ABSTRACT Bremen 2004 meeting
Drilling to determine the Origin and Influence of the Antarctic Circumpolar Current
Peter Barker1 and Ellen Thomas2
The conventional model of Antarctic Circumpolar Current (ACC) onset and influence, first proposed almost 30 years ago, has the ACC beginning at c.34 Ma (Eocene-Oligocene boundary), identified from a N-S planktonic biocalcareous-biosiliceous microfossil transition and responsible for the onset of Antarctic glaciation by thermally isolating the continent. It may not hold. For the ACC (as known today) to have developed, a continuous deep-water path around Antarctica is required. Recent drilling (Leg 189: Exon, Kennett et al. 2001) confirmed rapid subsidence of the S Tasman Rise south of Australia about 34 Ma ago, but tectonic modelling (Lawver and Gahagan submitted) argues that a deep-water gap may have opened earlier. The time of deep-water opening S of S America is disputed (eg. Barker 2001; Lawver and Gahagan submitted). Recent climate models suggest that (a) a steady drop in atmospheric CO2 could have caused glacial onset without an ACC (De Conto and Pollard 2002) and (b) oceanographic changes consequent on creation of a gap S of Australia would not have isolated Antarctica thermally (Huber et al. submitted). ACC transport is now known to be confined within continuous, deep-reaching oceanic fronts (eg Gille 1994; Heywood and King 2002). Sea-surface temperature changes, previously used to infer ACC existence, may instead merely reflect Antarctic glaciation (Barker and Thomas in press).
These developments greatly modify the optimal strategy for determining ACC development. Any drilling proposal should examine terrigenous sediment provenance and bottom current flow (away from regions of glacially-induced bottom-water flow). Sea-surface temperatures should be examined primarily to determine their true significance. Oceanic fronts meander and spawn rings, so fast flow at depth is intermittent at any one location: this limits inferences about flow strength from grain-size and related studies. Sites should be within meander regions, but away from mean positions to avoid non-deposition. Sedimentary sections at sites should be pelagic/hemipelagic and continuous, and at least some should extend sufficiently far back in time to cover the range of hypothesised times of onset. The Scotia Sea region is optimal (Barker and Thomas in press; submitted) because (a) the general level of knowledge there is high, allowing suitable sites to be identified with confidence and (b) it is close to the most likely region of final opening, minimising the risk of extraneous contamination of the ACC signal.
References: