Preliminary Results for Estimating the Bone Background Uncertainties at SUERC Using Statistical Analysis

Abstract

AbstractBone is frequently dated in archaeological studies and, especially for very old bones (more than 40,000 years old), it is critical to have an accurate and precise measure of the material-specific background value and its associated uncertainty. The SUERC Radiocarbon Laboratory has obtained a mammoth bone as a background bone standard and an appropriate number are now routinely prepared and measured in each AMS batch, resulting in the accumulation of a large number of background bone results over a two-year period. Additionally, information on which of the two accelerator mass spectrometry (AMS) instruments was used to make the radiocarbon (14C) measurements, and which sample pretreatment method (modified Longin or modified ultrafiltration) was used to extract and purify the collagen, is recorded for each sample. These data have been used to estimate the laboratory bone background (to be subtracted from each unknown bone sample prepared in the laboratory) and its associated analytical uncertainty. The statistical analysis of the bone results has made use of a linear mixed effects model to examine the variation, and to apportion the overall variation between and within batches on both AMS instruments, and the different pretreatment methods used at SUERC.