Bayesian radiocarbon age modeling

Published: 25 June 2021| Version 1 | DOI: 10.17632/sfdwkyh848.1
Shiyong Yu


Matlab implementation of a hierarchical Bayesian model for calibrating radiocarbon ages in a sequence, which consists of one or several phases each has an early and late boundary. Ages in each phase can be either chronologically "ordered" or "unordered", and the boundaries of two neighboring phases can be either "contiguous" or "disjoint". Note that the overlapping phase relationship was not implemented in this version.


Steps to reproduce

In the Matlab environment, open the file named "MatCalib_v1.0". First, enter the radiocarbon ages (including the reservoir ages if any) to be modeled, then specify the ordering of the ages in each phase (e.g., ordered/unordered), the relationship of the internal boundaries of each two neighboring phases (e.g., contiguous/disjoint), the beginning and the end of the study period, and the calibration curve to be used. Because MatCalib is simulation-based using the MCMC method, users should be aware that the results from each run may be more or less different.


Radiocarbon Dating