Surrogate modeling for the climate sciences dynamics with machine learning and data assimilation

Bocquet, Marc

doi:10.3389/fams.2023.1133226

Cited by 9 publications

(7 citation statements)

References 103 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Peyron et al (2021) proposed an ETKF-Q-L method that learned the latent structure of the dynamic using an autoencoder to reduce the computational cost and memory storage. These interdisciplinary approaches, which combine ML and DA, have shown promising results in improving the accuracy and interpretability of models across various domains (Bocquet, 2023). However, the demand for huge ensemble members requires more external storage and computational resources than the proposed end-to-end model.…”

Section: Ml-based Surrogate Models For Ensemble Damentioning

confidence: 99%

“…In the earth science domain, ML also offers a powerful toolkit to improve the computational efficiency of models and extract information from large amounts of data about Earth (Düben et al., 2021). Further, Bocquet (2023) and Cheng et al. (2023) highlight the potential of ML and DA for improving the accuracy and efficiency of models in Earth sciences.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Four‐Dimensional Variational Constrained Neural Network‐Based Data Assimilation Method

Wang,

Ren,

Duan

et al. 2024

J Adv Model Earth Syst

View full text Add to dashboard Cite

Advances in data assimilation (DA) methods and the increasing amount of observations have continuously improved the accuracy of initial fields in numerical weather prediction during the last decades. Meanwhile, in order to effectively utilize the rapidly increasing data, Earth scientists must further improve DA methods. Recent studies have introduced machine learning (ML) methods to assist the DA process. In this paper, we explore the potential of a four‐dimensional variational (4DVar) constrained neural network (NN) method for accurate DA. Our NN is trained to approximate the solution of the variational problem, thereby avoiding the need for expensive online optimization when generating the initial fields. In the context that the full‐field system truths are unavailable, our approach embeds the system's kinetic features described by a series of analysis fields into the NN through a 4DVar‐form loss function. Numerical experiments on the Lorenz96 physical model demonstrate that our method can generate better initial fields than most traditional DA methods at a low computational cost, and is robust when assimilating observations with higher error outside of the distributions where it is trained. Furthermore, our NN‐based DA model is effective against Lorenz96 physical models with larger variable numbers. Our approach exemplifies how ML methods can be leveraged to improve both the efficiency and accuracy of DA techniques.

show abstract

Section: Ml-based Surrogate Models For Ensemble Damentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Four‐Dimensional Variational Constrained Neural Network‐Based Data Assimilation Method

Wang,

Ren,

Duan

et al. 2024

J Adv Model Earth Syst

View full text Add to dashboard Cite

show abstract

“…Surrogate modeling addresses this challenge by creating data-driven models that approximate the behavior of physical models and predict the outputs of these complex systems with much less computational effort. This makes them extremely valuable in scenarios where rapid decision-making is crucial, such as emergency response planning, environmental impact assessments, or policy development [11,38,60,102,132,179,180,184]. Surrogate models are typically developed using advanced machine learning tools [11,60,144,158,179,184] (e.g., Koopman operator [15]) or statistical methods [43], and are trained on simulated data generated by complex physical models.…”

Section: Surrogate Modeling Solving Pdesmentioning

confidence: 99%

“…This makes them extremely valuable in scenarios where rapid decision-making is crucial, such as emergency response planning, environmental impact assessments, or policy development [11,38,60,102,132,179,180,184]. Surrogate models are typically developed using advanced machine learning tools [11,60,144,158,179,184] (e.g., Koopman operator [15]) or statistical methods [43], and are trained on simulated data generated by complex physical models. To fully capture the essential patterns and relationships inherent in environmental processes, scientific knowledge can also be leveraged to enhance the ML-based surrogate models.…”

Section: Surrogate Modeling Solving Pdesmentioning

confidence: 99%

Knowledge-Guided Machine Learning

Karpatne¹,

Kannan²,

Kumar³

2022

View full text Add to dashboard Cite

This paper presents an overview of scientific modeling and discusses the complementary strengths and weaknesses of ML methods for scientific modeling in comparison to process-based models. It also provides an introduction to the current state of research in the emerging field of scientific knowledge-guided machine learning (KGML) that aims to use both scientific knowledge and data in ML frameworks to achieve better generalizability, scientific consistency, and explainability of results. We discuss different facets of KGML research in terms of the type of scientific knowledge used, the form of knowledge-ML integration explored, and the method for incorporating scientific knowledge in ML. We also discuss some of the common categories of use cases in environmental sciences where KGML methods are being developed, using illustrative examples in each category.

show abstract

“…A solution to mitigate the biases would be to update the reanalysis dataset using the trained MLWP model, effectively combining data assimilation and machine learning as originally proposed by Brajard et al (2020); Bocquet et al (2020). However, the existing MLWP models are by construction designed for forecasting tasks and are far from being suitable for assimilation purposes (Bocquet, 2023). Beyond these considerations, they also present some limitations in the forecast applications.…”

Section: Introductionmentioning

confidence: 99%

Online model error correction with neural networks in the incremental 4D-Var framework

Farchi

Chrust

Bocquet

et al. 2022

Preprint

View full text Add to dashboard Cite

Recent studies have demonstrated that it is possible to combine machine learning with data assimilation to reconstruct the dynamics of a physical model partially and imperfectly observed. Data assimilation is used to estimate the system state from the observations, while machine learning computes a surrogate model of the dynamical system based on those estimated states. The surrogate model can be defined as an hybrid combination where a physical model based on prior knowledge is enhanced with a statistical model estimated by a neural network. The training of the neural network is typically done offline, once a large enough dataset of model state estimates is available. By contrast, with online approaches the surrogate model is improved each time a new system state estimate is computed. Online approaches naturally fit the sequential framework encountered in geosciences where new observations become available with time. In a recent methodology paper, we have developed a new weak-constraint 4D-Var formulation which can be used to train a neural network for online model error correction. In the present article, we develop a simplified version of that method, in the incremental 4D-Var framework adopted by most operational weather centres. The simplified method is implemented in the ECMWF Object-Oriented Prediction System, with the help of a newly developed Fortran neural network library, and tested with a two-layer two-dimensional quasi geostrophic model. The results confirm that online learning is effective and yields a more accurate model error correction than offline learning. Finally, the simplified method is compatible with future applications to state-of-the-art models such as the ECMWF Integrated Forecasting System.

show abstract

Surrogate modeling for the climate sciences dynamics with machine learning and data assimilation

Cited by 9 publications

References 103 publications

A Four‐Dimensional Variational Constrained Neural Network‐Based Data Assimilation Method

A Four‐Dimensional Variational Constrained Neural Network‐Based Data Assimilation Method

Knowledge-Guided Machine Learning

Online model error correction with neural networks in the incremental 4D-Var framework

Contact Info

Product

Resources

About