Oct 2026· Environmental Data Science· Vol 5· 0 citations· 24 references
TL;DR
A method to represent location-based uncertainty through random, spatiotemporally coherent displacements in physical space, which are used to sample states in the vicinity of a given reference state is proposed, which provides a structured way for modeling and training DL models for dealing with location-based uncertainty.
Abstract
Abstract While deep learning (DL) methods have expanded the approach to solving state estimation problems in data assimilation (DA), a robust approach requires extending them for uncertainty quantification. In this work, we propose a method to represent location-based uncertainty through random, spatiotemporally coherent displacements in physical space, which are used to sample states in the vicinity of a given reference state. We integrate this strategy into an end-to-end variational DA framework, termed 4DVarNet-LU, and compare its performance with a UNet baseline, demonstrating improved reconstruction skill through the explicit use of prior state information. Trained deterministically in output and stochastically in input, we demonstrate that at inference, we see improvement in skill when using information from initial conditions. Using observation system simulation experiments based on a single-layer quasi-geostrophic model with sparse pseudo-nadir altimetry observations, we evaluate reconstruction mean and ensemble skill. Results show that 4DVarNet-LU consistently outperforms the UNet baseline in terms of root mean square error (RMSE) and continuous ranked probability score (CRPS), with increased robustness to initial condition bias and improved temporal stability. While the method does not propose to account for all different uncertainty sources in DA, it provides a structured way for modeling and training DL models for dealing with location-based uncertainty. Future work may extend this approach to account for additional sources of uncertainty and end-to-end ensemble generation.
Data Assimilation (DA) aims to recover the full state of a dynamical system that is only partially observed. A solution is to use Score-based models to generate physically consistent trajectories that agree with the observations. These Autoregressive Diffusion models are trained by conditioning on the previous state; h...
IHEnKS with spatial and temporal localizations and hybridization achieves the best reconstructions compared to various configurations of ensemble‐based assimilation methods and become more pronounced as the proxy network becomes sparser.
Hao-Hao Sun, Li-Li Lei, Zhe-Min Tan et al.· Journal of Advances in Model...· 1 citation
A hybrid framework that couples a convolutional neural network with a learnable-kernel Gaussian Process to jointly perform DA and bias correction from extremely sparse observations is proposed, outperforms 3D-Var, a classical GP and a competing DL model, while reducing inference time.
Emanuele Mele, A. Coluccia, M. Cafaro et al.· 0 citations
Computationally expensive simulations and real-world observations often provide data at a lower temporal resolution than downstream tasks require. Averaging observations at fixed grid points can introduce errors, and motion estimates alone do not determine how the physical field values change. In this work, we study wh...
Pu Ren, Michael W. Mahoney, N. Benjamin Erichson· 0 citations
Weather reanalysis products rely on computationally intensive numerical weather predictions followed by data assimilation that corrects the forecast toward observations. Recent advances in deep generative models offer a cheaper alternative that shifts much of this cost from inference to offline training. However, exist...
Rui-Zhe Huang, Qi-Dong Yang, J. Giezendanner et al.· 0 citations
Many approaches have recently been proposed to combine data assimilation with machine learning for improved forecasts with geophysical models. Here, we focus on combining advanced sequential data assimilation schemes with a generative forecasting model which accounts for complex model errors. This generative model prov...
M. Bocquet, T. Finn, Si-Bo Cheng· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.