Toth et al. (2025) Weather forecasting: what sets the forecast skill horizon?

Identification

• Journal: Elsevier eBooks
• Year: 2025
• Date: 2025-11-14
• Authors: Zoltan Toth, Roberto Buizza, Jie Feng
• DOI: 10.1016/b978-0-443-31538-1.00021-x

Research Groups

• NOAA Global Systems Laboratory, Boulder, CO, United States
• Scuola Universitaria Superiore Sant'Anna, Pisa, Italy
• Department of Atmospheric and Oceanic Sciences, Fudan University, Shanghai, P.R. China

Short Summary

This paper examines the evolution of weather forecasting skill over the past 50 years, identifying the current practical limits of predictability for instantaneous weather and the transition to subseasonal-to-seasonal forecasting for weather statistics.

Objective

• To explore what sets the forecast skill horizon for weather prediction, examining the advancements in skill over time and the current limits of predictability for instantaneous weather and subseasonal-to-seasonal forecasts.

Study Configuration

• Spatial Scale: Synoptic-scale features (e.g., Northern Hemisphere extra tropics), larger spatial scale variations for subseasonal-to-seasonal (S2S) predictions.
• Temporal Scale: Short lead times (e.g., 5, 7, 10 days), up to approximately 15 days for instantaneous weather, and subseasonal-to-seasonal for statistics of weather.

Methodology and Data

• Models used: Not explicitly detailed, but refers to advanced prediction techniques and ensemble-based probabilistic approaches.
• Data sources: Not explicitly detailed, but implies operational forecast data and historical skill assessments.

Main Results

• Over the Northern Hemisphere extra tropics, 10-day forecasts of synoptic-scale features today are as skillful as 7-day forecasts were 40 years ago.
• 7-day forecasts today are as skillful as 5-day forecasts were 25 years ago.
• The skill in predicting instantaneous weather is generally lost around a 15-day lead time.
• Beyond 15 days, only lower, subseasonal time frequency and larger spatial scale variations in the statistics of weather are predictable (subseasonal-to-seasonal prediction).

Contributions

• Provides a historical perspective on the significant improvement of weather forecast skill over the past 50 years.
• Defines the current practical limits of predictability for instantaneous weather (approximately 15 days) and the transition to subseasonal-to-seasonal predictability for weather statistics.
• Establishes the context for understanding the challenges and potential of subseasonal-to-seasonal prediction.

Funding

• Not specified in the provided text.

Citation

@article{Toth2025Weather,
  author = {Toth, Zoltan and Buizza, Roberto and Feng, Jie},
  title = {Weather forecasting: what sets the forecast skill horizon?},
  journal = {Elsevier eBooks},
  year = {2025},
  doi = {10.1016/b978-0-443-31538-1.00021-x},
  url = {https://doi.org/10.1016/b978-0-443-31538-1.00021-x}
}