Statistical modelling of single grain quartz De distributions and an assessment of procedures for estimating burial dose

Abstract

It is possible to increase the accuracy of the optical dating of some Quaternary deposits by deriving age estimates from individual quartz grains. Age estimates from individual grains within single samples display varying degrees of inconsistency however and this paper address the difficulties in reducing complex distributions of single-grain age estimates to single accurate sample ages using existing statistical age models. A modelled response of the quartz OSL signal to irradiation and illumination (both in the laboratory and under natural conditions) was used to test the accuracy of various published statistical age-models. ‘Aeolian’ and ‘fluvial’ transport conditions were simulated by adjusting the wavelength range of the daylight spectrum and the incident photon flux during the simulated bleaching event. A heterogeneous mix of grains was produced using a Monte Carlo approach, in which key irradiation and bleaching parameters were varied for each simulated grain, and the accuracy of published statistical models assessed. Results suggest that no single age-model is appropriate for all modelled cases. A formal decision process is suggested and shown to make appropriate choices between the available age-models in each of the modelled cases and in ∼85% of cases using 43 known-age sediment samples.