Item: EARLY DETECTION OF SNOWPACK WETTING USING SENTINEL-1 DATA: LINKING SEASONAL SNOWPACK DYNAMICS AND GLACIER MELT UNDER CLOUD-COVERED CONDITIONS ON THE BRØGGER PENINSULA, SVALBARD (2022–2025)
-
-
Title: EARLY DETECTION OF SNOWPACK WETTING USING SENTINEL-1 DATA: LINKING SEASONAL SNOWPACK DYNAMICS AND GLACIER MELT UNDER CLOUD-COVERED CONDITIONS ON THE BRØGGER PENINSULA, SVALBARD (2022–2025)
Proceedings: Proceedings, International Snow Science Workshop, Whistler, BC, Canada, 2026
Authors:
- Jean-Michel Friedt [ Université Marie et Louis Pasteur, CNRS, Institut FEMTO-ST (UMR 6174), Besançon, France ]
- Eric Bernard [ Université Marie et Louis Pasteur, CNRS, Théma (UMR 6049), Besançon, France ]
Date: 2026-09-28
Abstract: Seasonal snow cover plays a fundamental role in controlling the timing, spatial distribution and intensity of glacier melt in the High Arctic. In Svalbard, winter warming, rain-on-snow events and earlier spring melt modify snowpack structure and promote transitions between dry snow, wet snow, refreezing and runoff. Observing these transitions at the glacier scale remains challenging because optical satellite observations are frequently affected by cloud cover and limited illumination. Synthetic Aperture Radar (SAR), and particularly Sentinel-1 C-band observations, provides an alternative because radar acquisitions are independent of cloud cover and solar illumination and are sensitive to changes in the dielectric properties of snow associated with liquid water. This study investigates the spatial and temporal dynamics of snowpack wetting across small glaciers of the Brøgger Peninsula, western Spitsbergen, using Sentinel-1 time series acquired between 2022 and 2025. Rather than attempting a direct inversion of snow water content, we use a relative, hypsometrically normalized approach designed to characterize the timing and vertical extent of snowpack transitions. Backscatter profiles are extracted along longitudinal glacier flowlines and transformed into a normalized elevation coordinate. Deviations from a winter reference state are used to identify statistically significant reductions in backscatter associated with the transition from cold, relatively dry snow towards wetter conditions. The resulting time–altitude information is summarized using a normalized hypsometric index, allowing comparison among glaciers with different elevation ranges. The Sentinel-1 observations are interpreted jointly with meteorological observations from Ny-Ålesund, including air temperature, precipitation and snow-cover information. Ground-based automated photography from the Austre Lovén glacier basin provides an independent local reference for interpreting major snowline and surface-state transitions. Across the 19 analysed glaciers, the radar time series reveal a coherent regional seasonal signal but also substantial differences in the timing, duration and vertical development of wet-snow conditions. The mean annual duration of detected wet-snow conditions is approximately 78 days over the 2022–2025 period. The onset of widespread wetting generally occurs after sustained positive-temperature forcing, while warm and wet periods associated with rain-on-snow conditions produce pronounced increases in the vertical extent of the radar-derived wet-snow signal. The results demonstrate the potential of Sentinel-1 time series for regionalizing observations of snowpack state across Arctic glaciers while emphasizing that radar backscatter remains an indirect proxy of snow wetness. The proposed framework therefore provides a spatially distributed diagnostic tool rather than a direct measurement of liquid-water content or glacier melt.
Object ID: ISSW2026_P3.25.pdf
DOI: https://doi.org/10.15788/1790099136
Language of Article: English
Presenter(s): Eric Bernard
Keywords: Sentinel-1, SAR, wet snow, snowpack, glacier melt, Svalbard
Page Number(s): 1130 - 1134
-