Correction of the IRD influence for paleo‐current flow speed reconstructions in hemipelagic sediments

Reconstructions of past changes in deep-sea current intensities are needed to understand ocean-climate interactions in the past. The mean size of the sortable silt fraction (10 – 63 µm, ) is one of the most used proxies in this domain. However, in polar and subpolar environments under relatively low...

Full description

Bibliographic Details
Published in:Paleoceanography and Paleoclimatology
Main Authors: Stevenard, N., Govin, A., Kissel, C., Van Toer, A.
Format: Article in Journal/Newspaper
Language:English
Published: American Geophysical Union (AGU) 2023
Subjects:
Zr
Online Access:https://archimer.ifremer.fr/doc/00821/93296/99978.pdf
https://archimer.ifremer.fr/doc/00821/93296/99979.pdf
https://doi.org/10.1029/2022PA004500
https://archimer.ifremer.fr/doc/00821/93296/
Description
Summary:Reconstructions of past changes in deep-sea current intensities are needed to understand ocean-climate interactions in the past. The mean size of the sortable silt fraction (10 – 63 µm, ) is one of the most used proxies in this domain. However, in polar and subpolar environments under relatively low flow speed conditions, the presence of Ice-Rafted Detritus (IRD) may alter the record and thus bias the interpretation of paleo-current strength changes. In this paper, we examine the influence of IRD on the record of three sedimentary cores from the subpolar North Atlantic and the Antarctic margin. The influence of unsorted IRD on records is clearly established. To remove this IRD influence on grain-size distributions, we propose a new method based on End-Member Analysis approach, and for which a MATLAB script is made available. This method characterizes the grain-size distribution of the unsorted IRD input, allowing it to be isolated and discarded, and the current sensitive variability to be robustly identified. The method therefore allows the recalculation of a modified sediment grain-size distribution free of unsorted IRD influence and the construction of modified and sortable silt percentage (the % of the 10-63 μm in the total <63 μm fraction) records. The application of the method to the three studied cores shows that (i) the unsorted IRD component is correctly removed from the grain-size signal and (ii) the new record is consistent with the XRF-based ln(Zr/Rb) grain-size proxy. This article is protected by copyright. All rights reserved. Key Points Evidence for unsorted IRD influence on records of hemipelagic sediments in subpolar environments New method proposed to provide grain-size distributions free of unsorted IRD influence Allows to define a paleo-current strength proxy free of unsorted IRD contribution