VAE-Var：变分自编码器增强的变分同化
VAE-Var: Variational-Autoencoder-Enhanced Variational Assimilation
AI总结：
本文提出VAE-Var，一种利用变分自编码器建模非高斯背景误差的变分同化算法，在低维混沌系统上实验证明其精度优于传统变分方法。
摘要：
数据同化是指一类算法，旨在通过使用观测数据细化先验预测（称为背景状态）来计算系统状态的最优估计。变分同化方法依赖于最大似然方法来构造变分代价，最优状态估计通过最小化该代价获得。尽管传统变分方法已取得巨大成功，并被广泛应用于许多数值天气预报中心，但它们通常假设背景状态中的误差服从高斯分布，由于该假设固有的不准确性，这限制了这些算法的精度。在本文中，我们提出VAE-Var，一种新颖的变分算法，利用变分自编码器（VAE）对背景误差分布的非高斯估计进行建模。我们在理论上推导了VAE估计下的变分代价，并给出了VAE-Var的一般公式；我们在低维混沌系统上实现了VAE-Var，并通过实验结果表明，在各种观测设置下，VAE-Var在精度上始终优于传统变分同化方法。
英文摘要：
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.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
