Constraints on the precision of SAR in equivalent dose estimations close to saturation in quartz

Abstract

Difficulties in evaluating accurate equivalent doses for Late Quaternary sediments from the south coasts of Greece arose in the course of SAR measurements. Large scatter among apparent equivalent doses perplexed the evaluation of accurate mean estimates in circumstances where very large doses were involved. Besides, dose recovery experiments revealed that when intersection of the natural signal occurs onto the saturating segment of the growth curve, a distinct relationship between D e and growth curve shape occurs. At very low slope angles, SAR tends to overestimate the recovered dose. The mechanism which controls the spread in equivalent dose estimation is investigated.

Exploring Paleogeographic Conditions at TwoPaleolithic Sites inNavarino, SouthwestGreece, Dated by Optically Stimulated Luminescence

“…In particular, errors associated with measurement of luminescence intensity will be amplified if they are interpolated onto an exponentially saturating dose‐response curve. A similar pattern was observed by Athanassas (: Figure ) in sedimentary quartz from the southwest Peloponnese. Therefore, the transition point from linearity to sublinearity in the dose‐response curve of quartz constitutes a divide, beyond which age assessments are characterized by relatively larger errors (Murray & Funder, ; Wintle & Murray, ; Athanassas, ).…”

“…In circumstances where complete sunlight bleaching can be guaranteed, Murray and Funder () and Athanassas () have predicted that paleodose distributions resulting from exponential fits developed from samples near saturation, inevitably demonstrate large dispersion and skewness toward high doses. Paleodose distributions from the Paleolithic layers at Navarino bear obvious resemblance to Murray and Funder's () and Athanassas’ () predictions, probably because they have been influenced by growth curves leaning toward saturation (Figure b). The same mechanism may also account for similar results acquired during older dating attempts (Parkinson & Cherry, ).…”

“…The same mechanism may also account for similar results acquired during older dating attempts (Parkinson & Cherry, ). Athanassas () suggested that paleodose distributions of this type not only display large uncertainty, but they also result in overestimates of mean paleodose when individual values are averaged. This overestimation is attributed to a portion of aliquots whose paleodoses are strongly controlled by the degree of curvature of their respective growth curves.…”

“…This dose value is characteristic of each growth curve and can be determined by solving the fitted single exponential to 2D o , namely: where D is the laboratory dose, I max the saturation level, and P the paleodose. All samples were subjected to the criteria of Wintle and Murray () and Athanassas (), and any aliquot that did not meet the dose parameter described above was rejected. Use of this procedure decreased the spread of the mean paleodoses (Table ), but only to a limited degree.…”

“…This probably influenced the OSL results from PRAP. Utilization of measures for dealing with samples close to saturation (see OSL dating procedures section) lately put forward by other authors (Wintle & Murray, 2006;Athanassas, 2011) were effective in reducing some of the observed uncertainty here, but only to a limited degree.…”