paper
2018
Probabilistic & ensemble forecasting
Post-processing & calibration New model or method
CNN / U-Net3–15 days Journal
2018-10-01
026
Oct 2018
Sebastian Scher, Gabriele Messori
Quarterly Journal of the Royal Meteorological Society
A CNN trained on past forecasts assigns a confidence score to medium-range forecasts without running an ensemble; it is cheaper but less skilful than ensembles.
Probabilistic & ensemble forecasting
Predicting weather forecast uncertainty with machine learning
In one sentence. A CNN trained on past forecasts assigns a confidence score to medium-range forecasts without running an ensemble; it is cheaper but less skilful than ensembles.
Labels
- Topic
- Probabilistic & ensemble forecasting, Post-processing & calibration
- Type
- New model or method
- Method
- CNN / U-Net
- Time scale
- 3–15 days
- Domain
- Global
- Data
- NWP analysis or forecasts (IFS, GFS, ...)
- Status
- Research
- Identifiers
- DOI:10.1002/qj.3410 OpenAlex:W2896006105
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.