Abstract. The last 5500 years of climate change and environmental response in the northern Benguela Coastal Upwelling are reconstructed by means of three sediment cores from the inner shelf off central Namibia. The study is based on nutrient (δ¹⁵N, δ¹³C) and productivity proxies (accumulation rates of total organic carbon; AR_TOC). Reconstructed sea surface temperatures (alkenone-derived SST) and temperatures at subsurface depths (T_δ18O; based on tests of planktonic foraminifers) reflect the physical boundary conditions. The selection of proxy indicators proved a valuable basis for robust palaeo-climatic reconstructions, with the resolution ranging from multi-decadal (NAM1) over centennial (core 178) to millennial scale (core 226620). The northern Benguela experienced pronounced and rapid perturbation during the middle and late Holocene, and apparently, not all are purely local in character. In fact, numerous correlations with records from the adjacent South African subcontinent and the northern hemisphere testify to global climatic teleconnections. The Holocene Hypsithermal, for instance, is just as evident as the Little Ice Age (LIA) and the Roman Warm Period. The marked SST-rise associated with the latter is substantiated by other marine and terrestrial data from the South African realm. The LIA (at least its early stages) manifests itself in intensified winds and upwelling, which accords with increased rainfall receipts above the continental interior. It appears that climate signals are transferred both via the atmosphere and ocean. The combined analysis of SST and T_δ18O proved a useful tool in order to differentiate between both pathways. SSTs are primarily controlled by the intensity of atmospheric circulation features, reflecting changes of upwelling-favourable winds. T_δ18O records the temperature of the source water and often correlates with global ocean conveyor speed due to varying inputs of warm Agulhas Water. It seems as though conveyor slowdown or acceleration not only affected the temperature of the source water but also its nutrient content. This relationship between source water quality and conveyor speed is already known from glacial times.

doi:10.1002/mmng.201100006