Learning Variational Data Assimilation Models and Solvers - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Journal Articles Journal of Advances in Modeling Earth Systems Year : 2021

Learning Variational Data Assimilation Models and Solvers


This paper addresses variational data assimilation from a learning point of view. Data assimilation aims to reconstruct the time evolution of some state given a series of observations, possibly noisy and irregularly-sampled. Using automatic differentiation tools embedded in deep learning frameworks, we introduce end-to-end neural network architectures for data assimilation. It comprises two key components: a variational model and a gradient-based solver both implemented as neural networks. A key feature of the proposed end-to-end learning architecture is that we may train the NN models using both supervised and unsupervised strategies. Our numerical experiments on Lorenz-63 and Lorenz-96 systems report significant gain w.r.t. a classic gradient-based minimization of the variational cost both in terms of reconstruction performance and optimization complexity. Intriguingly, we also show that the variational models issued from the true Lorenz-63 and Lorenz-96 ODE representations may not lead to the best reconstruction performance. We believe these results may open new research avenues for the specification of assimilation models in geoscience.
Fichier principal
Vignette du fichier
learning4DVar (16).pdf (1.04 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-02906798 , version 1 (25-07-2020)



Ronan Fablet, Bertrand Chapron, Lucas Drumetz, Etienne Mémin, Olivier Pannekoucke, et al.. Learning Variational Data Assimilation Models and Solvers. Journal of Advances in Modeling Earth Systems, 2021, 13 (10), pp.article n° e2021MS002572. ⟨10.1029/2021MS002572⟩. ⟨hal-02906798⟩
195 View
408 Download



Gmail Facebook X LinkedIn More