Weather forecasting in a changing climate: the rise of AI and Machine learning?
- 1European Centre for Medium-range Weather Forecasts, Reading, UK
Abstract
Climate change is directly linked to the increase in intensity and frequency of extreme weather events such as heatwaves, heavy rainfall, droughts, floods and hurricanes. In this context, anticipating such extreme events with accurate weather forecasts is of the highest importance in order to provide early warnings to the population and adapt to climate change. Forecasting has long relied on physical models to anticipate atmospheric changes, underpinning decisions that affect millions of lives and livelihoods. In recent years, machine learning has emerged as a powerful new tool, learning from decades of weather data to generate forecasts with remarkable speed and accuracy. Here we show that the integration of machine learning into operational weather prediction, as pioneered by the European Centre for Medium-Range Weather Forecasts, not only matches but regularly surpasses traditional physics-based methods in many metrics. These advances demonstrate that machine learning—far from being a mysterious or threatening ‘black box’—can already be responsibly embedded alongside physics-based methods, boosting predictive skill while fostering new forms of scientific and operational collaboration
Related
- A Station-Based Evaluation of Machine Learning-based Weather Forecasting Models in Northern Norway
- Sonny-ENS: demystifying machine-learning ensemble weather forecasting
- FastNet: Improving the Physical Consistency of Machine Learning Weather Prediction Models through Loss Function Design
- AIFS-DOP: End-to-End Medium-Range Weather Prediction from Observations Alone with Machine Learning
- Hybrid machine learning framework for drought forecasting: integration of LSBoost, RQNN, CNN-RNN, CNN-LSTM and advanced data decomposition
BibTeX
@article{rabier2026weather,
title = {{Weather forecasting in a changing climate: the rise of AI and Machine learning?}},
author = {Florence Rabier and Andrew Brown and Matthew Chantry and Florian Pappenberger},
journal = {Journal of the European Meteorological Society.},
year = {2026},
month = may,
doi = {10.1016/j.jemets.2026.100040},
url = {https://doi.org/10.1016/j.jemets.2026.100040},
note = {CC BY},
}
RIS
TY - JOUR
TI - Weather forecasting in a changing climate: the rise of AI and Machine learning?
AU - Florence Rabier
AU - Andrew Brown
AU - Matthew Chantry
AU - Florian Pappenberger
PY - 2026
DA - 2026/05/19
JO - Journal of the European Meteorological Society.
DO - 10.1016/j.jemets.2026.100040
UR - https://doi.org/10.1016/j.jemets.2026.100040
AB - Climate change is directly linked to the increase in intensity and frequency of extreme weather events such as heatwaves, heavy rainfall, droughts, floods and hurricanes. In this context, anticipating such extreme events with accurate weather forecasts is of the highest importance in order to provide early warnings to the population and adapt to climate change. Forecasting has long relied on physical models to anticipate atmospheric changes, underpinning decisions that affect millions of lives and livelihoods. In recent years, machine learning has emerged as a powerful new tool, learning from decades of weather data to generate forecasts with remarkable speed and accuracy. Here we show that the integration of machine learning into operational weather prediction, as pioneered by the European Centre for Medium-Range Weather Forecasts, not only matches but regularly surpasses traditional physics-based methods in many metrics. These advances demonstrate that machine learning—far from being a mysterious or threatening ‘black box’—can already be responsibly embedded alongside physics-based methods, boosting predictive skill while fostering new forms of scientific and operational collaboration
SN - 2950-6301
ER -