Chou et al. (2025) Human influence on recent trends in extratropical low-level wind speed

Identification

• Journal: npj Climate and Atmospheric Science
• Year: 2025
• Date: 2025-12-20
• Authors: Hsing-Hung Chou, Tiffany A. Shaw, Gan Zhang
• DOI: 10.1038/s41612-025-01292-6

Research Groups

• Department of the Geophysical Sciences, The University of Chicago, Chicago, IL, USA
• Department of Climate, Meteorology & Atmospheric Sciences, University of Illinois at Urbana-Champaign, Urbana, IL, USA

Short Summary

This study compares satellite-era trends in extratropical low-level mean and extreme wind speeds from reanalyses with climate model simulations to assess human influence and model fidelity. It finds human influence drives summertime trends in both hemispheres, while wintertime trends exhibit regional model discrepancies linked to sea surface temperature biases.

Objective

• Are low-level extreme wind trends in climate model simulations consistent with the reanalysis trends?
• If yes, can the trends be attributed to human influence?
• If no, what is the source of the reanalysis-model discrepancy?

Study Configuration

• Spatial Scale: Extratropical regions, specifically Southern Hemisphere midlatitudes (45°S to 60°S), Europe, Eurasia, and the subtropical North Pacific. Wind speeds are analyzed at 850 hPa. All datasets are interpolated to a 1.5° × 1.5° grid.
• Temporal Scale: Satellite era, covering 1980 to 2019. Analysis is conducted separately for extended summer (Southern Hemisphere: October to March; Northern Hemisphere: April to September) and extended winter (Southern Hemisphere: April to September; Northern Hemisphere: October to March).

Methodology and Data

• Models used:
  • Coupled Model Intercomparison Project Phase 6 (CMIP6) coupled climate models (25 models).
  • Community Earth System Model version 2 Large Ensemble (CESM2-LE) simulations (50 ensemble members).
  • Detection and Attribution Model Intercomparison Project (DAMIP) single forcing experiments (anthropogenic aerosol, greenhouse gas, stratospheric ozone, natural forcing).
  • Atmospheric Model Intercomparison Project Phase 6 (AMIP6) models with prescribed observed sea surface temperatures (29 models).
  • CESM2 tropical Pacific pacemaker (PacPACE) experiments (10 ensemble members).
• Data sources:
  • Reanalysis datasets: ERA5, JRA-3Q, and MERRA-2.
  • Observed sea surface temperatures: ERSSTv5 (for PacPACE nudging).
  • Daily mean 850 hPa zonal and meridional wind speeds are used to calculate daily mean wind speed. Extreme winds are defined as exceeding the 90th percentile of the daily distribution.

Main Results

• Southern Hemisphere Summer (October to March): Low-level mean and extreme wind speeds strengthened in the midlatitudes (45°S-60°S). CMIP6 and CESM2-LE models successfully capture these trends. These trends are attributed to greenhouse gas (approximately 49.0% for mean, 47.7% for extreme winds) and stratospheric ozone (approximately 22.7% for mean, 33.3% for extreme winds) forcings.
• Southern Hemisphere Winter (April to September): Low-level extreme winds strengthened, particularly over the South Pacific. CMIP6 models show discrepancies, underestimating this strengthening. AMIP6 models (with prescribed observed sea surface temperatures) better capture the trends, reducing the discrepancies. Tropical Pacific sea surface temperature trends, through their remote impact on midlatitude low-level baroclinicity, significantly influence these extreme wind trends.
• Northern Hemisphere Summer (April to September): A robust low-level wind stilling occurred over Europe and parts of western Asia (Eurasia). CMIP6 models capture these trends, although they underestimate the magnitude for extreme winds. This stilling is attributed to aerosol (approximately 51.6% for mean, 65% for extreme winds) and greenhouse gas (approximately 47.9% for mean winds) forcings.
• Northern Hemisphere Winter (October to March): Low-level mean and extreme winds weakened over Europe and the subtropical North Pacific. CMIP6 models fail to capture these trends over both regions. Over Europe, AMIP6 models still show discrepancies. Over the subtropical North Pacific, AMIP6 models reduce the discrepancy, indicating the importance of sea surface temperature trends in shaping regional low-level baroclinicity and wind speed trends.

Contributions

• Provides a comprehensive comparison of satellite-era low-level mean and extreme wind trends from state-of-the-art reanalyses with a suite of climate model simulations.
• Quantifies and attributes the human influence (greenhouse gas, stratospheric ozone, and aerosol forcings) on summertime low-level extreme wind trends in both hemispheres.
• Identifies and investigates regional wintertime discrepancies between reanalysis and coupled climate model trends, linking them to sea surface temperature biases, particularly in the tropical Pacific, and their impact on low-level baroclinicity.
• Highlights the hemispheric asymmetry in the dominant human influences on low-level wind speed trends.

Funding

• National Oceanic and Atmospheric Administration (award NA23OAR4310597)
• National Science Foundation (NSF) (award AGS2300037)
• U.S. NSF (awards AGS-2327959 and RISE-2530555)
• Climate Variability and Change Working Group and the CESM2 Large Ensemble Community Project at the NSF National Center of Atmospheric Research
• University of Chicago’s Research Computing Center

Citation

@article{Chou2025Human,
  author = {Chou, Hsing-Hung and Shaw, Tiffany A. and Zhang, Gan},
  title = {Human influence on recent trends in extratropical low-level wind speed},
  journal = {npj Climate and Atmospheric Science},
  year = {2025},
  doi = {10.1038/s41612-025-01292-6},
  url = {https://doi.org/10.1038/s41612-025-01292-6}
}