Stability of North Atlantic water masses in face of pronounced climate variability during the Pleistocene

Abstract

[1] Geochemical profiles from the North Atlantic Ocean suggest that the vertical d 13 C structure of the water column at intermediate depths did not change significantly between glacial and interglacial time over much of the Pleistocene, despite large changes in ice volume and iceberg delivery from nearby landmasses. The most anomalous d 13 C profiles are from the extreme interglaciations of the late Pleistocene. This compilation of data suggests that, unlike today (an extreme interglaciation), the two primary sources of northern deep water, Norwegian-Greenland Sea and Labrador Sea/subpolar North Atlantic, had different characteristic d 13 C values over most of the Pleistocene. We speculate that the current open sea ice conditions in the Norwegian-Greenland Sea are a relatively rare occurrence and that the high-d 13 C deep water that forms in this region today is geologically unusual. If northern source deep waters can have highly variable d 13 C, then this likelihood must be considered when inferring past circulation changes from benthic d 13 C records.

The middle Pleistocene transition: characteristics, mechanisms, and implications for long-term changes in atmospheric pCO2

“…The classic picture that has evolved since the work of Boyle and Keigwin (1987) and Duplessy et al (1988) is one of the present interglaciation as having little to no d 13 C gradient, reflecting the penetration of NADW to the abyss, versus the LGM 21 ka having a positive d 13 C gradient, suggesting that the North Atlantic was stratified with a divide at $2000 m separating enriched Glacial North Atlantic Intermediate Water from underlying depleted CDW. The Raymo et al (2004) analysis, however, found that glacial-interglacial d 13 C gradients over the last 1800 kyr often differed from this picture, with the most recent examples being relatively unique. Instead, the ''typical'' vertical d 13 C gradient, characteristic of both glaciations and interglaciations, is negative, implying a source of depleted d 13 C water in the North Atlantic.…”

“…Instead, the ''typical'' vertical d 13 C gradient, characteristic of both glaciations and interglaciations, is negative, implying a source of depleted d 13 C water in the North Atlantic. Raymo et al (2004) speculate that such a source was from brine formation beneath permanent sea ice covering the Norwegian-Greenland Seas, with depleted d 13 C values reflecting an attendant reduction in air-sea exchange.…”

“…In order to account for global ocean d 13 C variations associated with changes in the carbon cycle, we follow the approach of Shackleton et al (1983) by calculating interocean and vertical d 13 C gradients to estimate the relative strengths of deep-water formation (Raymo et al, 1990(Raymo et al, , 2004Mix et al, 1995b).…”

“…Raymo et al (1990) argued that the MPT change to more depleted glacial d 13 C at Site 607 (and a decrease in the Dd 13 C (AÀP) signal) represented a weakening of NADW formation during glaciations, allowing greater penetration of depleted CPW into the North Atlantic basin. More recently, Raymo et al (2004) argued that the progressive depletion of d 13 C in the Site 607 record may reflect instead a greater influence of CDW. Support for this latter argument comes from the intra-Atlantic d 13 C gradient (Dd 13 C (SÀA) ) derived by subtracting the Site 607 d 13 C values from Site 1090 d 13 C values (Fig.…”

“…In particular, the average Dd 13 C (SÀA) signal remains constant throughout the Pleistocene, suggesting that the depletion in glacial d 13 C seen in both records at the start of the MPT is related to a change in the ventilation of CDW. Raymo et al (2004) further examined changes in the history of NADW formation by reconstructing vertical (depth) gradients of d 13 C in the North Atlantic using 10,000-yr averaged d 13 C values representing glaciations and interglaciations over the last 1800 ka. The classic picture that has evolved since the work of Boyle and Keigwin (1987) and Duplessy et al (1988) is one of the present interglaciation as having little to no d 13 C gradient, reflecting the penetration of NADW to the abyss, versus the LGM 21 ka having a positive d 13 C gradient, suggesting that the North Atlantic was stratified with a divide at $2000 m separating enriched Glacial North Atlantic Intermediate Water from underlying depleted CDW.…”