mlsanityData to 14 Oct 2026

U-Cast: A Surprisingly Simple and Efficient Frontier Probabilistic AI Weather Forecaster

arXiv 2604.09041 ↗First posted 10 Apr 2026cs.LG cs.AI physics.ao-phcited by 12

Not on the map. Why?

Abstract (arXiv)

AI-based weather forecasting now rivals traditional physics-based ensembles, but state-of-the-art (SOTA) models rely on specialized architectures and massive computational budgets, creating a high barrier to entry. We demonstrate that such complexity is unnecessary for frontier performance. We introduce ours, a probabilistic forecaster built on a standard U-Net backbone trained with a simple recipe: deterministic pre-training on Mean Absolute Error followed by short probabilistic fine-tuning on the Continuous Ranked Probability Score (CRPS) using Monte Carlo Dropout for stochasticity. As a result, our model matches or exceeds the probabilistic skill of GenCast and IFS ENS at 1.5^circ resolution while reducing training compute by over 10× compared to leading CRPS-based models and inference latency by over 10× compared to diffusion-based models. U-Cast trains in under 12 H200 GPU-days and generates a 15-day ensemble forecast in 3 seconds. These results suggest that scalable, general-purpose architectures paired with efficient training curricula can match complex domain-specific designs at a fraction of the cost, opening the training of frontier probabilistic weather models to the broader community.

Key facts. Cited by 12 papers in the map (data through 14 Oct 2026). Builds on 30 of its 54 references that are in the map. First posted on arXiv 10 Apr 2026.

Builds on

30 of 54 references.

References · 30

  1. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,255
  2. 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,448
  3. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,926
  4. 2022Elucidating the Design Space of Diffusion-Based Generative Modelscited by 3,045
  5. 2023FuXi: A cascade machine learning forecasting system for 15-day global weather forecastcited by 241
  6. 2023Neural General Circulation Models for Weather and Climatecited by 239
  7. 2023WeatherBench 2: A benchmark for the next generation of data-driven global weather modelscited by 236
  8. 2023Spherical Fourier Neural Operators: Learning Stable Dynamics on the Spherecited by 233
  9. 2022FourCastNet: Accelerating Global High-Resolution Weather Forecasting using Adaptive Fourier Neural Operatorscited by 188
  10. 2022Forecasting Global Weather with Graph Neural Networkscited by 168
  11. 2023Scaling transformer neural networks for skillful and reliable medium-range weather forecastingcited by 110
  12. 2023Uncertainty Quantification over Graph with Conformalized Graph Neural Networkscited by 79
18 more
  1. 2025Skillful joint probabilistic weather forecasting from marginalscited by 63
  2. 2025FourCastNet 3: A geometric approach to probabilistic machine-learning weather forecasting at scalecited by 62
  3. 2024The Importance of Being Scalable: Improving the Speed and Accuracy of Neural Network Interatomic Potentials Across Chemical Domainscited by 40
  4. 2022SwinVRNN: A Data-Driven Ensemble Forecasting Model via Learned Distribution Perturbationcited by 32
  5. 2024ArchesWeather & ArchesWeatherGen: a deterministic and generative model for efficient ML weather forecastingcited by 32
  6. 2024Probabilistic Weather Forecasting with Hierarchical Graph Neural Networkscited by 30
  7. 2024Continuous Ensemble Weather Forecasting with Diffusion modelscited by 30
  8. 2023Advancing Parsimonious Deep Learning Weather Prediction using the HEALPix Meshcited by 28
  9. 2024A Practical Probabilistic Benchmark for AI Weather Modelscited by 25
  10. 2023Scaling Spherical CNNscited by 21
  11. 2024Probabilistic Emulation of a Global Climate Model with Spherical DYffusioncited by 21
  12. 2022WeatherBench Probability: A benchmark dataset for probabilistic medium-range weather forecasting along with deep learning baseline modelscited by 17
  13. 2025Elucidated Rolling Diffusion Models for Probabilistic Forecasting of Complex Dynamicscited by 13
  14. 2025Swift: An Autoregressive Consistency Model for Efficient Weather Forecastingcited by 13
  15. 2025OmniCast: A Masked Latent Diffusion Model for Weather Forecasting Across Time Scalescited by 12
  16. 2025AERIS: Argonne Earth Systems Model for Reliable and Skillful Predictionscited by 8
  17. 2026Probabilistic Retrofitting of Learned Simulatorscited by 7
  18. 2025Controllable Probabilistic Forecasting with Stochastic Decomposition Layerscited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 12.

12 papers.

  1. 2026WeatherNext 3: Increasing resolution and performance of global weather models with raw observationscited by 2
  2. 2026Otter Weather: Skillful and Computationally Efficient Medium-Range Weather Forecastingcited by 1
  3. 2026On the sensitivity of machine-learned probabilistic weather forecast models to scale-aware scoring rulescited by 1
  4. 2026Reliability of Probabilistic Emulation of Physical Systemscited by 0
  5. 2026Kastor: An efficient fine-tuning strategy for generative emulation of PDE simulationscited by 0
  6. 2026A Checklist to assess the energy and carbon impacts of ML/AI applications in Earth System Modelingcited by 0
  7. 2026Stochastically Perturbed Weights: Ensembles from Deterministic Machine-Learning Weather Modelscited by 0
  8. 2026A Station-Based Evaluation of Machine Learning-based Weather Forecasting Models in Northern Norwaycited by 0
  9. 2026Every Fixed Metric Has a Blind Spot: A Learned Atmospheric Critic for Scoring Forecast Realismcited by 0
  10. 2026Lightweight Probabilistic Downscaling from a Deterministic Base Modelcited by 0
  11. 2026Proper Scoring Rule-based Diffusion for Probabilistic Weather Forecastingcited by 0
  12. 2026Xaurora: Generative Weather Forecasting with Denoising Stochastic Interpolants from a Foundation Model Priorcited by 0