Publication Date


Document Type

Honors Thesis




Glacial erosion-Computer simulation, Glaciers-Alaska, Valleys-Alaska, Glaciology, Geology, Stratigraphic-Pleistocene, Geology, Stratigraphic-Quaternary, Computer models, Basal erosion, Pleistocene, Glacier piracy, Glen-Nye equation, Quaternary geology


The Gilkey Trench penetrates the western edge of the Juneau Icefield in Southeast Alaska where it is filled by many smaller surrounding glaciers. One tributary, the Thiel Glacier, flows northeast towards the center of the Juneau Icefield before abruptly turning west to join the westward-flowing Gilkey Glacier. This intersection suggests that the Thiel Glacier could have initially flowed radially outward from the Juneau Icefield until it was later captured by the larger Gilkey Glacier. A mathematical model (Model S) is created to simulate the development of this landscape through glacial erosion. The model calculates the evolution of the longitudinal profiles of the beds of the Gilkey and Thiel glaciers and assumes that glacial erosion scales linearly with basal slip, which increases with glacier thickness and decreases with channel constriction. Since the current longitudinal bed profiles of the Gilkey and Thiel glaciers are unknown, bedrock elevations are estimated by fitting parabolic curves to valley cross sections using a Digital Elevation Model (DEM). Estimates of Pleistocene surface elevations of the Juneau Icefield are used to calculate past ice thicknesses and hence past erosion rates. These rates were scaled using a multiplier derived from the analysis of tributary glacier hang heights. The model adds bedrock progressively back to the longitudinal profiles of the Gilkey and Thiel glaciers through a series of iterative time steps, re-calculating ice thickness and basal sliding at each step, to recreate longitudinal profiles of the Gilkey Trench area at various times in the past going back 100,000 years or more. A separate balance velocity model, Model B, operates like Model S but calculates basal sliding rates from accumulation-driven ice discharge through constricted glacial channels. Model B reconstructs a pre-Wisconsinan Thiel valley as largely the same as it appears today, while Model S indicates that a drainage divide once existed halfway up the Thiel valley. Model B produces extreme ranges in erosion rates and functions poorly as a glacial erosion model, while Model S yields erosion rates that agree more closely with rates derived from fjord sedimentation studies of Alaskan glaciers. Model S is appropriate as a method of estimating glacier erosion rates for comparison with erosion rates derived from fjord sedimentation studies.




v, 106 p. : ill. (some col.) Honors project-Smith College, Northampton, Mass., 2012. Includes bibliographical references (p. 85-88)