Deep learning has advanced weather forecasting, but accurate prediction requires identifying the current state of the atmosphere from observational data. We introduce Appa, a score-based data assimilation model generating global atmospheric trajectories at 0.25° resolution and 1-hour intervals. Powered by a 565M-parameter latent diffusion model trained on ERA5, Appa can be conditioned on arbitrary observations to infer posterior distributions of plausible states without retraining. Our probabilistic framework handles reanalysis, filtering, and forecasting, within a single model, producing physically consistent reconstructions from various inputs. Results establish latent score-based data assimilation as a promising foundation for future global atmospheric modeling systems.
Reanalysis aims to assimilate weather observations to produce plausible full-state trajectories. We demonstrate that Appa can generate trajectories conditioned on a sequence of partial satellite and ground-station observations. We display here six key variables: surface temperature, surface wind speed (eastward and northward), total precipitation, and atmospheric temperature (850hPa pressure level) and geopotential (500hPa).
Appa's flexibility enables framing forecasting in different manners, including autoregressively, by sequentially generating blankets and conditioning on the initial ground truth and previous estimates. Appa reaches performance comparable to state-of-the-art machine learning forecasting models, despite not being specifically designed for that task.
Thanks to its blanket mechanism, Appa can generate arbitrarily long sequences of weather states. Furthermore, this generation is embarrassingly parallel, reaching as low as about 20-second sampling regardless of the length of the sequence. We demonstrate this by generating a consistent trajectory of approximately 6 months.
@misc{andry2025appabendingweatherdynamics,
title={Appa: Bending Weather Dynamics with Latent Diffusion Models for Global Data Assimilation},
author={Gérôme Andry and François Rozet and Sacha Lewin and Omer Rochman and Victor Mangeleer and Matthias Pirlet and Elise Faulx and Marilaure Grégoire and Gilles Louppe},
year={2025},
eprint={2504.18720},
archivePrefix={arXiv},
primaryClass={cs.LG},
url={https://arxiv.org/abs/2504.18720},
}