Estimating the ages of glacial landforms from the statistical distributions of cosmogenic exposure dates

Glacial landforms, especially moraines, have long been used as indicators of decreased temperature or increased precipitation in the past. Cosmogenic exposure dating of moraine boulders provides a method for estimating moraine ages. However, geomorphic processes interfere with cosmogenic exposure dating. To improve the accuracy of the cosmogenic exposure dating method, quantitative methods for assessing the effects of geomorphic processes on cosmogenic exposure dating are needed.;To address this need, this dissertation describes models of two geomorphic processes and their effects on the cosmogenic exposure dating of moraines. These processes are moraine degradation and inheritance. Both models use Monte Carlo techniques to estimate the statistical distributions of exposure dates from moraine boulders, given specific assumptions about the histories of the boulders. The moraine degradation model is based on prior examples from the literature; the inheritance model is novel.;Some implications of this work for cosmogenic exposure dating of moraines follow. Different geomorphic processes give rise to different statistical distributions of cosmogenic exposure dates. Moraine degradation produces distributions that are skewed toward the young tail of the distribution, whereas inheritance produces distributions that are skewed toward the old tail of the distribution.;Simple procedures for estimating moraine ages from cosmogenic exposure dates perform well in some cases and poorly in others, sometimes producing moraine age estimates that are incorrect by thousands of years. Simple estimators tested here include the mean, the mean after discarding outliers, the youngest date, and the oldest date.;Explicit inversion of the models against collections of cosmogenic exposure dates may represent an improvement over simple methods of estimating moraine ages. These inverse iv methods yield estimates of the rates and magnitudes of geomorphic processes acting in glaciated basins, as well as the moraines' ages.;Field sampling criteria that preferentially choose pristine boulders may yield exposure dates that underestimate the ages of moraines by thousands of years. This statement assumes that the moraines lose several meters of material from their crests over their lifetimes, that the boulders erode at a constant rate after being exhumed, and that inherited nuclides are not present in the boulders.