García et al. (2026) Seasonal Forecasts as a Tool for Agricultural Drought Early Warning in Portugal

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

Identification

• Journal: Climate
• Year: 2026
• Date: 2026-11-03
• Authors: Daniel Rey García, Diogo Simão, João Rolim, João A. Santos, Paula Paredes
• DOI: 10.3390/cli14110239

Research Groups

Not explicitly stated, but likely involves institutions in Portugal and the European Centre for Medium-Range Weather Forecasts (ECMWF).

Short Summary

This study evaluated ECMWF SEAS5 seasonal forecasts for drought monitoring and early warning in Portugal, finding they offer potential for early warning despite limitations like coarse resolution and a warm-season wet bias, with performance varying by lead time and season.

Objective

• To evaluate ECMWF SEAS5 seasonal forecasts for drought monitoring and early warning in Portugal, considering the accumulation period, lead time, event type, seasonality, and spatial variability.

Study Configuration

• Spatial Scale: Portugal (country-level, with analysis of spatial gradients and variability).
• Temporal Scale: Seasonal forecasts (2017–2024), considering accumulation periods, lead times, and seasonality (cool/warm seasons).

Methodology and Data

• Models used: ECMWF SEAS5 seasonal forecasts.
• Data sources: SEAS5 forecasts, bias-corrected ERA5-Land reanalysis. Forecast performance was assessed by comparing the Standardized Precipitation Evapotranspiration Index (SPEI) derived from both sources, using Cohen’s Kappa and confusion-matrix metrics (True Positive Fraction—TPF; False Positive Fraction—FPF).

Main Results

• Seasonal forecasts (SF) generally reproduced the main spatial gradients and seasonal variability of climatic water availability and captured major drought episodes.
• Limitations included the SF’s coarse spatial resolution and a persistent warm-season wet bias, leading to drought severity underestimation.
• Exact categorical agreement between SF and reanalysis SPEI was limited (29–42%), though quadratically weighted Kappa indicated most disagreements occurred between adjacent SPEI classes.
• Forecast performance decreased with lead time but improved with the accumulation period.
• SPEI-9 retained substantial agreement (quadratically weighted Kappa ≥ 0.6) through the sixth lead month in the warm season.
• Dry conditions were better identified during the cool season (TPF up to 60%), while wet conditions were better identified in the warm season (TPF up to 89%), though the latter was associated with higher false alarm rates (FPF 6–11%).
• Extreme SF percentiles occasionally reduced the wet bias and improved detection but generally showed larger errors; the SF median provided the most balanced skill.

Contributions

• Provides a comprehensive evaluation of ECMWF SEAS5 seasonal forecasts for agricultural drought early warning in Portugal.
• Identifies specific strengths (reproduction of spatial/seasonal variability, capture of major droughts) and limitations (coarse resolution, warm-season wet bias, severity underestimation) of the forecasts in a Mediterranean context.
• Quantifies forecast performance across varying lead times, accumulation periods, and seasons, offering insights into optimal usage.
• Suggests promising pathways for enhancing usability, including bias correction, downscaling methodologies, and multi-system integration for agricultural decision-making.

Funding

Not explicitly stated in the provided text.

Citation

@article{García2026Seasonal,
  author = {García, Daniel Rey and Simão, Diogo and Rolim, João and Santos, João A. and Paredes, Paula},
  title = {Seasonal Forecasts as a Tool for Agricultural Drought Early Warning in Portugal},
  journal = {Climate},
  year = {2026},
  doi = {10.3390/cli14110239},
  url = {https://doi.org/10.3390/cli14110239}
}