Ocean circulation in the Atlantic Ocean is dominated by the operation of the ‘conveyor belt’, which transports warm waters northward towards and past Europe, which then cool and sink to the north of Iceland before flowing back to the South Atlantic in the deep ocean. The strength of this conveyor is linked to both the temperature and the salinity of the northward flowing Atlantic waters, which control water density and the ability of the surface waters to sink in the north. A lot of research has demonstrated that by adding freshwater to the North Atlantic the conveyor system can become weaker, because the water density is reduced and more difficult to sink.
An additional control of Atlantic surface water properties is found in the South Atlantic: to the south of Cape Town, the Indian and Atlantic Oceans meet, and some of the warm and salty waters of the Indian Ocean are transferred into the Atlantic, where they are transported northward and into the conveyor. This process, termed ‘Agulhas leakage’, is not constant, but varies depending upon the position and intensity of the westerly winds as well as the position of the oceanic fronts associated with the Antarctic Circumpolar Current. Over the last few cycles of ice sheet expansion and retreat (the ‘glacial-interglacial cycles’) we have observed changes to the strength of Agulhas leakage, but it is not clear what caused these changes, nor how they developed over the last 5 million years when we are aware of other major transitions occurring elsewhere in the climate system (e.g. growth of larger ice sheets, intensification of wind systems, shifting Antarctic Circumpolar Current position). In this project we take advantage of the first continuous sediment sequence to extend through the last 5 million years located within the direct pathway of the Agulhas leakage. The new record (Site U1479) was drilled in spring 2016, and spans the ice ages of the Pleistocene (0-2.7 million years ago) and the globally warmer climate of the Pliocene (2.7-5.3 million years ago). Previously we have identified evidence for increasing leakage (more warming) in the South Atlantic during times where ice-sheets were larger than present (which we would have expected to cause more cooling), but our records are limited in the time window they cover. We will reconstruct sea-surface temperatures at Site U1479 from the present day back to 5 million years ago. We will compare our results to existing data from the South Atlantic, to consider whether there is evidence for changes in Agulhas leakage in response to changes in wind strength and position, and changes to the position of the Antarctic Circumpolar Current. As the first continuous reconstruction of surface ocean properties spanning the last 5 million years from the corridor of Agulhas leakage, our new data will also potentially shed light onto how and why the ocean conveyor has also evolved over this time window.