The Role of Uncertainty in Estimating Lead/Lag Relationships in Marine Sedimentary Archives: A Case Study From the Tropical Pacific. (23rd November 2017)

• Record Type: Journal Article
• Authors: Khider, D.; Ahn, S.; Lisiecki, L. E.; Lawrence, C. E.; Kienast, M.
• Is Part Of: Paleoceanography. Volume 32:Number 11(2017)
• Page Start: 1275
• Page End: 1290
• Publication Date: 2017-11-23
• Subjects: sea surface temperature -- equatorial Pacific -- lead/lag -- uncertainty -- benthic d18O -- deglaciation timing; Paleoceanography -- Periodicals; 551.46
• DOI: 10.1002/2016PA003057
• Languages: English
• ISSNs: 0883-8305

Abstract

Abstract: Understanding the mechanisms behind any changes in the climate system often requires establishing the timing of events imprinted on the geological record. However, these proxy records are prone to large uncertainties, which may preclude meaningful conclusions about the relative timing of events. In this study, we put forth a framework to estimate the uncertainty in phase relationships inferred from marine sedimentary records. The novelty of our method lies in the accounting of the various sources of uncertainty inherent to paleoclimate reconstruction and timing analysis. Specifically, we use a Monte Carlo process allowing sampling of possible realizations of the time series as functions of uncertainties in time, the climate proxy, and the identification of the termination timing. We then apply this technique to 15 published sea surface temperature records from the equatorial Pacific to evaluate whether we observed any significant changes in the termination timing between the east and the west. We find that the uncertainty on the relative timing estimates is on the order of several thousand years and mainly stems from age model uncertainty (90%). However, even small differences in mean termination timings can be detected with a sufficiently large number of samples. Improvements in the dating of sediment records provide an opportunity to reduce uncertainty in studies of this kind. Key Points: Uncertainty analysis is based on Monte Carlo realizations of the time series Age uncertainties are the main contributor to the overall uncertainty Phase relationship needs to be established on a large number of records