Terrestrial pollen records are abundant and widely distributed, making them an excellent proxy for past vegetation dynamics. Age-depth models relate pollen samples from sediment cores to a depositional age based on the relationship between sample depth and available chronological controls. Large-scale synthesis of pollen data benefit from consistent treatment of age uncertainties. Generating new age models helps to reduce potential artifacts from legacy age models that used outdated techniques. Traditional age-depth models, often applied for comparative purposes, infer ages by fitting a curve between dated samples. Bacon, based on Bayesian theory, simulates the sediment deposition process, accounting for both variable deposition rates and temporal/spatial autocorrelation of deposition from one sample to another within the core. Bacon provides robust uncertainty estimation across cores with different depositional processes. We use Bacon to estimate pollen sample ages from 554 North American sediment cores. This dataset standardizes age-depth estimations, supporting future large spatial-temporal studies and removes a challenging, computationally-intensive step for scientists interested in questions that integrate across multiple cores.
Bibliographical noteFunding Information:
The work is funded by NSF (grant number: DEB-1655898). SJG was funded through NSF grants 1740694 and 1550707. We thank Jadyn M. Sethna for recording Bayesian age into Neotoma. We thank M. Allison Stegner, Jordan Schutz, and Jack Williams for tutoring data-upload methods. We thank Jessica Blois for discussion about this project. And we thank the reviewers for the comments to improve the manuscript.
© 2019, The Author(s).
Copyright 2020 Elsevier B.V., All rights reserved.
PubMed: MeSH publication types
- Journal Article
- Research Support, U.S. Gov't, Non-P.H.S.