mlsanityData to 14 Oct 2026

Appa: Bending Weather Dynamics with Latent Diffusion Models for Global Data Assimilation

arXiv 2504.18720 ↗First posted 25 Apr 2025cs.LG physics.ao-phcited by 20

On the map: Diffusion models → Flow matching, time series and science

Abstract (arXiv)

Deep learning has advanced weather forecasting, but accurate predictions first require identifying the current state of the atmosphere from observational data. In this work, we introduce Appa, a score-based data assimilation model generating global atmospheric trajectories at 0.25sidegree 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 plausible trajectories, 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.

Key facts. Cited by 20 papers in the map (data through 14 Oct 2026). Placed in Flow matching, time series and science, one of 14 directions in Diffusion models. Builds on 38 of its 65 references that are in the map. First posted on arXiv 25 Apr 2025.

Builds on

38 of 65 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024A Generative Framework for Probabilistic, Spatiotemporally Coherent Downscaling of Climate Simulation19 Dec
  2. 2024On conditional diffusion models for PDE simulations21 Oct
  3. 2024DiffDA: a Diffusion Model for Weather-scale Data Assimilation11 Jan

Other references · 35

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 9,122
  4. 2017Graph Attention Networkscited by 8,923
  5. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  6. 2020Scaling Laws for Neural Language Modelscited by 6,586
  7. 2022Scalable Diffusion Models with Transformerscited by 6,054
  8. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,101
  9. 2013Auto-Encoding Variational Bayescited by 4,053
  10. 2022Elucidating the Design Space of Diffusion-Based Generative Modelscited by 3,045
  11. 2020Implicit Neural Representations with Periodic Activation Functionscited by 2,238
  12. 2022Diffusion Posterior Sampling for General Noisy Inverse Problemscited by 1,416
23 more
  1. 2022FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operatorscited by 776
  2. 2024Movie Gen: A Cast of Media Foundation Modelscited by 644
  3. 2021Explaining Neural Scaling Lawscited by 363
  4. 2023FuXi: A cascade machine learning forecasting system for 15-day global weather forecastcited by 241
  5. 2023WeatherBench 2: A benchmark for the next generation of data-driven global weather modelscited by 236
  6. 2024A Survey on Diffusion Models for Inverse Problemscited by 180
  7. 2023FengWu: Pushing the Skillful Global Medium-range Weather Forecast beyond 10 Days Leadcited by 175
  8. 2023Score-based Data Assimilationcited by 96
  9. 2023Machine learning with data assimilation and uncertainty quantification for dynamical systems: a reviewcited by 75
  10. 2023Deep Learning for Day Forecasts from Sparse Observationscited by 72
  11. 2023Geographic Location Encoding with Spherical Harmonics and Sinusoidal Representation Networkscited by 61
  12. 2024Learning Diffusion Priors from Observations by Expectation Maximizationcited by 61
  13. 2024Generative Data Assimilation of Sparse Weather Station Observations at Kilometer Scalescited by 27
  14. 2024Samudra: An AI Global Ocean Emulator for Climatecited by 27
  15. 2024GraphDOP: Towards skilful data-driven medium-range weather forecasts learnt and initialised directly from observationscited by 27
  16. 2023FengWu-4DVar: Coupling the Data-driven Weather Forecasting Model with 4D Variational Assimilationcited by 23
  17. 2025Erwin: A Tree-based Hierarchical Transformer for Large-scale Physical Systemscited by 22
  18. 2025SILO: Solving Inverse Problems with Latent Operatorscited by 19
  19. 2024Fuxi-DA: A Generalized Deep Learning Data Assimilation Framework for Assimilating Satellite Observationscited by 18
  20. 2023Score-based Data Assimilation for a Two-Layer Quasi-Geostrophic Modelcited by 12
  21. 2021Multi-Level Attention Pooling for Graph Neural Networks: Unifying Graph Representations with Multiple Localitiescited by 8
  22. 2024Neural Compression of Atmospheric Statescited by 3
  23. 2022Multimodal learning-based inversion models for the space-time reconstruction of satellite-derived geophysical fieldscited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 7; 2026: 13.

20 papers.

Sort:

By area:

  1. 2025Lost in Latent Space: An Empirical Study of Latent Diffusion Models for Physics Emulationcited by 19
  2. 2025A Probabilistic Approach to Wildfire Spread Prediction Using a Denoising Diffusion Surrogate Modelcited by 11
  3. 2025Multimodal Atmospheric Super-Resolution With Deep Generative Modelscited by 8
  4. 2025Optimization Benchmark for Diffusion Models on Dynamical Systemscited by 7
  5. 2025LaDCast: A Latent Diffusion Model for Medium-Range Ensemble Weather Forecastingcited by 6
  6. 2025DAISI: Data Assimilation with Inverse Sampling using Stochastic Interpolantscited by 5
  7. 2026LEVDA: Latent Ensemble Variational Data Assimilation via Differentiable Dynamicscited by 4
  8. 2026Training-Free Bayesian Filtering with Generative Emulatorscited by 4
  9. 2025How to use score-based diffusion in earth system science: A satellite nowcasting examplecited by 3
  10. 2026WIND: Weather Inverse Diffusion for Zero-Shot Atmospheric Modelingcited by 3
  11. 2026ForcingDAS: Unified and Robust Data Assimilation via Diffusion Forcingcited by 3
  12. 2026Amortized Filtering and Smoothing with Conditional Normalizing Flowscited by 2
8 more
  1. 2026Neural-HSS: Hierarchical Semi-Separable Neural PDE Solvercited by 1
  2. 2026Towards a Foundation Model for the Martian Atmospherecited by 1
  3. 2026ADDA: a Modular Framework for Representing, Simulating and Assimilating Dynamics with End-to-end Differentiabilitycited by 1
  4. 2026FlowDA: Accurate, Low-Latency Weather Data Assimilation via Flow Matchingcited by 0
  5. 2026WindINR: Latent-State INR for Fast Local Wind Query and Correction in Complex Terraincited by 0
  6. 2026Multimodal Spatiotemporal Atmospheric Data Assimilation with Latent Video Flow-matchingcited by 0
  7. 2026Iterative Refinement Diffusion for Super-Resolved Data Assimilation of Multiscale Physical Systemscited by 0
  8. 2026Generative Atmospheric Super-Resolution from Heterogeneous In Situ Observations through Composable Interfacescited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025PhyDA: Physics-Guided Diffusion Models for Data Assimilation in Atmospheric Systems
  2. 2026LD-EnFF: Latent-Dynamics Ensemble Flow Filtering for Data Assimilation with Sparse Observations
  3. 2026Marchuk: Efficient Global Weather Forecasting from Mid-Range to Sub-Seasonal Scales via Flow Matching
  4. 2026Benchmarking Generative Models for Weather Data Assimilation on Real Station Observations
  5. 2026DAWIS: Data Assimilation with Windowed Inverse Sampling via Multitask Interpolants
  6. 2026PuYun-LDM: A Latent Diffusion Model for High-Resolution Ensemble Weather Forecasts
  7. 2025OmniCast: A Masked Latent Diffusion Model for Weather Forecasting Across Time Scales
  8. 2025A PDE-Informed Latent Diffusion Model for 2-m Temperature Downscaling