MCDoseE 2.0 A new Markov Chain Monte Carlo program for ESR dose response curve fitting and dose evaluation
Embargoed until: 2020-03-01
MetadataShow full item record
This study presents MCDoseE 2.0, a new fitting program for ESR dating dose response curve (DRC) fitting and dose calculation. The standalone software was specifically designed to remove assumed data weighting, and instead to obtain a full probabilistic solution of the DRC by propagating the uncertainties associated with the measured ESR intensities. It uses a non-linear Bayesian framework, specifically a Markov Chain Monte Carlo (MCMC) scheme based on the Metropolis-Hastings algorithm, where the solution is a probability distribution for the equivalent dose, according to the precision of the measurements. In this paper, we investigate the capabilities and limitations of MCDoseE 2.0 by comparing our results to those obtained with OriginPro 9.1®, a proven and commonly used commercial software package. The two programs were evaluated against both known-dose samples and random archaeological tooth enamel and quartz samples, using three commonly used DRC fitting functions. We found that both programs provide highly consistent results. When comparing the dose estimates obtained by both programs we found that 90% of the solutions are statistically indistinguishable regardless of the data weighting assumption used in OriginPro. We also found that MCDoseE 2.0 offers an increased precision on the ending results compared to the commercial software, as long as each measured ESR uncertainty remains within 2-sigma range of the mean error value of all measured ESR uncertainties of the dataset. The accuracy of the fitting results given by MCDoseE 2.0 are undeniably dependent on the measurement accuracy, and emphasises the need of a proper assessment of the experimental errors in the ESR intensities. A copy of the program is available in Supplementary information, and some basic instructions for its use are provided, as well as recommendations to ensure reliable and accurate fitting results.
© 2018, Elsevier. Licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International Licence (http://creativecommons.org/licenses/by-nc-nd/4.0/) which permits unrestricted, non-commercial use, distribution and reproduction in any medium, providing that the work is properly cited.