We’re excited to announce the publication of the first peer-reviewed scientific paper produced within the LIQUIDICE project. Titled “Glacier internal structure revealed by automatic image processing‑powered classification of radar images,” the study by Kamil Kachniarz, Mariusz Grabiec, Krzysztof Wróbel & Dariusz Ignatiuk, from the University of Silesia – LIQUIDICE partner, appears in the July 2025 issue of Applied Geomatics. This publication represents a significant step forward in understanding the internal dynamics of Arctic glaciers using cutting-edge image processing techniques.
The study focuses on Hansbreen, a polythermal glacier located in southern Spitsbergen, and draws on data collected from ground-penetrating radar (GPR) surveys conducted between 2007 and 2021. By applying a newly developed, fully automated classification algorithm, the authors were able to identify and track changes in the glacier’s internal thermal structure with a high degree of accuracy. The method—based on local image binarization and morphological filtering—eliminates the need for manual layer delineation, drastically improving efficiency while maintaining results comparable to expert interpretation.
Importantly, the paper reveals a striking transformation in Hansbreen’s internal composition over the 14-year period. Where the glacier once showed a two-layered structure—cold, water-free ice (W-FI) overlaying a temperate, water-saturated layer (W-STI)—the upper cold layer has thinned significantly and had largely disappeared by 2021. This shift appears to be driven by rising air temperatures, increased snow insulation, and the growing complexity of internal meltwater drainage systems.
The automated method was validated against a borehole temperature profile from 1997, confirming the boundary between cold and temperate ice with a high degree of precision. The resulting analysis offers new possibilities for large-scale, long-term monitoring of polythermal glaciers and their response to climate change.
This publication marks a major achievement for LIQUIDICE, demonstrating how advanced computational techniques can be combined with field data to deliver timely, repeatable, and robust insights into cryospheric processes. As the project progresses, this approach could be extended to other glacier systems, offering a powerful tool for tracking the evolving dynamics of ice masses in a warming world.
We encourage you to read the full open-access paper here: https://link.springer.com/article/10.1007/s12518-025-00635-5