Vergnano et al. (2025) Integrating GPR and ice-thickness models for improved bedrock detection: the case study of Rutor temperate glacier

Identification

• Journal: ˜The œcryosphere
• Year: 2025
• Date: 2025-12-19
• Authors: Andrea Vergnano, Diego Franco, Alberto Godio
• DOI: 10.5194/tc-19-6965-2025

Research Groups

• Department of Environment, Land and Infrastructure Engineering (DIATI), Politecnico di Torino, Torino, Italy
• Department of Earth Sciences, Università degli studi di Torino, Torino, Italy

Short Summary

This study integrates Ground Penetrating Radar (GPR) data with multiple ice-thickness models to improve bedrock detection in temperate glaciers, where englacial water often causes signal scattering. Applying this combined methodology to the Rutor Glacier, the research provides a significantly revised and more accurate ice volume estimate of approximately 450 million cubic meters, about three times the previous value.

Objective

• To examine the advantages of integrating glaciological ice-thickness models into the workflow for geophysical data processing to better recognize the ice-bedrock interface in Ground Penetrating Radar (GPR) data, especially in temperate glaciers affected by englacial water content.
• To test the hypothesis that previous ice volume estimates for Rutor Glacier were underestimated due to misinterpretation of scattered GPR data.

Study Configuration

• Spatial Scale: Rutor temperate glacier (RGI ID: RGI60-11.03039) in the European Alps (Aosta Valley, Italy), covering an area of 7.5 km² (in 2021) with elevations ranging from approximately 2550 m a.s.l. to 3440 m a.s.l.
• Temporal Scale: GPR surveys conducted in May 2012 (helicopter-based) and May 2022 (ground-based). Digital Elevation Models (DEMs) from 2008 and 2021 were used for surface topography and change analysis. The study's analysis period for ice loss was 2008-2021.

Methodology and Data

• Models used:
  • OGGM (Open Global Glacier Model) v1.6.2
  • GlaTE (Glacier Thickness Estimation)
  • Original-GlabTop2 (Glacier bed Topography 2)
  • GlabTop2-Py (open-source Python implementation of GlabTop2)
• Data sources:
  • Ground Penetrating Radar (GPR) data:
    • 2012 helicopter-based survey using a GSSI 70 MHz single-frequency antenna.
    • 2022 ground-based survey using an IDS RIS ONE 40 MHz single-frequency antenna.
  • Digital Elevation Models (DEMs): 2008 (regional cartography) and 2021 (topographical survey).
  • High-resolution orthophoto (for glacier margin delineation).
  • Climate data: W5E5 dataset (for OGGM).
  • Software used for processing and visualization: QGIS, GRASS plugin (v.sample tool), RGPR, Paraview, ReflexW.

Main Results

• The previous GPR-based ice volume estimate of 150 million m³ for Rutor Glacier in 2008 was found to be unrealistic, as the glacier lost approximately 100 million m³ between 2008 and 2021.
• New GPR measurements from 2012 and 2022 confirmed widespread signal scattering due to englacial water, making reliable bedrock detection challenging and leading to potential misinterpretation of clutter zones (10–50 m depth) as the true bedrock.
• Initial unconstrained ice-thickness models provided estimates ranging from 470 million m³ (Original-GlabTop2) to 630 million m³ (OGGM), with an average of approximately 540 million m³.
• Combined 2D and 3D visualization of GPR data and model estimates significantly improved the manual selection of the ice-bedrock interface, preventing misinterpretation of scattered GPR signals.
• The final GlaTE model, constrained by the manually selected GPR bedrock interfaces, estimated the Rutor Glacier's ice volume in 2021 to be approximately 450 million m³, which is nearly three times the previous estimate.
• A sensitivity analysis demonstrated that the constrained GlaTE model was robust to variations in glaciological input parameters, showing a negligible standard deviation of about 5 million m³ in ice volume, compared to approximately 75 million m³ for the unconstrained model.
• The methodology proved effective in increasing the "pickable" regions in scattered GPR sections, particularly in areas of thickest ice, steepest bedrock slopes, and high englacial water content.

Contributions

• Introduces and validates a novel, openly available workflow that integrates Ground Penetrating Radar (GPR) data with multiple ice-thickness models to significantly improve bedrock detection and ice volume estimation in temperate glaciers, where traditional GPR interpretation is hindered by signal scattering.
• Provides a substantially revised and more accurate ice volume estimate for the Rutor Glacier (450 million m³), correcting a previous underestimation and establishing a more reliable baseline for local glaciological and hydrological studies.
• Demonstrates the critical role of model guidance in preventing misinterpretation of scattered GPR data (e.g., distinguishing englacial clutter from true bedrock) and effectively filling spatial gaps in GPR surveys, thereby enhancing the overall reliability of glacier bedrock topography maps.
• Offers a cost-effective and reproducible methodology that can inform the design of future GPR surveys, improve the calibration of regional glacier models, and deepen the understanding of ice-flux behavior in non-equilibrium states.

Funding

• "CC-Glacier lab" of the MIUR project "Department of excellence" at the Politecnico di Torino – DIATI.

Citation

@article{Vergnano2025Integrating,
  author = {Vergnano, Andrea and Franco, Diego and Godio, Alberto},
  title = {Integrating GPR and ice-thickness models for improved bedrock detection: the case study of Rutor temperate glacier},
  journal = {˜The œcryosphere},
  year = {2025},
  doi = {10.5194/tc-19-6965-2025},
  url = {https://doi.org/10.5194/tc-19-6965-2025}
}