Glacier Basal Sliding in Two-Dimensions Quantified from Correlation of High-Resolution Satellite Imagery: A Case Study on Kennicott Glacier, Alaska

Abstract:

The coupling of glacial hydrology and sliding is a source of uncertainty for both ice flow modeling and prediction of future sea level rise. As basal sliding is required for a glacier to erode its bed, the spatial pattern of glacier sliding is also important for understanding alpine landscape evolution. We use multi-temporal WorldView satellite imagery (0.5 m pixel) to monitor the seasonal progression of glacier velocity across the terminal ~50 km²of Kennicott Glacier, Alaska. We employ the free image correlation software COSI-Corr to construct multiple velocity maps, using 2013 imagery with repeat times from 15 to 38 days. These short intervals between images allow us to analyze variations in glacier velocity over weekly to monthly timescales associated with hydrologically-induced basal sliding.

By assuming that spring (March-April) glacier velocity results solely from viscous deformation, we produce spatially distributed maps of glacier sliding speed by differencing summer and spring ice surface speeds. For a given time, a large portion of our study reach slides with roughly uniform speed, despite significant variation in deformation speed. This suggests that glacier flow models in which basal sliding is taken simply to scale as ice surface velocity are unfounded. The upglacier end of our study reach slides at speeds that vary through the summer, whereas the terminal reach slides at a steady speed. The proportion of glacier motion due to sliding increases dramatically moving downglacier, making basal sliding especially important in the terminal region.

Many formulations express glacier sliding as a function of effective pressure (ice pressure minus water pressure). If such formulations are correct, effective pressure varies little over large areas or is averaged over lengthscales equivalent to ~10 glacier thicknesses. Also, effective pressure is steady in the terminal region through the summer. We explore existing sliding laws to find which best describes the observed spatiotemporal pattern of sliding.