An empirical study on the variability of luminescence ages for coeval loess samples

Abstract

An increasing number of studies exploit the advantages of a single-aliquot regenerative-dose protocol (SAR) for equivalent dose determination with sampling at relatively closely-spaced vertical intervals (of the order of 10-30 cm). The resulting ages are, at least in principal, ideally suited for age-depth modelling. The modelling, however, is made difficult owing to the variety and complex combination of uncertainties associated with luminescence dating. Moreover, we previously reported on a variability in age results for coeval loess samples that is significantly larger than expected and remains to be understood.In this study, we examine this problem explicitly by observing the degree of precision and accuracy that can be achieved by luminescence dating of multiple coeval loess samples of known age. The main goal is to improve our understanding of how luminescence ages are to be incorporated into age-depth models, thus increasing their robustness and accuracy.Fourteen samples were taken at closely-spaced horizontal intervals from loess deposits immediately over- and underlying the Campanian Ignimbrite/Y5 tephra layer (40Ar/39Ar dated to 39.2±0.1 ka), as exposed at a section in the Lower Danube Basin in southeastern Europe. Luminescence analyses were carried out using the single-aliquot regenerative-dose (SAR) protocol and OSL signals from 63-90 µm quartz fraction. We report an average age of 46 ka for the samples collected below the tephra layer and 40 ka for the samples collected above it. The individual random and systematic uncertainties contributing to the individual ages vary from 1.9 % to 5.4 % and from 6.0 % to 6.1 %, respectively. We obtain an improved overall precision on the age of the sedimentary context by calculating the weighted average age and combining the individual random and systematic uncertainties following Aitken (1985, Appendix B). Thus we report weighted average ages of 46 ka and 40 ka for the horizontally sampled sediment layers intercalating the ash layer and associated overall random uncertainties of 1.3 % and 1.1 %. The overall systematic uncertainties are 6.1 % and 6.0 %. The insights gained from this are discussed in relation to age-modelling studies of luminescence-dated paleoclimate archives, and loess deposits in particular.Keywords: luminescence dating; precision; random uncertainty; quartz