Cooper et al. (2025) Greenland ice sheet runoff reduced by meltwater refreezing in bare ice

Identification

• Journal: Nature Communications
• Year: 2025
• Date: 2025-09-12
• Authors: Matthew G. Cooper, L. C. Smith, Å. K. Rennermalm, Jonathan C. Ryan, L. H. Pitcher, Glen E. Liston, C. Miège, Sarah Cooley, Dirk van As
• DOI: 10.1038/s41467-025-62281-0

Research Groups

• Department of Geography, University of California, Los Angeles, CA, USA
• Sierra Crest Analytics, Portland, OR, USA
• Institute at Brown for Environment and Society, Brown University, Providence, RI, USA
• Department of Earth, Environmental and Planetary Sciences, Brown University, Providence, RI, USA
• Department of Geography, Rutgers, The State University of New Jersey, New Brunswick, NJ, USA
• Division of Earth and Climate Sciences, Nicholas School of the Environment, Duke University, Durham, NC, USA
• Cooperative Institute for Research in the Atmosphere, Colorado State University, Fort Collins, CO, USA
• Department of Geography, University of Utah, Salt Lake City, UT, USA
• Geological Survey of Denmark and Greenland, Copenhagen, Denmark

Short Summary

This study combines field measurements and numerical modeling to demonstrate that extensive meltwater retention and refreezing occur in bare glacier ice, significantly reducing runoff from the Greenland Ice Sheet's ablation zone. This overlooked process explains why current climate models overestimate runoff by 9–15% in southwest Greenland, impacting sea-level rise projections.

Objective

• To formally test the hypothesis that meltwater retention and refreezing occur in bare glacier ice, explaining the observed overestimation of runoff by current climate models.

Study Configuration

• Spatial Scale: Field experiments conducted at the RB catchment (60–63 km²) in southwest Greenland (1200–1360 meters above sea level). Regional simulations covered all bare ice areas of the Greenland Ice Sheet's southwest sector (61,975 km²) at 5 km horizontal resolution.
• Temporal Scale: Field campaigns in July 2015 and 6–13 July 2016. Regional simulations for the decade 2009–2018, with specific refreezing estimates for 15 June–15 August and 1 June–31 August.

Methodology and Data

• Models used:
  • SkinModel: A process-based numerical model emulating traditional climate model surface energy balance assumptions (no subsurface melt/refreezing).
  • IceModel: A process-based numerical model simulating spectral radiation and thermodynamic heat transfer in glacier ice, including subsurface melt and refreezing.
  • Regional Climate Models (RCMs): MAR3.11, RACMO2.3p3, RACMO2.3p2.
  • Global Climate Reanalysis: MERRA-2.
• Data sources:
  • Field measurements: Hourly supraglacial river discharge (Acoustic Doppler Current Profiler - ADCP), three-hourly ice surface lowering (ablation stakes), hourly ice surface elevation change (pressure transducer at KANM AWS), shallow ice cores (ice density, porosity, liquid water saturation).
  • Automatic Weather Station (AWS): PROMICE/GAP KANM (near-surface air temperature, wind speed, relative humidity, air pressure, albedo, incoming shortwave and longwave radiation).
  • Satellite data: WorldView-1 and WorldView-2 (catchment topography, surface classification), MODIS (daily ice sheet surface albedo, bare-ice extent).
  • Publicly available datasets: PROMICE Greenland Liquid Water Discharge dataset, ArcticDEM v7, BedMachine v3.

Main Results

• Meltwater refreezing in bare, porous glacier ice reduced runoff by an estimated 11–17 gigatonnes per year (Gt a⁻¹) in southwest Greenland alone from 2009 to 2018. This is equivalent to 9–15% of the annual meltwater runoff simulated by climate models for this sector.
• Climate models overpredicted observed meltwater runoff during the July 2016 field experiment by 7% (MERRA-2) to 58% (RACMO2.3p3).
• The developed IceModel, which accounts for subsurface melt and refreezing, closely reproduced observed cumulative runoff within catchment boundary uncertainty.
• IceModel predicted 32% less runoff than SkinModel (which emulates climate model assumptions) at the Leverett Glacier (LG) catchment and reproduced cumulative proglacial runoff to within 6% at the Akuliarusiarsuup Kuua River's northern tributary (AK4) catchment.
• Climate models overpredicted supraglacial lake volume infilling by 58–81%, while IceModel closely tracked observations.
• Across six independent sites, IceModel (forced with MODIS albedo) exhibited the lowest mean bias (-2% ± 18%), whereas climate models predicted +9% ± 46% to +47% ± 32% higher runoff than observations.
• Nocturnal refreezing of subsurface liquid meltwater in bare ice was observed at rates approaching 1 millimeter per hour (mm h⁻¹).

Contributions

• Identifies and quantifies nocturnal refreezing of meltwater in bare glacier ice as a significant, previously overlooked process in Greenland, explaining the consistent overestimation of bare ice runoff by current climate models.
• Provides the first empirical verification of bare-ice refreezing in Greenland, a process previously observed only on mountain glaciers.
• Demonstrates that this refreezing constitutes a substantial energy sink and mass retention mechanism, buffering projected runoff increases and potentially exceeding the impact of refreezing in snow and firn.
• Highlights an urgent need to incorporate bare-ice retention and refreezing processes into future climate models to improve forecasts of ice sheet runoff and its contribution to global sea-level rise.

Funding

• NASA Cryospheric Science Program (grants NNX14AH93G, 80NSSC19K0942, 80NSSC25K7960)
• NASA Earth and Space Sciences Fellowship Program (grant 80NSSC17K0374)

Citation

@article{Cooper2025Greenland,
  author = {Cooper, Matthew G. and Smith, L. C. and Rennermalm, Å. K. and Ryan, Jonathan C. and Pitcher, L. H. and Liston, Glen E. and Miège, C. and Cooley, Sarah and As, Dirk van},
  title = {Greenland ice sheet runoff reduced by meltwater refreezing in bare ice},
  journal = {Nature Communications},
  year = {2025},
  doi = {10.1038/s41467-025-62281-0},
  url = {https://doi.org/10.1038/s41467-025-62281-0}
}