The source of De variability in periglacial sand wedges: Depositional processes versus measurement issues

Relict periglacial wedge structures are widespread in mid-latitude and polar regions. The wedges have a high preservation potential and are often infilled with quartz-rich aeolian sand, making them potentially suitable for luminescence dating. This paper presents initial work from an anti-syngenetic sand wedge in the Tuktoyaktuk Coastlands, Arctic Canada. When samples were measured at the single grain level they showed poor palaeodose (D-e) reproducibility and consequent high age uncertainties. The aim of the study was to determine whether this intra-sample D-e scatter reflected methodological issues arising from OSL measurement or processes of wedge development. Initial single grain dose recovery tests show that D-e scatter was not easily explained by poor recycling, sensitivity changes, variable OSL components, recuperation problems, or large D-e uncertainties from dim grains. Single grain preheating tests did show that some D-e scatter might be attributable to individual grains requiring different preheat temperatures, whilst dose recovery tests revealed that highly sensitive grains showed less scatter than dim grains. However, selection of these bright grains from natural samples still resulted in scattered D-e values. An alternative explanation for the D-e scatter relates to the formation of anti-syngenetic sand wedges by thermal contraction cracking over thousands of years in an eroding landscape, which may result in sediment of very different ages being deposited in adjacent cracks. Finite mixture modelling was used to identify D-e components within each sample. Ages calculated from these components suggest sand wedge formation at ca. 5, 8.5, 12.5 and 18 ka which correlate well with known cooling events over the last 18 ka and support a multiple phase activity model for anti-syngenetic sand wedges. (C) 2009 Elsevier B.V. All rights reserved.