Annually resolved (varved) lake sequences are important palaeoenvironmental archives as they offer a direct incremental dating technique for high-frequency reconstruction of environmental and climate change. Despite the importance of these records, establishing a robust chronology and quantifying its precision and accuracy (estimations of error) remains an essential but challenging component of their development. We outline an approach for building reliable independent chronologies, testing the accuracy of layer counts and integrating all chronological uncertainties to provide quantitative age and error estimates for varved lake sequences. The approach incorporates (1) layer counts and estimates of counting precision; (2) radiometric and biostratigrapic dating techniques to derive independent chronology; and (3) the application of Bayesian age modelling to produce an integrated age model. This approach is applied to a case study of an annually resolved sediment record from Lake Ohau, New Zealand. The most robust age model provides an average error of 72 years across the whole depth range. This represents a fractional uncertainty of ∼5%, higher than the <3% quoted for most published varve records. However, the age model and reported uncertainty represent the best fit between layer counts and independent chronology and the uncertainties account for both layer counting precision and the chronological accuracy of the layer counts. This integrated approach provides a more representative estimate of age uncertainty and therefore represents a statistically more robust chronology.
Bibliographical noteFunding Information:
Support for this research came from GNS Science's direct crown funded research program ‘Global Change through Time’ and the Sarah Beanland Memorial Scholarship (contract C05X1702 ). Many thanks to the additional GNS staff who have supported this research: Kelly Lyons, Cathy Ginnane, Helen Zhang, Johannes Kaiser, Albert Zondervan. We thank the two anonymous reviewers for providing constructive comment that has greatly improved the manuscript.
- Bayesian age modelling
- C dating
- Integrated age model
- Varve chronologies