High-Resolution Force Balance Analyses of Tidewater Glacier Dynamics

Abstract:

Changes in glacier velocity, thickness, and terminus position have been used to infer the dynamic response of tidewater glaciers to environmental perturbations, yet few analyses have attempted to quantify the associated variations in the glacier force balance. Where repeat high-resolution ice thickness and velocity estimates are available, force balance time series can be constructed to investigate the redistribution of driving and resistive forces associated with changes in terminus position. Comparative force balance analyses may, therefore, help us understand the variable dynamic response observed for glaciers in close proximity to each other.

Here we construct force balance time series for Helheim Glacier, SE Greenland, and Columbia Glacier, SE Alaska, to investigate differences in dynamic sensitivity to terminus position change. The analysis relies on in situ and remotely sensed observations of ice thickness, velocity, and terminus position. Ice thickness time series are obtained from stereo satellite image-derived surface elevation and continuity-derived bed elevations that are constrained by airborne radar observations. Surface velocity time series are obtained from interferometric synthetic aperture radar (InSAR) observations. Approximately daily terminus positions are from a combination of satellite images and terrestrial time-lapse photographs. Helheim and Columbia glaciers are two of the best-studied Arctic tidewater glaciers with comprehensive high-resolution observational time series, yet we find that bed elevation uncertainties and poorly-constrained stress-coupling length estimates still hinder the analysis of spatial and temporal force balance variations. Here we use a new observationally-based method to estimate the stress-coupling length which successfully reduces noise in the derived force balance but preserves spatial variations that can be over-smoothed when estimating the stress-coupling length as a scalar function of the ice thickness. Further, we show that while systematic uncertainty in bed elevations will bias force balance estimates for each individual observational epoch (i.e., snapshot estimates), its effect mostly cancels when calculating force balance anomalies and does not prevent the analysis of dynamic change.