Dust correlation and oxygen isotope stratigraphy in the Southern Ocean over the last 450 kyrs: An Indian sector perspective

Abstract

The chronology of Southern Ocean (SO) marine sediment cores forms the basis of understanding the SO paleoceanography, with significant implications for global climate. Because tuning of the oxygen isotope (δ18O) record of a marine sediment core to a δ18O stack (δ18O stratigraphy) is difficult in the SO because of a general paucity of calcareous foraminifera, tuning of the dust proxy signal of a marine sediment core to the dust record of an ice core (dust correlation) is a promising way to construct an age–depth model. However, the reliability of dust correlation has not been established, especially beyond the last ∼300 kyrs, and such work has been performed more in the Atlantic and Pacific sectors of the SO than in the Indian sector. Here we present a new dust correlation using the rock magnetic record, together with continuous δ18O stratigraphy, for a marine sediment core in the Indian sector of the SO over the last ∼410 kyrs. The δ18O stratigraphy is consistent with the dust correlation within their chronological uncertainties, supporting the reliability of the latter chronology for glacial–interglacial timescales. However, the dust correlation often produces older ages than the δ18O stratigraphy by up to 3 kyrs (410–126 ka and 84–50 ka). We additionally compiled, based on dust correlation, available dust proxy records in the circum-Antarctic SO for the last 450 kyrs to discuss variations in the dust proxy record within the Indian sector and between all three sectors. For the Indian sector, there is a marked difference in dust proxy signals for Marine Isotope Stage (MIS) 6 between two cores that have markedly different latitudinal positions relative to the oceanic fronts. Increased biogenic magnetite production by iron fertilization in the Subantarctic Zone during MIS 6 may partly explain the difference between the two cores. For the circum-Antarctic SO, moving correlation coefficients were computed between the marine sediment-core dust proxy signals and the dust flux of an Antarctic ice core. Strong correlation was recognized for intervals of high dust flux in the ice core, except for MIS 6. Although minor, the slightly lower correlation in the Indian sector than in the other sectors indicates a contribution from local dust source and volcanic materials and thus suggests the necessity for caution when performing dust correlation in the Indian sector of the SO.