INVESTIGATING DYNAMIC GLACIER PROCESSES, MASS LOSS, AND COUPLED INTERACTIONS WITH THE SOLID EARTH USING SATELLITE GEODESY

Abstract

Glaciers outside of the Greenland and Antarctic ice sheets currently comprise ~1/3 of the cryosphere’s contribution to sea level rise (SLR) and are expected to remain a significant contributor during the 21st century. Glaciers represent vast reservoirs of fresh water, and their melt waters are often an important component of hydropower and agriculture for communities living in glacierized watersheds. In addition to surface melt, glaciers that terminate in the ocean (i.e., marine terminating or tidewater glaciers) can rapidly retreat and discharge large amounts ice from upper regions into the ocean. These processes are largely decoupled from climate and represent a major source of uncertainty in SLR projections. As glaciers redistribute their mass through ice discharge and melting, they evoke a deformational response of the solid Earth that can be used to better understand these processes or, conversely, probe the mechanical properties of the Earth’s interior. In this thesis, I use a variety of geodetic data and numerical models to quantify the changes to the cryosphere and the response of the sold Earth in Alaska and Iceland. I begin by investigating the processes promoting the advance of the marine terminating Yahtse Glacier in southern Alaska. Using the pixel tracking technique with satellite imagery, I construct a time series of the glacier’s velocity spanning the years 1985 – 2016. Rates of ice elevation change during years 2000 – 2014 are estimated from a time series of satellite-derived digital elevation maps (DEMs) by fitting a trend to the elevations on a pixel-by-pixel basis. We find that the development of a submarine shoal stabilizes the glacier’s terminus, causing the terminus to compressively thicken by ~6 m/yr and decelerate by ~45%. A steep (up to 35% slope) icefall prevents these stabilities from reaching the majority of the glacier. The continued influx of ice to an increasingly stable terminus promotes the glacier’s advance and highlights the important role of geometric controls in dynamic glacier processes. Using the same DEM timeseries approach, I estimate the rates of mass loss of the Juneau and Stikine icefields and Glacier Bay region in southeast Alaska between 2000-2017. The loss of 3 to 5 gigatonnes (Gt) of ice in each of the three regions drives the deformation of the solid Earth with uplift rates of up to 10 mm/yr. Measurements of this deformation response may be used to gain insight into the mass loss of these icefields or, if ice mass change is constrained, to probe the elastic structure of the Earth’s crust. Southeast Alaska’s mountainous, geologically complex setting, as well as brittle, grain-scale deformation in the uppermost crust can introduce significant biases into estimates of glacier mass change if not accounted for. Recent improvements in satellite orbital corrections open the possibility for long-wavelength glacial isostatic adjustment (GIA) to be mapped using interferometric synthetic aperture radar (InSAR) over the expansive southeast Alaska region. In this method, deformation is estimated by integrating differences in phase between repeat satellite images. However, the phase information of these images is highly affected by large quantities of atmospheric water vapor in the temperate rainforest setting of southeast Alaska. Using a stochastic estimation approach, I stack “synthetic” interferograms until the deformation signal is recovered to a pre-determined threshold of 2 mm RMSE. In order to account for 95% of 1000 trials, 90 one-year long interferograms are needed in the stack. There currently exists a sufficient number of Sentinel-1 imagery to create the equivalent number of independent interferograms, and GIA in southeast Alaska is likely to be measurable with InSAR. In the final chapter, I investigate the assumption of purely elastic deformation resulting from seasonal ice mass changes across Iceland. I model the Earth’s elastic response to seasonal changes in ice mass using seismically derived estimates of the elastic properties of the Icelandic upper mantle. The rigidity of the mantle is modified to find the best fit to the deformation observations. Reductions of at least 30% in rigidity of the upper mantle are necessary in order to best fit the data, indicating viscoelastic relaxation over seasonal time scales and in agreement with rheological models that describe viscoelastic deformation following large earthquakes in Iceland. These results suggest that glacier-induced solid Earth deformation over seasonal timescales could provide a new and valuable dataset for better understanding the viscoelastic deformation that follows large earthquakes.