Folland et al. (2025) A Review of 25 Annual Forecasts of Global Mean Surface Temperature Including the Record Warm Years 2023 and 2024

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Geophysical Research Letters
• Year: 2025
• Date: 2025-09-22
• Authors: Chris K. Folland, Andrew Colman, Nick Dunstone, Doug Smith, Adam A. Scaife
• DOI: 10.1029/2025gl117308

Research Groups

Not explicitly mentioned in the abstract.

Short Summary

This study evaluates the skill of real-time global mean surface temperature forecasts issued annually for 2000–2025, finding high skill in capturing interannual variability for 2000–2024, though the extreme 2023 warming was missed, while 2024's record warmth was better predicted, especially by dynamical models.

Objective

• To evaluate the skill of real-time global mean surface temperature forecasts against observations for the period 2000–2024, and discuss forecasts for 2023 and 2024.

Study Configuration

• Spatial Scale: Global (global mean surface temperature)
• Temporal Scale: Annual forecasts for 2000–2025, with evaluation covering 2000–2024.

Methodology and Data

• Models used: Real-time forecasts, statistical forecasts, and physics-based dynamical forecasts.
• Data sources: Observations (for comparison and skill assessment).

Main Results

• Forecasts for 2000–2024 demonstrated high skill.
• Correlations between observations and real-time, statistical, and dynamical forecasts were 0.96, 0.95, and 0.94, respectively.
• Root mean square errors (RMSE) for these forecasts were 0.09 °C, 0.11 °C, and 0.10 °C, respectively.
• Interannual forecast correlations, independent of the trend, were high and statistically significant at 0.73, 0.76, and 0.61 for the respective forecast types.
• The observed strong rate of warming interannual variability was well captured by the forecasts.
• The extreme and unexpected jump in global warming in 2023 was not well captured by the forecasts.
• The record warmth of 2024 was better forecast, particularly by the physics-based dynamical model.

Contributions

• Provides a comprehensive assessment of the skill of various real-time global mean surface temperature forecasting methods over a 25-year period.
• Quantifies the performance of statistical and dynamical models in predicting interannual variability and overall warming trends.
• Identifies limitations in capturing abrupt, extreme warming events (e.g., 2023) while highlighting the improved performance of physics-based dynamical models for subsequent record warmth (e.g., 2024).

Funding

Not mentioned in the abstract.

Citation

@article{Folland2025Review,
  author = {Folland, Chris K. and Colman, Andrew and Dunstone, Nick and Smith, Doug and Scaife, Adam A.},
  title = {A Review of 25 Annual Forecasts of Global Mean Surface Temperature Including the Record Warm Years 2023 and 2024},
  journal = {Geophysical Research Letters},
  year = {2025},
  doi = {10.1029/2025gl117308},
  url = {https://doi.org/10.1029/2025gl117308}
}