numerical model outputs were collected from two sources for training and regularization: the hybrid circulation ocean model and the fifth ECMWF reanalysis experiment. A hyperparameter determines the ratio of observational and modeled data to be used in the training procedure, so we conducted a grid search to find the most performant ratio. Overall, the technique improved the general forecast performance compared with nonregularized models. Under specific circumstances, the regularization mechanism enabled the PINN models to be more accurate than the numerical models. This demonstrates the utility of combining various climate models and sensor observations to improve surrogate modeling.

show abstract

GSWR-DARN:GNSS-R Sea Surface Wind Speed Retrieval Based on Data Augmentation and Residual Network

Zhou,

Duan,

Ren

et al. 2023

2023 IEEE International Conference on Systems, Man, and Cybernetics (SMC)

View full text Add to dashboard Cite

Neural Network Driven by Space-time Partial Differential Equation for Predicting Sea Surface Temperature

Cited by 5 publications

References 45 publications

A Space-Time Partial Differential Equation Based Physics-Guided Neural Network for Sea Surface Temperature Prediction

A Space-Time Partial Differential Equation Based Physics-Guided Neural Network for Sea Surface Temperature Prediction

Forecasting Buoy Observations Using Physics-Informed Neural Networks

GSWR-DARN:GNSS-R Sea Surface Wind Speed Retrieval Based on Data Augmentation and Residual Network

Contact Info

Product

Resources

About