Constraining Quaternary ice covers and erosion rates using cosmogenic 26Al/10Be nuclide concentrations

Paired cosmogenic nuclides are often used to constrain the exposure/burial history of landforms repeatedly covered by ice during the Quaternary, including tors, high-elevation surfaces, and steep alpine summits in the circum-Arctic regions. The approach generally exploits the different production rates and half-lives of Be-10 and Al-26 to infer past exposure/burial histories. However, the two-stage minimum limiting exposure and burial model regularly used to interpret the nuclides ignores the effect of variable erosion rates, which potentially may bias the interpretation. In this study, we use a Monte Carlo model approach to investigate systematically how the exposure/burial and erosion history, including variable erosion and the timing of erosion events, influence concentrations of Be-10 and Al-26. The results show that low Al-26/Be-10 ratios are not uniquely associated with prolonged burial under ice, but may as well reflect ice covers that were limited to the coldest part of the late Pleistocene combined with recent exhumation of the sample, e.g. due to glacial plucking during the last glacial period. As an example, we simulate published Al-26/Be-10 data from Svalbard and show that it is postible that the steep alpine summits experienced ice-free conditions during large parts of the late Pleistocene and varying amounts of glacial erosion. This scenario, which contrasts with the original interpretation of more-or-less continuous burial under non-erosive ice over the last similar to 1 Myr, thus challenge the conventional interpretation of such data. On the other hand, high Al-26/Be-10 ratios do not necessarily reflect limited burial under ice, which is the common interpretation of high ratios. In fact, high Al-26/Be-10 ratios may also reflect extensive burial under ice, combined with a change from burial under erosive ice, which brought the sample close to the surface, to burial under non-erosive ice at some point during the mid-Pleistocene. Importantly, by allowing for variable erosion rates, the model results may reconcile spatially varying Al-26/Be-10 data from bedrock surfaces preserved over multiple glacial cycles, suggesting that samples from the same high elevation surface or neighbouring alpine summits may have experienced similar long-term burial under ice, but varying amounts of glacial erosion. (C) 2017 The Authors. Published by Elsevier Ltd.