Predicting weather forecast uncertainty with machine learning
Problem
Estimating confidence in weather forecasts currently requires expensive ensemble simulations, so a cheaper alternative was sought.
Why it matters
Forecasts are only valuable for many applications if an uncertainty estimate can be attached to them.
What the AI did
A deep-learning convolutional neural network, trained on past forecasts, assigns a scalar confidence value to medium-range forecasts from the initial atmospheric state.
Results
Lower skill than ensemble weather forecast models (abstract truncated before further metrics)
Impact
Peer-reviewed alternative approach, but it does not yet match ensemble skill.
Record
| Venue | Quarterly Journal of the Royal Meteorological Society |
|---|---|
| Published | 2018 |
| Work type | article |
| Authors | Sebastian Scher, Gabriele Messori |
| Citations | 223 (OpenAlex, at import time) |
| References | 23 |
| Model or method | convolutional neural networks (deep learning) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1002/qj.3410 |
| Open access | green |
| Keywords | weather forecasting, uncertainty prediction, deep learning, CNN, ensemble forecasts |
Importance rating
real but narrower advance
Abstract states forecasts are only valuable for many applications if uncertainty estimates exist and ensembles are computationally very expensive; the ML alternative is efficient but has lower skill, a narrower methodological advance.
More in Climate & Environment
| # | Paper | Year | Citations | Importance |
|---|---|---|---|---|
| 1 | Learning skillful medium-range global weather forecasting | 2023 | 1,416 |
8
|
| 2 | Accurate medium-range global weather forecasting with 3D neural networks | 2023 | 1,631 |
8
|
| 3 | FuXi: a cascade machine learning forecasting system for 15-day global weather forecast | 2023 | 355 |
8
|
| 4 | Skilful nowcasting of extreme precipitation with NowcastNet | 2023 | 451 |
8
|
| 5 | Forecasting Renewable Energy Generation with Machine Learning and Deep Learning: Current Advances and Future Prospects | 2023 | 378 |
8
|