Boulder height - exposure age relationships from a global glacial 10Be compilation

Cosmogenic exposure dating of glacial boulders is commonly used to estimate the timing of past glaciations because the method enables direct dating of the duration a boulder has been exposed to cosmic rays. For successful dating, the boulders must have been fully shielded from cosmic rays prior to deposition and continuously exposed to cosmic rays ever since. A common assumption is that boulder height (the distance between the top of the boulder and the surrounding surface) is important, and that tall boulders are more likely to have been continuously exposed to cosmic rays than short boulders and therefore yield more accurate exposure ages. Here we test this assumption 'based on exposure age clustering for groups of glacial boulders (and single cobbles) Be-10 exposure ages that have recorded boulder heights (3741 boulders; 579 boulder groups with >= 3 boulders). Of the full set of boulder groups with >= 3 boulders, 21% fulfill a reduced chi square criterion (chi(2)(R) < 2) for well-clustered exposure ages. For boulder groups containing only tall boulders, the fraction of well-clustered exposure age groups is consistently larger. Moreover, this fraction of well-clustered exposure age groups increases with the minimum boulder height in each group. This result confirms the common assumption that tall boulders are generally better targets for cosmogenic exposure dating compared to short boulders. Whereas the tall boulder groups have a significantly larger fraction of well-clustered exposure age groups, there is nonetheless a dominant fraction (>50%) of the boulder groups with scattered exposure ages, highlighting the problem with prior and incomplete exposure for cosmogenic dating of glacial boulders. (C) 2016 Elsevier B.V. All rights reserved.