Southwest Atlantic water mass evolution during the last deglaciation

This is the publisher’s final pdf. The published article is published by John Wiley & Sons Ltd. and copyrighted by American Geophysical Union. It can be found at: http://agupubs.onlinelibrary.wiley.com/agu/journal/10.1002/%28ISSN%291944-9186/The rise in atmospheric CO₂ during Heinrich Stadial 1 (HS1; 14.5–17.5 kyr B.P.) may have been driven by the release of carbon from the abyssal ocean. Model simulations suggest that wind-driven upwelling in the Southern Ocean can liberate ¹³C-depleted carbon from the abyss, causing atmospheric CO₂ to increase and the δ¹³C of CO₂ to decrease. One prediction of the Southern Ocean hypothesis is that water mass tracers in the deep South Atlantic should register a circulation response early in the deglaciation. Here we test this idea using a depth transect of 12 cores from the Brazil Margin. We show that records below 2300 m remained ¹³C-depleted until 15 kyr B.P. or later, indicating that the abyssal South Atlantic was an unlikely source of light carbon to the atmosphere during HS1. Benthic δ¹⁸O results are consistent with abyssal South Atlantic isolation until 15 kyr B.P., in contrast to shallower sites. The depth dependent timing of the δ¹⁸O signal suggests that correcting δ¹⁸O for ice volume is problematic on glacial terminations. New data from 2700 to 3000 m show that the deep SW Atlantic was isotopically distinct from the abyss during HS1. As a result, we find that mid-depth δ¹³C minima were most likely driven by an abrupt drop in δ¹³C of northern component water. Low δ¹³C at the Brazil Margin also coincided with an ~80‰ decrease in Δ¹⁴C. Our results are consistent with a weakening of the Atlantic meridional overturning circulation and point toward a northern hemisphere trigger for the initial rise in atmospheric CO₂ during HS1