Master of Science (M.S.)
Degree Granting Department
Brad E. Rosenheim, Ph.D.
Julie N. Richey, Ph.D.
Ryan P. Moyer, Ph.D.
gas ion source, accelerator mass spectrometry, isotope dilution, foraminifera, Pigmy Basin
The development of an accurate and precise geochronology is imperative to understanding archives containing information about Earth’s past. Unable to date all intervals of an archive, researchers use methods of interpolation to approximate age between dates. Sections of the radiocarbon calibration curve can induce larger chronological uncertainty independent of instrumental precision, meaning even a precise date may carry inflated error in its calibration to a calendar age. Methods of interpolation range from step-wise linear regression to, most recently, Bayesian statistical models. These employ prior knowledge of accumulation rate to provide a more informed interpolation between neighboring dates. This study uses a Bayesian statistical accumulation model to inform non-sequential dating of a sediment core using a high-throughput gas-accepting accelerator mass spectrometer. Chronological uncertainty was iteratively improved but approached an asymptote due to a blend of calibration uncertainty, instrument error and sampling frequency. This novel method resulted in a superior chronology when compared to a traditional sediment core chronology with fewer, but more precise, dates from the same location. The high-resolution chronology was constructed for a gravity core from the Pigmy Basin with an overall 95% confidence age range of 360 years, unmatched by the previously established chronology of 460 years. This research reveals that a larger number of low-precision dates requires less interpolation, resulting in a more robust chronology than one based on fewer high-precision measurements necessitating a higher degree of age interpolation.
Scholar Commons Citation
Firesinger, Devon Robert, "Quantity Trumps Quality: Bayesian Statistical Accumulation Modeling Guides Radiocarbon Measurements to Construct a Chronology in Real-time" (2017). USF Tampa Graduate Theses and Dissertations.