Feature tracking of sub-metre resolution Capella SAR imagery to measure mountain glacier ice flow

Kurzfassung

In recent years, the number and capability of Synthetic Aperture Radar (SAR) sensors in low earth orbit has grown considerably, with multiple commercial satellites now capable of capturing sub-metre resolution imagery. In this study, we present the first application of such very fine resolution SAR imagery to measure ice velocity of a high mountain glacier. To achieve this, we apply feature tracking to a pair of Capella images in spotlight mode (0.35 m resolution) acquired in July 2021 over Baltoro Glacier in the Karakoram, Pakistan, and compare the results to ice velocities derived from feature tracking using more commonly employed TerraSAR-X Stripmap (3 m) and Sentinel-1 Interferometric Wide (IW) (5 x 20 m) imagery. We show that Capella-derived velocities reveal subtle features that are not evident in velocities derived using coarser resolution imagery. In particular, slower moving ice at the glacier margin, variations in velocity between different flow units, and lateral fluctuations reflecting the local topography are all more clearly resolved. However, the small footprint of the imagery and lack of stable ground within the frame poses a challenge for co-registration, resulting in an inherent, and spatially varying, bias of up to 0.5 m/day in the measured offsets. We reduce this error to ~0.1 m/day using a global shift and propose that the derived data then show potential to advance knowledge in a range of glaciological disciplines.