Assessing Assumptions and Applying Cosmogenic Nuclide Methods to Earth Surface Processes

Terrestrial in situ cosmogenic radionuclides are powerful tools for quantifying rates and durations of Earth surface processes. Among the many applications of cosmogenic radionuclides are quantification of basin-scale erosion rates, burial durations of paleosols and terrace sediments, and unraveling of glacial chronologies. In this dissertation I use large data sets to investigate methodological assumptions critical for using cosmogenic radionuclides to interpret Earth surface processes, and I apply this isotope system to reconstruct the deglacial history of the Laurentide Ice Sheet in the northeastern United States. I use new and previously-published measurements of in situ 26Al and 10Be concentrations, over 300 from glacially-scoured rock surfaces and over 500 from fluvial sediments, to assess the veracity of cosmogenic radionuclide methodological assumptions. I use the glacial samples to investigate global variations in the 26Al/10Be production ratio, a critical value in dual-nuclide studies that has long been assigned a globally-constant value. I find robust evidence for a negative correlation between the 26Al/10Be production ratio and altitude, and a positive correlation between the production ratio and latitude, patterns that agree well with models of nuclide production. I use the fluvial sediment samples to investigate a key assumption in basin-wide erosion rate studies: that fluvial sediment samples experience, on average, a simple hillslope mobilization and fluvial transport history. My dual-nuclide analyses suggest that this assumption is violated in many samples from around the world, implying that erosion rates may be overestimated for these basins. I use regressions between the apparent complexity of sediment transport history and basin parameters to investigate if systematic patterns exist that explain the occurrence of non-simple sediment histories. I find that steep and/or small catchments are correlated with simple sediment transport histories, while low-slope and/or large catchments are correlated with more complex apparent transport histories. I use 10Be and other deglacial chronometers to constrain the patterns of Laurentide Ice Sheet margin retreat and thinning in the northeastern United States following the Last Glacial Maximum. Patterns of deglaciation implied by all chronometers agree at locations more than 150 km away from the maximum ice margin. Near the maximum ice margin, 10Be and organic 14C ages disagree by thousands of years, likely due to 10Be inheritance producing ‘too-old’ ages and delayed re-vegetation causing ‘too-young’ organic 14C ages. Based on these lines of evidence and independent glacial varve chronologies, I constrain the initial timing of deglaciation to between 24 and 20 thousand years ago. I then use over 80 new and 100 previously-published 10Be ages sampled along vertical transects of mountains to constrain the ice thinning history of the Laurentide Ice Sheet. I find evidence for ice thinning by at least 19 ka, with ~600 m of ice thinning between 17 and 15 thousand years ago, coincident with the slow initial ice margin retreat indicated by varve records. Further up-ice (\u3e400 km north of the ice sheet maximum extent), exposure ages over ~1000 m elevation spans are between 15 and 13 ka, suggesting rapid and extensive thinning coincident with a period of abrupt warming in the North Atlantic