Observing traveling waves on glaciers using remote sensing: Quantifying how far and how fast localized changes in stress propagate through marine-terminating glaciers

Abstract:

Marine-terminating glaciers respond dynamically to a variety of forcings with timescales ranging from hours to years. Such forcings perturb the stress fields of glaciers and include uplift from ocean tides, seasonal to multiannual migration of calving fronts in tidewater glaciers, calving of tabular icebergs from ice shelves, and melting at ice-ocean interfaces. The nonlocal response of glacier flow and surface elevation to these perturbations can provide valuable insight into the nature of the forcing and the dynamics of marine terminating glaciers, including the rheology of ice and mechanisms that facilitate slip along glacier beds. Here, we present new methods for quantifying spatiotemporal variations in ice surface velocity and elevation fields from remote sensing observations, and discuss results for Sermeq Kujalleq (Jakobshavn Isbræ), Greenland, and Pine Island Glacier (PIG), Antarctica. We introduce a flexible time-series reconstruction and decomposition method for forming continuous, time-dependent surface velocity and elevation fields from discontinuous remote sensing data collected from a variety of instruments and platforms. Reconstruction is facilitated through a sparsity-regularized least-squares regression that allows us to model time-series with complex variations as a linear combination of generic basis functions of multiple temporal scales. This approach allows us to decompose the resulting time-series into multiple frequencies. Applying this method to synthetic aperture radar data, we are able to quantify the rate and distance over which nonlocal perturbations propagate through Jakobshavn and PIG. We cast these observations in terms of wave propagation and show that on Jakobshavn, the response of ice-flow to seasonal migration of the calving front manifests as velocity variations that propagate upstream at ~400 m/day, with an attenuation length scale of ~7 km. Velocity variations due to multiannual migration of the calving front travel slower but about twice as far as seasonal variations. These results demonstrate dispersive behavior in the long-period (annual and longer) response of Jakobshavn to forcing, which has important implications for understanding its dynamics. We will discuss these implications and similar results for PIG.