arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2501.13312cs.LG
收藏

Tensor-Var:高效的四维变分数据同化

Tensor-Var: Efficient Four-Dimensional Variational Data Assimilation

University College London(伦敦大学学院)

Yiming Yang, Xiaoyuan Cheng, Daniel Giles, Sibo Cheng, Yi He, Xiao Xue, Boli Chen, Yukun Hu

AI总结:

针对传统四维变分数据同化计算成本高、深度学习集成难且缺理论保证的问题,提出结合核条件均值嵌入的Tensor-Var框架,实现线性化凸优化与理论保证,还支持神经网络深度特征扩展,在混沌系统和全球天气预测中精度更优且速度提升10-20倍。

摘要:

变分数据同化通过最小化使数值模型与观测数据相拟合的代价函数来估计动力系统状态。尽管四维变分同化(4D-Var)应用广泛,但它在复杂非线性系统中存在计算成本高的问题,且依赖不完善的状态-观测映射。深度学习(DL)可提供表达能力更强的近似器,但由于其非线性特性以及同化结果缺乏理论保证,将深度学习模型集成到4D-Var中颇具挑战。本文提出Tensor-Var,这是一种将核条件均值嵌入(CME,kernel conditional mean embedding)与4D-Var相结合的新型框架,可对非线性动力学进行线性化,在学习到的特征空间中实现凸优化。此外,我们的方法为以线性方式求解4D-Var提供了新视角,为原始空间与特征空间之间的同化结果一致性提供了理论保证。为处理大规模问题,我们提出了一种在Tensor-Var框架内使用神经网络学习深度特征的方法。在混沌系统以及结合实时观测的全球天气预测任务上的实验表明,Tensor-Var在精度上优于传统的以及深度学习混合的4D-Var基线方法,同时实现了10到20倍的速度提升。

英文摘要:

Variational data assimilation estimates the dynamical system states by minimizing a cost function that fits the numerical models with the observational data. Although four-dimensional variational assimilation (4D-Var) is widely used, it faces high computational costs in complex nonlinear systems and depends on imperfect state-observation mappings. Deep learning (DL) offers more expressive approximators, while integrating DL models into 4D-Var is challenging due to their nonlinearities and lack of theoretical guarantees in assimilation results. In this paper, we propose Tensor-Var, a novel framework that integrates kernel conditional mean embedding (CME) with 4D-Var to linearize nonlinear dynamics, achieving convex optimization in a learned feature space. Moreover, our method provides a new perspective for solving 4D-Var in a linear way, offering theoretical guarantees of consistent assimilation results between the original and feature spaces. To handle large-scale problems, we propose a method to learn deep features using neural networks within the Tensor-Var framework. Experiments on chaotic systems and global weather prediction with real-time observations show that Tensor-Var outperforms conventional and DL hybrid 4D-Var baselines in accuracy while achieving a 10- to 20-fold speed improvement.

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。

↑