Pellicciotti et al. (2026) DCG-MIP: the Debris-Covered Glacier melt Model Intercomparison exPeriment

Identification

• Journal: ˜The œcryosphere
• Year: 2026
• Date: 2026-04-02
• Authors: Francesca Pellicciotti, Adrià Fontrodona-Bach, David R. Rounce, Catriona L. Fyffe, Leif S. Anderson, Álvaro Ayala, Ben Brock, Pascal Buri, Stefan Fugger, Koji Fujita, Prateek Gantayat, Alexander Raphael Groos, Walter Immerzeel, Marin Kneib, Christoph Mayer, Shelley MacDonell, Michael McCarthy, James McPhee, Evan Miles, Heather Purdie, Ekaterina Rets, Akiko Sakai, T. E. Shaw, Jakob Steiner, Patrick Wagnon, Alex Winter-Billington
• DOI: 10.5194/tc-20-1895-2026

Research Groups

• Institute of Science and Technology Austria (ISTA)
• Swiss Federal Institute for Forest, Snow and Landscape Research WSL
• Carnegie Mellon University, Department of Civil and Environmental Engineering
• Northumbria University, School of Geography and Natural Sciences
• University of Utah, Department of Geology and Geophysics
• Centro de Estudios Avanzados en Zonas Áridas (CEAZA)
• University of Alaska Fairbanks, Geophysical Institute
• University of Fribourg, Department of Geosciences
• Nagoya University, Graduate School of Environmental Studies
• Lancaster University, Lancaster Environment Center
• University of Bern, Institute of Geography
• Friedrich-Alexander-Universität Erlangen-Nürnberg, Institute of Geography
• Utrecht University, Department of Physical Geography
• ETH Zurich, Laboratory of Hydraulics, Hydrology and Glaciology (VAW)
• Bavarian Academy of Sciences and Humanities, Geodesy and Glaciology
• University of Canterbury, Waterways Centre & School of Earth & Environment
• University of Chile, Department of Civil Engineering & Advanced Mining Technology Center
• University of Zurich, Glaciology and Geomorphodynamics Group
• Polish Academy of Sciences, Institute of Geophysics
• Himalayan University Consortium
• University of Graz, Institute of Geography and Regional Science
• Univ. Grenoble Alpes, CNRS, IRD, IGE
• Victoria University of Wellington, Te Puna Pātiotio Antarctic Research Centre
• International Association of Cryospheric Sciences (IACS) Debris Covered Glacier Working Group

Short Summary

This study intercompared 15 debris-covered glacier melt models across nine global sites to assess their performance in simulating ice melt under debris. It found that models with higher complexity at the atmosphere-debris interface perform best, but identified critical data gaps, particularly regarding debris thermal properties, which hinder accurate global modeling and necessitate further model development and standardized data collection.

Objective

• To assess the performance of various debris-covered glacier melt models across different sites globally.
• To identify the strengths and limitations of different model categories (energy balance, simplified energy balance, enhanced temperature-index, and temperature-index models).
• To advance understanding of the impact of model choice on the accuracy and uncertainty of simulated melt.
• To attribute differences in model performance to model physics, underlying assumptions, and/or inaccuracies in input data.

Study Configuration

• Spatial Scale: Point scale, at automatic weather station locations on nine debris-covered glaciers in the European Alps (Arolla, Miage, Suldenferner), Caucasus (Djankuat), Chilean Andes (Pirámide, Tapado), Nepalese Himalaya (Changri Nup, Lirung), and Southern Alps of New Zealand (Tasman).
• Temporal Scale: One melt season (variable duration per site), with meteorological input data at hourly resolution and model evaluation against daily melt and surface temperature.

Methodology and Data

• Models used: 15 models categorized into:
  • Eight energy balance models (e.g., DEB CF, ROU15, d2EB, GRO17 A/B, THRED, A-Melt).
  • One simplified energy balance model (MCC19).
  • Two enhanced temperature-index models (DETI m, KO2).
  • Four temperature-index models (Hyper-fit, KM1, KP1, DDF debris).
• Data sources:
  • On-site Automatic Weather Stations (AWS) providing hourly meteorological data: air temperature (°C), relative humidity (%), wind speed (m s⁻¹), air pressure (hPa), incoming/outgoing shortwave and longwave radiation (W m⁻²), precipitation (mm h⁻¹), snow depth (cm), and sensor height (m).
  • Debris properties: thickness (cm), surface roughness length (m), thermal conductivity (W m⁻¹ K⁻¹), porosity (dimensionless), and emissivity (dimensionless), which were either measured, derived, assumed, or optimized.
  • Validation data: on-site measurements of sub-debris melt (from ablation stakes, ultrasonic depth gauges, or draw-wires) and debris surface temperature (derived from outgoing longwave radiation).
  • All data and model outputs are publicly accessible on the Zenodo community for debris-covered glaciers.

Main Results

• Model performance varied considerably across sites, with high accuracy and consistency at European Alps sites (Arolla, Suldenferner) but consistently poor performance at Djankuat, Lirung, and Tasman.
• Physically-based energy balance models and empirical temperature-index models exhibited distinct performance characteristics. Calibrated temperature-index models showed high accuracy, sometimes surpassing energy balance models, but performed poorly when uncalibrated.
• The most accurate energy balance models were those incorporating the highest degree of complexity in representing processes at the atmosphere-debris interface (e.g., DEB CF, ROU15, d2EB).
• A significant data gap identified was the poor knowledge and ambiguous derivation of debris properties, particularly thermal conductivity, which was a primary cause of poor model performance at several sites.
• Model sensitivity to parameter uncertainty (typically around 5%) was generally smaller than the overall model error, suggesting that larger uncertainties stem from data quality or model structural limitations.
• Temperature-index models offered substantial computational efficiency (10³–10⁴ times faster than energy balance models) and, when calibrated, provided satisfactory daily melt estimates, though they could not accurately reproduce sub-daily melt cycles.
• Models of intermediate complexity (simplified energy balance and enhanced temperature-index models) performed comparably to the best energy balance models and could reproduce sub-daily melt variability.
• The treatment of turbulent fluxes and assumptions regarding debris surface relative humidity were critical factors distinguishing the performance of energy balance models.

Contributions

• Conducted the first systematic intercomparison of 15 diverse debris-covered glacier melt models across nine globally distributed sites, providing a robust assessment of current modeling capabilities.
• Identified specific strengths and limitations of different model categories, clarifying their applicability based on data availability and research objectives.
• Quantified the impact of model choice on melt simulation accuracy and uncertainty, highlighting the importance of calibration for empirical models and detailed atmospheric-debris interface physics for energy balance models.
• Attributed significant model performance differences to the quality and availability of debris properties, particularly thermal conductivity, underscoring a critical knowledge gap.
• Provided concrete recommendations for future research, including the need for consistent and standardized field measurements of debris properties, internal validation data for model components, and the development of modular model frameworks to systematically test process representations (e.g., debris-snow interactions, moisture content, refreezing, and convection).

Funding

• European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement No 772751, RAVEN, "Rapid mass losses of debris covered glaciers in High Mountain Asia").
• SNSF RENOIR project ("Resolving the thickness of debris on Earth's glaciers and its rate of change (RENOIR)"), project number 204322.
• NASA-ROSES program grants NNX17AB27G and 80NSSC17K0566.
• European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (grant agreement no. 676819).
• EU/FP7 ACQWA (Assessing Climate impacts on the Quantity and quality of WAter) project.
• NERC grant NE/C514282/1.
• British Council-Italian Ministry of University and Research Partnership programme.
• Carnegie Trust for the Universities of Scotland.

Citation

@article{Pellicciotti2026DCGMIP,
  author = {Pellicciotti, Francesca and Fontrodona-Bach, Adrià and Rounce, David R. and Fyffe, Catriona L. and Anderson, Leif S. and Ayala, Álvaro and Brock, Ben and Buri, Pascal and Fugger, Stefan and Fujita, Koji and Gantayat, Prateek and Groos, Alexander Raphael and Immerzeel, Walter and Kneib, Marin and Mayer, Christoph and MacDonell, Shelley and McCarthy, Michael and McPhee, James and Miles, Evan and Purdie, Heather and Rets, Ekaterina and Sakai, Akiko and Shaw, T. E. and Steiner, Jakob and Wagnon, Patrick and Winter-Billington, Alex},
  title = {DCG-MIP: the Debris-Covered Glacier melt Model Intercomparison exPeriment},
  journal = {˜The œcryosphere},
  year = {2026},
  doi = {10.5194/tc-20-1895-2026},
  url = {https://doi.org/10.5194/tc-20-1895-2026}
}