Sonny-ENS: demystifying machine-learning ensemble weather forecasting
Abstract
Every weather forecast is uncertain, and decisions from flood warnings to energy trading depend on how well that uncertainty is quantified. Machine-learning (ML) ensembles now quantify it better than the best physics-based ensemble, yet two things about them remain opaque: whether their skill requires hundreds to thousands of accelerator-days of training, and how a network with a simple source of randomness produces a spread that follows the flow. Here we introduce Sonny-ENS, an ensemble of three 68.9-million-parameter networks with Monte Carlo dropout, trained on the continuous ranked probability score (CRPS) in 2.7 A40 GPU-days. On 2022, it has a lower CRPS than the European Centre for Medium-Range Weather Forecasts ensemble (ENS) on 96% of 570 targets and an 8.3% lower mean CRPS at lead times of 1-10 days. It outperforms the operational GenCast on 69% of the targets, and a 51-member, 15-day forecast takes 69 s on one GPU. Ablations identify the ingredients of this skill, and a Kolmogorov-Arnold sparse autoencoder shows that dropout injects uncertainty mainly in the tropics, whereas the spread grows where the atmosphere is baroclinically unstable (spatial correlation 0.91 with the Eady growth rate). Skillful probabilistic weather forecasts are therefore within reach of a single GPU server.
Related
- Weather forecasting in a changing climate: the rise of AI and Machine learning?
- A Station-Based Evaluation of Machine Learning-based Weather Forecasting Models in Northern Norway
- Conditional-Risk Optimization and KKT Geometry of a Regime- and Horizon-Adaptive Probabilistic Ensemble for Local Weather Forecasting
- Uncertainty analysis of a machine learning-aided ensemble forecast for coastal flood maps
- Hybrid machine learning framework for drought forecasting: integration of LSBoost, RQNN, CNN-RNN, CNN-LSTM and advanced data decomposition
BibTeX
@misc{cheon2026sonnyens,
title = {{Sonny-ENS: demystifying machine-learning ensemble weather forecasting}},
author = {Minjong Cheon},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.01614},
eprint = {2611.01614},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.01614},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Sonny-ENS: demystifying machine-learning ensemble weather forecasting
AU - Minjong Cheon
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.01614
UR - https://arxiv.org/abs/2611.01614
AB - Every weather forecast is uncertain, and decisions from flood warnings to energy trading depend on how well that uncertainty is quantified. Machine-learning (ML) ensembles now quantify it better than the best physics-based ensemble, yet two things about them remain opaque: whether their skill requires hundreds to thousands of accelerator-days of training, and how a network with a simple source of randomness produces a spread that follows the flow. Here we introduce Sonny-ENS, an ensemble of three 68.9-million-parameter networks with Monte Carlo dropout, trained on the continuous ranked probability score (CRPS) in 2.7 A40 GPU-days. On 2022, it has a lower CRPS than the European Centre for Medium-Range Weather Forecasts ensemble (ENS) on 96% of 570 targets and an 8.3% lower mean CRPS at lead times of 1-10 days. It outperforms the operational GenCast on 69% of the targets, and a 51-member, 15-day forecast takes 69 s on one GPU. Ablations identify the ingredients of this skill, and a Kolmogorov-Arnold sparse autoencoder shows that dropout injects uncertainty mainly in the tropics, whereas the spread grows where the atmosphere is baroclinically unstable (spatial correlation 0.91 with the Eady growth rate). Skillful probabilistic weather forecasts are therefore within reach of a single GPU server.
ER -