UC Santa Barbara

Interpretations of paleoclimate records from ocean sediment cores rely on age-depth models, which provide estimates of age as a function of core depth. Here I compare four methods used to generate age models for sediment cores for the past 140 thousand years (kyr). The first method is based on radiocarbon dating using the Bayesian statistical software, Bacon [Blaauw and Christen, 2011]. The second method aligns benthic δ18O to a target core using the probabilistic alignment algorithm, HMM-Match, which generates 95% confidence intervals [Lin et al., 2014]. The third and fourth methods are to perform planktonic δ18O and sea surface temperature (SST) alignments to the same target core, using Match [Lisiecki and Lisiecki, 2002]. Unlike HMM-Match, Match requires parameter tuning and does not produce uncertainty estimates. I compare multiple age model types for nine high-resolution cores from the Iberian margin. The root mean square error between the individual age model results and each core’s average estimated age is 1.4 kyr. Additionally, HMM-Match and Bacon age estimates agree to within uncertainty and have similar 95% confidence widths of 1-2 kyr for the highest resolution records. In one core, the planktonic and SST alignments did not fall within the HMM-Match 95% confidence intervals. For this core, the surface proxy alignments likely produce more reliable results due to millennial-scale SST variability. Thus, I find evidence that HMM-Match may have underestimated alignment uncertainty for one of the six benthic δ18O alignments performed.