Hu et al. (2026) New insights from the bias-corrected simulations of CMIP6 in Northern Hemisphere’s snow drought

Identification

• Journal: Communications Earth & Environment
• Year: 2026
• Date: 2026-01-13
• Authors: Yong Yu Hu, Lichun Zhu, Zican He, Fei Zhang, Boliang Dong, Yuanfang Chai
• DOI: 10.1038/s43247-026-03187-7

Research Groups

• State Key Laboratory of Loess Science, Institute of Earth Environment, Chinese Academy of Sciences
• National Observation and Research Station of Earth Critical Zone on the Loess Plateau
• Changjiang River Scientific Research Institute
• College of Geography and Environmental Sciences, Zhejiang Normal University
• China-Mozambique Belt and Road Joint Laboratory on Smart Agriculture, Zhejiang Normal University
• Guangdong Research Institute of Water Resources and Hydropower

Short Summary

This study bias-corrects CMIP6 Snow Water Equivalent (SWE) outputs to robustly project future Northern Hemisphere snow drought characteristics. It reveals a fundamental shift towards more frequent, prolonged, and severe extreme droughts under high-emission scenarios, primarily driven by reduced snowfall.

Objective

• To refine Snow Water Equivalent outputs from 29 CMIP6 models using a Cumulative Distribution Function-transform (CDF-t) bias correction method.
• To enable a robust assessment of future snow drought characteristics across the Northern Hemisphere using the Snow Water Equivalent Index (SWEI).
• To systematically analyze the spatiotemporal evolution of future snow droughts and identify their underlying physical drivers.

Study Configuration

• Spatial Scale: Northern Hemisphere land area (excluding Greenland and regions with annual cumulative SWE less than 10 mm), standardized to a 0.5° × 0.5° latitude–longitude grid, analyzed across 11 sub-regions.
• Temporal Scale: Historical period (1982–2014) and future projection period (2030–2100), focusing on the extended snow season (November–April).

Methodology and Data

• Models used: 29 CMIP6 models (e.g., GISS-E2-1-G, CESM2, NorESM2-MM), Cumulative Distribution Function-transform (CDF-t) bias correction method, multivariate linear regression for attribution.
• Data sources:
  • Observational: ERA5-Land dataset for Snow Water Equivalent (SWE) (1982–2014).
  • Simulated: CMIP6 models for monthly SWE, monthly snowmelt runoff, daily precipitation, and daily temperature under SSP126, SSP245, SSP370, and SSP585 scenarios.

Main Results

• The CDF-t bias correction significantly improved CMIP6 SWE simulations, reducing the multi-model ensemble mean absolute bias by 89.0% (from 16.3 kg m⁻² to 1.8 kg m⁻²) compared to observations.
• Under high-emission scenarios (e.g., SSP585), a fundamental structural shift in snow drought regimes is projected:
  • The frequency of the most extreme droughts (D4) increases dramatically by 63.0% (from 0.27 events/year under SSP126 to 0.44 events/year under SSP585).
  • The duration of D4 events lengthens by 54.5% compared to the historical period (from 70.1 days under SSP126 to 86.7 days under SSP585).
  • The intensity of D4 events escalates (from -2.96 under SSP126 to -3.14 under SSP585), while less severe droughts (D0, D1) may stabilize or decline in frequency and duration.
• Hotspots for increasing frequency, intensity, and duration of severe snow droughts (D4) are projected across Europe, western Asia, and central North America.
• Reduced snowfall, driven by warming-induced precipitation phase changes, is identified as the dominant mechanism for increasing snow drought frequency, contributing 78.2% to the change. Changes in snow ablation play a secondary, modulating role (15.2% contribution), constrained by the already diminished snowpack.

Contributions

• Systematically applied and validated the CDF-t bias-correction method to CMIP6 Snow Water Equivalent (SWE) outputs, substantially improving the reliability of future snow drought projections.
• Revealed a fundamental structural shift in Northern Hemisphere snow drought regimes, showing that while milder droughts may stabilize or decline, the most extreme events (D4) will become dramatically more frequent, intense, and prolonged, especially under high-emission scenarios.
• Quantitatively attributed the increase in snow drought frequency primarily to the widespread decline in snowfall due to warming-induced precipitation phase changes, clarifying the secondary role of snowmelt processes.
• Provided a more robust quantitative foundation for assessing future snow drought risks, emphasizing the critical need for urgent emission reductions and targeted adaptation strategies for water resources, agriculture, and ecosystems.

Funding

• National Natural Science Foundation of China (42407124, 42301018 & U2443223)
• National Key Research and Development Program of China (2024YFF1306900 & 2024YFE0214000)
• China Postdoctoral Science Foundation (GZC20241698)
• Open Research Funding of the Key Laboratory of Lower Yellow River Channel and Estuary Regulation of Ministry of Water Resources (LYRCER202401)

Citation

@article{Hu2026New,
  author = {Hu, Yong Yu and Zhu, Lichun and He, Zican and Zhang, Fei and Dong, Boliang and Chai, Yuanfang},
  title = {New insights from the bias-corrected simulations of CMIP6 in Northern Hemisphere’s snow drought},
  journal = {Communications Earth & Environment},
  year = {2026},
  doi = {10.1038/s43247-026-03187-7},
  url = {https://doi.org/10.1038/s43247-026-03187-7}
}