cs.LGNov 2, 2026

When Latent Dynamics Meet Data Assimilation: A Benchmark of Koopman-Type Representations

Authors: Zhiwei Zhang, Wenxuan Yuan, Xiaoyuan Cheng, Pouria Behnoudfar, Chuanqi Chen, Charlotte Moser, Yuta Tarumi, Jin-Long Wu, 

Organizations: CEREA, ENPC, EDF R&D, Institut Polytechnique de Paris, Île-de-France, France · College of Computing and Data Science, Nanyang Technological University, Singapore · Dynamic Systems Lab, University College London, London, UK · Department of Mathematics, University of Wisconsin–Madison, Madison, WI, USA · 

Abstract

Koopman representations describe nonlinear dynamics through linear evolution in a transformed space, offering an attractive framework for their linear simplicity and generality. Yet these advantages do not necessarily translate into improved performance in deep latent data assimilation (LDA). We systematically compare seven representative methods and structural baselines across five systems spanning approximately periodic to strongly chaotic dynamics and a broad range of state dimensions. We evaluate forecast and assimilation accuracy, physical state recovery, latent-trajectory behavior, and dynamical consistency, and assess Gaussian error assumptions, objective convexity, and uniqueness of analysis solutions. Our results reveal a trade-off between representational flexibility and optimization tractability. Linear evolution of the latent state can stabilize short-term forecasts but restrict expressiveness in strongly chaotic and high-dimensional systems, without consistently improving assimilation accuracy. Parameterized linear operators offer greater flexibility, yet their joint optimization with the encoder--decoder remains nonconvex. Methods with closed-form analysis updates generally perform better and exhibit greater robustness in our experiments, although matrix inversion remains a computational bottleneck. These findings show that the practical benefits of Koopman structure depend on both sufficient observational information and the choice of assimilation procedure.

Figures & tables

Appendix figures & tables22 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Latent Autoencoder Ensemble Kalman Filter for Nonlinear Data assimilation

    Mar 6, 2026Xin T. Tong, Yanyan Wang, Liang YanData AssimilationEnsemble Kalman Filtering

  2. Deep Embedded Multiplicative DMD for Algebra-Preserving Koopman Learning

    Jun 3, 2026Kelan Gray, Finlay Brown, Nicolas Boullé +1Koopman Operator LearningLatent Dynamics Modeling

  3. Learning the Koopman Operator using Attention Free Transformers

    Jun 22, 2026Mohammed Nagdi, Evangelos-Marios Nikolados, Alexey Yermakov +3Long-Term Time Series ForecastingKoopman Operator Learning