VAE-Var: Variational-Autoencoder-Enhanced Variational Assimilation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916257345830912 |
|---|---|
| author | Xiao, Yi Jia, Qilong Xue, Wei Bai, Lei |
| author_facet | Xiao, Yi Jia, Qilong Xue, Wei Bai, Lei |
| contents | Data assimilation refers to a set of algorithms designed to compute the optimal estimate of a system's state by refining the prior prediction (known as background states) using observed data. Variational assimilation methods rely on the maximum likelihood approach to formulate a variational cost, with the optimal state estimate derived by minimizing this cost. Although traditional variational methods have achieved great success and have been widely used in many numerical weather prediction centers, they generally assume Gaussian errors in the background states, which limits the accuracy of these algorithms due to the inherent inaccuracies of this assumption. In this paper, we introduce VAE-Var, a novel variational algorithm that leverages a variational autoencoder (VAE) to model a non-Gaussian estimate of the background error distribution. We theoretically derive the variational cost under the VAE estimation and present the general formulation of VAE-Var; we implement VAE-Var on low-dimensional chaotic systems and demonstrate through experimental results that VAE-Var consistently outperforms traditional variational assimilation methods in terms of accuracy across various observational settings. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_13711 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | VAE-Var: Variational-Autoencoder-Enhanced Variational Assimilation Xiao, Yi Jia, Qilong Xue, Wei Bai, Lei Machine Learning Artificial Intelligence Dynamical Systems Atmospheric and Oceanic Physics Data assimilation refers to a set of algorithms designed to compute the optimal estimate of a system's state by refining the prior prediction (known as background states) using observed data. Variational assimilation methods rely on the maximum likelihood approach to formulate a variational cost, with the optimal state estimate derived by minimizing this cost. Although traditional variational methods have achieved great success and have been widely used in many numerical weather prediction centers, they generally assume Gaussian errors in the background states, which limits the accuracy of these algorithms due to the inherent inaccuracies of this assumption. In this paper, we introduce VAE-Var, a novel variational algorithm that leverages a variational autoencoder (VAE) to model a non-Gaussian estimate of the background error distribution. We theoretically derive the variational cost under the VAE estimation and present the general formulation of VAE-Var; we implement VAE-Var on low-dimensional chaotic systems and demonstrate through experimental results that VAE-Var consistently outperforms traditional variational assimilation methods in terms of accuracy across various observational settings. |
| title | VAE-Var: Variational-Autoencoder-Enhanced Variational Assimilation |
| topic | Machine Learning Artificial Intelligence Dynamical Systems Atmospheric and Oceanic Physics |
| url | https://arxiv.org/abs/2405.13711 |