Lee et al. (2026) Increasing Frequency and Persistence of the Summertime Greenland High Regime Not Captured by a Seasonal Prediction Model Very Large Ensemble

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

Identification

• Journal: Geophysical Research Letters
• Year: 2026
• Date: 2026-01-09
• Authors: Simon H. Lee, Lorenzo M. Polvani
• DOI: 10.1029/2025gl119421

Research Groups

Not explicitly mentioned in the provided abstract.

Short Summary

This study identifies significant increases in the frequency, persistence, and interannual variability of the Greenland High (GH) weather regime in North America from 1981 to 2024, and demonstrates that the SEAS5 seasonal model fails to capture these observed trends due to insufficient persistence trends.

Objective

• To identify and analyze summertime circulation trends, specifically focusing on the Greenland High (GH) weather regime, using a year-round North American regime classification from 1981 to 2024.
• To assess whether a simple Markov model can explain the observed changes in GH frequency and variability through increased persistence.
• To evaluate the ability of the SEAS5 seasonal model to capture these observed trends in GH frequency and understand the underlying reasons for any discrepancies.

Study Configuration

• Spatial Scale: North American region, focusing on the Greenland High (GH) regime, implying a regional to sub-continental scale covering North America and the North Atlantic/Greenland area.
• Temporal Scale: Summertime circulation trends from 1981 to 2024 (44 years), with analysis based on a year-round regime classification and consideration of subseasonal timescales for model deficiencies.

Methodology and Data

• Models used:
  • Simple Markov model (to investigate the role of persistence).
  • SEAS5 seasonal model (10,000-member ensemble for climate trend analysis).
• Data sources:
  • Observational data (implied by "observed trend," specific source not detailed in abstract).
  • SEAS5 seasonal model data.

Main Results

• Large increases were found in the frequency, persistence, and interannual variability of the Greenland High (GH) regime from 1981 to 2024.
• A simple Markov model indicated that the observed increased GH frequency and variability can be attributed to increased persistence of the regime.
• The SEAS5 seasonal model, despite producing summers with more GH days and individual regimes that are more persistent than observed, fails to capture the observed trend in GH frequency.
• This failure in SEAS5 is attributed to its persistence trends being too weak.
• The missing trends in SEAS5 are suggested to arise from fundamental model deficiencies that develop on subseasonal timescales and are not rectified by initialization.

Contributions

• Provides a comprehensive analysis of summertime circulation trends using a year-round North American weather regime classification, specifically highlighting significant changes in the Greenland High regime.
• Quantifies the increasing frequency, persistence, and interannual variability of the Greenland High regime over a 44-year period (1981-2024).
• Demonstrates, using a simple Markov model, that increased persistence is a key driver for the observed changes in GH frequency and variability.
• Identifies a critical deficiency in the SEAS5 seasonal model's ability to capture observed climate trends, specifically its underestimation of persistence trends, despite other aspects of regime behavior being well-represented.
• Emphasizes that these model deficiencies develop on subseasonal timescales and are not resolved by model initialization, pointing to fundamental model limitations for climate variability studies.

Funding

Not explicitly mentioned in the provided abstract.

Citation

@article{Lee2026Increasing,
  author = {Lee, Simon H. and Polvani, Lorenzo M.},
  title = {Increasing Frequency and Persistence of the Summertime Greenland High Regime Not Captured by a Seasonal Prediction Model Very Large Ensemble},
  journal = {Geophysical Research Letters},
  year = {2026},
  doi = {10.1029/2025gl119421},
  url = {https://doi.org/10.1029/2025gl119421}
}