Speakers Club: Alan Jones (UCSB)

Event Date: 

Thursday, April 20, 2017 -
2:00pm to 2:45pm

Event Location: 

  • Webb 1100

Title: Testing the consistency of four age modeling techniques using Iberian Margin sediment cores

Abstract: Interpretations of paleoclimate records from ocean sediment cores rely on age-depth models, which provide estimates of sediment age as a function of core depth. Here we compare four methods used to generate age models for cores from the past 40 kyr. The first method is based on radiocarbon dating and uses the Bayesian statistical software, Bacon [Blaauw and Christen, 2011], to produce continuous age estimates (i.e., between radiocarbon dates) and their uncertainties. The second method is to perform benthic d18O alignments to a target core; here we use the probabilistic alignment algorithm, HMM-Match, which runs in an automated way (requiring no tuning) and generates 95% confidence intervals [Lin et al., 2014]. The third and fourth methods are to perform planktonic d18O and sea surface temperature alignments to a target core; here we use Match [Lisiecki and Lisiecki, 2002]. Like HMM-Match, Match is automated, but it requires tuning and does not produce uncertainty estimates. We present results from 9 high-resolution cores from the Iberian margin for which multiple age model types were produced. Our benthic alignment target is core MD95-2042 on the regional age model for the deep North Atlantic developed by Stern and Lisiecki [2014]. We find that all four age model strategies usually produce results which agree with one another to within age uncertainty estimates. We also compare the 95% confidence widths produced by HMM-Match and Bacon. We conclude that, for Iberian margin cores with sufficient resolution, all techniques appear equally effective; however, planktonic d18O and SST alignments should only be applied to nearby cores and cannot be used to generate age uncertainty estimates.