Goelzer et al. (2025) Interactive coupling of a Greenland ice sheet model in NorESM2

Identification

• Journal: Geoscientific model development
• Year: 2025
• Date: 2025-10-27
• Authors: Heiko Goelzer, Petra M. Langebroek, Andreas Born, Stefan Hofer, Konstanze Haubner, Michele Petrini, Gunter Leguy, William H. Lipscomb, Katherine Thayer‐Calder
• DOI: 10.5194/gmd-18-7853-2025

Research Groups

• NORCE Research, Bjerknes Centre for Climate Research, Bergen, Norway
• Department of Earth Science, University of Bergen, Bjerknes Centre for Climate Research, Bergen, Norway
• School of Geographical Sciences, University of Bristol, Bristol, UK
• Department of Geosciences, University of Oslo, Oslo, Norway
• Climate and Global Dynamics Laboratory, NSF National Center for Atmospheric Research, Boulder, CO, USA

Short Summary

This study describes the first interactive coupling of the Greenland Ice Sheet model (CISM) within the Norwegian Earth System Model (NorESM2), detailing the model setup and initialization procedure. Experiments under historical and high-emission future scenarios reveal a limited impact of dynamic ice sheet changes on the global climate response compared to a fixed ice sheet simulation.

Objective

• To implement and describe the interactive coupling of the Greenland Ice Sheet (GrIS) model (CISM) into the Norwegian Earth System Model (NorESM2), including its setup and initialization procedure.
• To evaluate the impact of this dynamic coupling on global climate projections under historical and future high-emission scenarios (SSP5-8.5).

Study Configuration

• Spatial Scale:
  • Atmosphere model (NorESM2-MM): 1° × 1° horizontal resolution.
  • Atmosphere model (NorESM2-LM): 2° × 2° horizontal resolution.
  • Community Ice Sheet Model (CISM): 4 km × 4 km horizontal resolution, 11 unequally spaced vertical levels.
  • Community Land Model (CLM): 1° or 2° horizontal resolution.
  • Ocean model: Approximately 1° × 1° horizontal resolution.
  • Elevation classes (ECs) for SMB calculation: 10 classes with boundaries at 0, 200, 400, 700, 1000, 1300, 1600, 2000, 2500, 3000, and 10 000 meters.
• Temporal Scale:
  • Ice sheet model spinup: 5000 years (standalone).
  • Coupled model initialisation/relaxation (c1850): 50 years.
  • Historical experiment (cHIST): 1850–2014.
  • Future projection (cSSP585): 2015–2100.
  • Extended future projection (cSSP585Ext): 2101–2300.
  • Control experiment (cControl): 350 years.
  • Atmospheric topography updates: Every five years.
  • Meltwater runoff coupling: Hourly timescales.
  • Ice sheet calving fluxes: Annually.

Methodology and Data

• Models used:
  • Norwegian Earth System Model (NorESM2)
  • Community Ice Sheet Model (CISM) v2.1
  • Community Land Model (CLM)
  • Community Atmosphere Model (CAM)
  • Bergen Ocean Model (BLOM)
  • Hamburg Ocean Carbon Cycle model (iHAMOCC)
  • Model for Scale Adaptive River Transport (MOSART)
  • Regional Climate Model MAR v3.12 (for comparison and evaluation)
• Data sources:
  • CMIP6 simulations (NorESM2fixed) for uncoupled comparisons.
  • ERA5 reanalysis data for surface mass balance (SMB) evaluation.
  • Present-day observed ice sheet elevation (Morlighem et al., 2017) for CISM initialization target.
  • Historical forcing data for 1850–2014.
  • Shared Socioeconomic Pathway 5-8.5 (SSP5-8.5) greenhouse gas forcing for 2015–2300.
  • Reconstructions (Zuo and Oerlemans, 1997; Box and Colgan, 2013) for historical mass balance comparison.

Main Results

• The first interactive coupling of the Greenland Ice Sheet (GrIS) in NorESM was successfully implemented, including an efficient initialization procedure that minimizes model drift and aligns the ice sheet geometry with present-day observations.
• Under historical and high-emission SSP5-8.5 scenarios until 2300, the dynamic GrIS coupling has a limited impact on most global climate variables (e.g., global mean temperature, precipitation, Atlantic Meridional Overturning Circulation (AMOC), Northern Hemisphere sea-ice extent) compared to a NorESM2 simulation with a fixed ice sheet.
• A notable difference is observed in global sea surface salinity, which is reduced in the coupled model due to additional freshwater input from GrIS mass loss.
• Over the historical period (1850–2014), the GrIS shows a small mass loss (positive sea-level contribution) consistent with reconstructions, maintaining surface elevation and velocities close to observed initialisation targets.
• NorESM2's simulated mean surface mass balance (SMB) for 1960–1989 captures main features but exhibits biases, including smoothed precipitation in the interior, excessively high SMB around the margins, a cold bias in near-surface temperatures over GrIS margins, and a positive precipitation bias over southern Greenland.
• Spatially integrated SMB in NorESM2 (380 Gt/year) is higher than dynamically downscaled products (NorESM2-MAR: 284 Gt/year; ERA5-MAR: 230 Gt/year) for 1960–1989.
• Under SSP5-8.5, GrIS mass loss accelerates after 2080, leading to a cumulative sea-level contribution of nearly 1.5 meters by 2300. This acceleration is initially offset by a rapidly weakening AMOC causing regional North Atlantic cooling.
• By 2300, the ice sheet thins by over 1 kilometer around the coast and retreats by several tens of kilometers along its entire margin.

Contributions

• Presents the first interactive coupling of the Greenland Ice Sheet (GrIS) model (CISM) within the Norwegian Earth System Model (NorESM2).
• Introduces and describes a novel, efficient initialization procedure for the coupled climate-ice sheet system in NorESM2, which effectively minimizes model drift and aligns the ice sheet state with observed present-day geometry.
• Evaluates the global and regional impacts of dynamic GrIS changes on climate projections under historical and high-emission future scenarios, demonstrating that global-scale feedbacks are limited in this specific model setup and forcing.
• Contributes a new high-complexity Earth System Model configuration to the scientific community, supporting future CMIP-type climate-ice sheet simulations and the Ice Sheet Model Intercomparison Project (ISMIP7).

Funding

• Research Council of Norway (KeyClim grant no. 295046, INES grant no. 270061, GREASE grant no. 324639)
• NSF National Center for Atmospheric Research (Cooperative Agreement no. 1852977)
• Sigma2 – National Infrastructure for High Performance Computing and Data Storage in Norway (projects NN9560K, NN9252K, NN2345K, NN8006K, NS9560K, NS9252K NS2345K, NS9034K, NS8006K)

Citation

@article{Goelzer2025Interactive,
  author = {Goelzer, Heiko and Langebroek, Petra M. and Born, Andreas and Hofer, Stefan and Haubner, Konstanze and Petrini, Michele and Leguy, Gunter and Lipscomb, William H. and Thayer‐Calder, Katherine},
  title = {Interactive coupling of a Greenland ice sheet model in NorESM2},
  journal = {Geoscientific model development},
  year = {2025},
  doi = {10.5194/gmd-18-7853-2025},
  url = {https://doi.org/10.5194/gmd-18-7853-2025}
}