| The shelves adjacent to Antarctica are characterised by cold, saline waters,and feature some of the densest waters of the global ocean. Sites where these shelf waters can descend off the shelf and contribute to deep waters are important for the global thermohaline circulation, and hence global climate. Coupled interannual modes of climate variability have the potential to significantly influence dense water formation via their effects on the atmosphere, cryosphere and surface ocean. Recent indications are that deep water formation in the Northern Hemisphere is influenced by at least one such coupled interannual mode of variability, however similar influences are largely unexplored in the Southern Ocean due to the general paucity of long time series from the deep oceans adjacent to Antarctica. Here, we present a long series (1992-2001) of bottom potential temperature from 1040m depth on the continental slope north of Elephant Island, near the tip of the Antarctic Peninsula. Despite the depth, a strong annual signal is apparent (peak-to-peak around 0.5˚C) with minimum potential temperatures in the months of the austral winter. This range is very similar to that observed in much shallower water on the nearby Antarctic shelf, and is too large to be explained by the continuation of the Polar Slope Current from Weddell Sea. Also similar to records from the Antarctic shelf, our record shows large-amplitude cold spikes in potential temperature that are confined almost entirely to the austral winter. These are the signals of individual convection events caused by brine release during localised sea ice production. The water in these convective plumes is dense enough to seasonally displace the Circumpolar Deep Water, and directly inject dense, cold water into the deep open ocean (commonly colder than -0.3˚C; at their coldest -0.9˚C). The coldest water in the downslope plume was observed during 1998, a period for which anomalous sea ice conditions in the region have already been noted, and demonstrated to be due to the strong 1997/98 ENSO event. Given the quasi-periodic nature of ENSO, we expect further repeats of such strong convective events on interannual timescales. However, the Antarctic Peninsula is known to be warming more rapidly than any other region of the Southern Hemisphere, with consequences for the sea ice field. If seasonal sea ice production near the tip of the Peninsula is reduced, it must be expected that this mode of ventilating the deep Southern Ocean will be suppressed. |
|
|