Verjans et al. (2025) The Greenland Ice Sheet Large Ensemble (GrISLENS): simulating the future of Greenland under climate variability

Identification

• Journal: ˜The œcryosphere
• Year: 2025
• Date: 2025-09-12
• Authors: Vincent Verjans, Alexander A. Robel, Lizz Ultee, Hélène Seroussi, Andrew F. Thompson, Lars Ackermann, Youngmin Choi, Uta Krebs‐Kanzow
• DOI: 10.5194/tc-19-3749-2025

Research Groups

• Barcelona Supercomputing Center, Barcelona, Spain
• School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, GA, USA
• NASA Goddard Space Flight Center, Greenbelt, MD, USA
• Thayer School of Engineering, Dartmouth College, Hanover, NH, USA
• Environmental Science & Engineering, California Institute of Technology, Pasadena, CA, USA
• Alfred Wegener Institute Helmholtz Centre for Polar and Marine Research, Bremerhaven, Germany
• Earth System Science Interdisciplinary Center, University of Maryland, College Park, MD, USA

Short Summary

This study introduces GrISLENS, the first large-ensemble ice sheet model resolving individual glaciers and calibrated to observations, to quantify the impact of internal climate variability on Greenland Ice Sheet evolution. It finds that internal climate variability significantly contributes to ice sheet mass change uncertainty on decadal timescales, but its relative importance diminishes on longer timescales compared to anthropogenic forcing.

Objective

• To quantify the role of internal climate variability in 185-year projections of the Greenland Ice Sheet (GrIS) evolution, including ice mass change, thickness change, and glacier retreat, under both high-emission (RCP8.5) and pre-2000 climate scenarios, while resolving individual outlet glaciers and calibrating climate variability to observations.

Study Configuration

• Spatial Scale: Variable horizontal resolution ranging from 25 kilometers in the ice sheet interior to less than 1 kilometer (800 meters) at dynamic calving fronts. Atmospheric fields are downscaled to 5 kilometers horizontal resolution.
• Temporal Scale: Transient simulations span 185 years (2018 to 2203), following an 11-year calibration period (2007–2017). Climate model forcing data covers 1850–2203.

Methodology and Data

• Models used:
  • Greenland Ice Sheet Large Ensemble (GrISLENS) – overarching modeling experiment.
  • Stochastic Ice-Sheet and Sea-Level System Model (StISSM) – ice sheet model.
  • Alfred Wegener Institute Earth system model (AWI-ESM) – coupled atmosphere-ocean model (reference climate simulation).
  • FESOM1.4 – ocean component of AWI-ESM.
  • ECHAM6 – atmosphere component of AWI-ESM.
  • Diurnal Energy Balance Model (dEBM) – for atmospheric downscaling and surface mass balance (SMB) calculation.
  • Autoregressive moving-average (ARMA) models – for stochastic representation of climatic variables.
  • Graphical lasso method – for computing sparse correlation matrices between climatic variables and catchments.
  • Von Mises calving parameterization – for dynamic calving front migration.
  • Budd sliding law – for basal friction.
• Data sources:
  • EN4 ocean monthly objective analyses – for bias correction of ocean thermal forcing (TF).
  • BedMachine v4 – for bed topography, ice thickness, and ice mask.
  • 2007 ice velocity field from Joughin et al. (2010).
  • Geothermal heat flux from Shapiro and Ritzwoller (2004).
  • Surface temperature from Ettema et al. (2009).
  • Mouginot et al. (2017) – 253 catchment delimitations.
  • Wood et al. (2021) – observed terminus retreats and glacier front locations.
  • ECCO2-Arctic – high-resolution ocean model reanalysis for downscaling TF to fjords.
  • Ice Sheet Mass Balance Intercomparison Exercise (IMBIE) (Otosaka et al., 2023) – for total GrIS mass loss (2007–2017) calibration.
  • RCP8.5 emissions scenario (Riahi et al., 2007) – for future climate forcing.

Main Results

• Internal climate variability significantly contributes to uncertainty in total GrIS mass change during the first 20–30 years of simulations, representing 10 %–300 % of the ensemble mean mass loss, which is crucial for decadal coastal planning.
• Beyond 2050, the relative importance of internal climate variability diminishes, accounting for a small fraction of the total ice sheet change, as anthropogenic emissions and model uncertainties become dominant.
• At the ice sheet scale, uncertainty in ice loss is primarily driven by surface mass balance (SMB) variability and its spatial correlations.
• Ocean variability has a relatively smaller influence on total ice sheet uncertainty but is critical within individual catchments, driving uncertainty in the timing of rapid glacier retreats (e.g., Petermann Glacier, Zachariae Isstrøm).
• The GrISLENS ensemble spread is 1 to 2 orders of magnitude smaller than previous studies using coarse models, indicating that resolving small-scale features in climate forcing and ice sheet dynamics substantially improves the quantification of internal variability.
• Under a high-emission (RCP8.5) scenario, SMB variability alone can explain the full-ensemble spread, with approximately one-third of the ensemble spread by 2203 attributed to the warming-driven increase in SMB variability amplitude.
• The pre-2000 control ensemble exhibits a 4 % greater mean ice loss than its deterministic counterpart, suggesting a noise-induced drift, likely caused by ocean variability driving bifurcations in glacier retreat.

Contributions

• Presents the first large-ensemble study of Greenland Ice Sheet evolution that resolves individual outlet glaciers and calibrates climate variability to observations, providing an advanced quantification of internal climate variability's role.
• Quantifies the relative importance of internal climate variability compared to forced trends and structural/parametric uncertainties in GrIS projections, particularly highlighting its significance on decadal timescales.
• Introduces a novel stochastic modeling approach that estimates GrIS sensitivity to internal climate variability down to sub-kilometer scales, incorporating calibrated spatiotemporal stochastic models for downscaling climate outputs.
• Provides an openly available, high-resolution model output dataset (GrISLENS) to the research community, enabling further investigations into the role of internal climate variability in driving ice sheet change.
• Emphasizes the critical role of initializing ice sheet models to match recently observed ice sheet behavior, which could reduce inter-model spread in future projections.
• Identifies the need for more robust initialization methods and suggests extending these large-ensemble methods to the Antarctic ice sheet.

Funding

• Heising-Simons Foundation (grant no. 2020-1965)
• Novo Nordisk Fonden (grant no. NNF23OC00807040)
• Deutsche Forschungsgemeinschaft (grant no. 390741603)
• Bundesministerium für Bildung und Forschung, BonaRes (grant no. 01LP2313A)

Citation

@article{Verjans2025Greenland,
  author = {Verjans, Vincent and Robel, Alexander A. and Ultee, Lizz and Seroussi, Hélène and Thompson, Andrew F. and Ackermann, Lars and Choi, Youngmin and Krebs‐Kanzow, Uta},
  title = {The Greenland Ice Sheet Large Ensemble (GrISLENS): simulating the future of Greenland under climate variability},
  journal = {˜The œcryosphere},
  year = {2025},
  doi = {10.5194/tc-19-3749-2025},
  url = {https://doi.org/10.5194/tc-19-3749-2025}
}