A model for Holocene glacial erosion at Peyto Glacier, Alberta.

Glaciers play an important role in the evolution of many mountain landscapes. The primary goal of this study is to model rates of contemporary glacier erosion through numerical modeling. The first part of this study uses a regional glaciation model and couples to it (i) a process-driven erosion model that includes abrasion and quarrying and (ii) a sliding-based power relation erosion model as used in landscape evolution models. A sensitivity analysis reveals that abrasion is sensitive to the grain-size distribution, the choice of sliding parameters, and the debris hardness/concentration parameter. Quarrying rates primarily depend on the height and spacing of bed irregularities. The sliding model is sensitive to a proportionality constant which is not physically constrained. The second part of the study uses the developed erosion models to estimate glacier erosion for Peyto Glacier. The models predict higher rates of primary glacier erosion than those estimated by sediment recovery from Peyto Lake. Contrary to observations of sediment yield from the catchment, both models predict a decrease in erosion rates during glacial retreat. Changes in sediment storage beneath the glacier are believed to account for the discrepancy between observed and modeled sediment yield. Over the past 2000 years, major differences are seen between the models during extreme climate periods such as the Medieval climate anomoly (MCA) and the Little Ice Age (LIA). These differences do not depend on the exact choice of modeling parameters but rather the inter-decadal to century-scale variability inherent in the climate proxy data used to force the models. --Leaf i.