(Table S1) Radiocarbon data of sediment core SO161/5 22SL

At the end of the Last Glacial Maximum (19,000 to 11,000 years ago), atmospheric carbon dioxide concentrations rose while the Delta14C of atmospheric carbon dioxide declined**1, 2. These changes have been attributed to an injection of carbon dioxide with low radiocarbon activity from an oceanic abyssal reservoir that was isolated from the atmosphere for several thousand years before deglaciation**3. The current understanding points to the Southern Ocean as the main area of exchange between these reservoirs4. Intermediate water formed in the Southern Ocean surrounding Antarctica would have then carried the old carbon dioxide signature to the lower-latitude oceans**5, 6. Here we reconstruct the Delta14C signature of Antarctic Intermediate Water off the coast of Chile for the past 20,000 years, using paired 14C ages of benthic and planktonic foraminifera. In contrast to the above scenario, we find that the delta14C signature of the Antarctic Intermediate Water closely matches the modelled surface ocean Delta14C, precluding the influence of an old carbon source. We suggest that if the abyssal ocean is indeed the source of the radiocarbon-depleted carbon dioxide, an alternative path for the mixing and propagation of its carbon dioxide may be required to explain the observed changes in atmospheric carbon dioxide concentration and radiocarbon activity.

(Table 1) U and Th isotopic composition and 230Th ages of samples from the A1 and C1 corals of Legare Anchorage

The 14C reservoir age of the surface ocean was determined for two Holocene periods (4908-4955 and 3008-3066 calendar (cal) B.P.) using U/Th-dated corals from Biscayne National Park, Florida, United States. We found that the average reservoir ages for these two time periods (294 ± 33 and 291 ± 27 years, respectively) were lower than the average value between A.D. 1600 and 1900 (390 ± 60 years) from corals. It appears that the surface ocean was closer to isotopic equilibrium with CO2 in the atmosphere during these two time periods than it was during recent times. Seasonal d18O measurements from the younger coral are similar to modern values, suggesting that mixing with open ocean waters was indeed occurring during this coral's lifetime. Likely explanations for the lower reservoir age include increased stratification of the surface ocean or increased D14C values of subsurface waters that mix into the surface. Our results imply that a more correct reservoir age correction for radiocarbon measurements of marine samples in this location from the time periods ~3040 and ~4930 cal years B.P. is ~292 ± 30 years, less than the canonical value of 404 ± 20 years.

(Table 2) Radiocarbon age determination from the polar planktonic foraminifera Neogloboquadrina pachyderma s for sediment core HU90-013-029

An additional Heinrich ice-rafting event is identified between Heinrich events 5 and 6 in eight cores from the Labrador Sea and the northwest Atlantic Ocean. It is characterized by sediment rich in detrital carbonate (40% CaCO3) with high concentration of floating dropstones, high coarse-fraction (% > 150 µm) content, and has a sharp contact with the underlying but grades into the overlying hemipelagic sediment. It also shows lighter d18ONpl values, indicating freshening due to iceberg rafting and/or meltwater discharge. This event is correlated with Dansgaard-Oeschger event 14 and interpreted as an additional Heinrich event, H5a. The thickness of H5a in the Labrador Sea reaches up to 220 cm. This additional Heinrich event has also been reported in cores PS2644 and SO82-5 from the northern North Atlantic. With the recognition of H5a the temporal spacing between Heinrich events 1 to 6 becomes more uniform (~7 ka).

Physical and chemical properties, 14C ages and palynology of sediment cores from the Western Baltic Sea

Sea level related radiocarbon, palynological and stratigraphical data from sediment cores in the Western Baltic have been tested against the existing sea level curves for the region. The relative sea level rise curves for the beginning of the Holocene show no significant deviations between the Kiel, Mecklenburg und Lübeck Bays and hence do not support the previously reported differences in the averaged regional subsidence rates for this time interval. Local subsidence and upheaval due to salt tectonics probably played a greater role than previously suspected in the region. The sea level possibly stagnated around -28 m during the early Holocene before rising very rapidly to -14 m. The submarine terraces at -30 m and perhaps also at -27 m were formed during the lacustrine phase of the Western Baltic when the water levels were controlled by the main thresholds in the Great Belt.

AMS C-14 ages of coretops collected on RRS Charles Darwin cruise CD159, July 2004, N. Atlantic, for the NERC RAPID programme

Material and data were collected at 41 sites in the subpolar North Atlantic Ocean between Scotland and Newfoundland, during the RRS CharlesDarwin CD159 cruise in July 2004 (McCave, 2005). Sites were selected to reflect the major inputs of water that becomes the North Atlantic Deep Water (NADW); the Iceland-Scotland Overflow Water (ISOW), the Denmark Strait Overflow Water (DSOW) and the Labrador Sea Water (LSW). Areas cored were the south Iceland Rise, SE Greenland slope/rise and Eirik Drift, and the Labrador margin. A total of 29 box cores, 19 piston cores, 6 kasten cores, 9 short gravity cores and 20 CTD casts as well as 28 surface water samples were collected during the cruise. Here we present sediment core-top sample ages. The cores were sampled at 1 or 0.5 cm intervals and we used the top 1 or 2 cm, depending on availability of foraminifera in the samples. Sediment samples were disaggregated on an end-over-end wheel, wet sieved at >63 um, and dry sieved to 63-150 and >150 um. Accelerator Mass Spectrometer (AMS) radiocarbon dating was done for each core top based on between 900-1600 monospecific planktonic foraminifera (Globigerina bulloides or Neogloboquadrina pachyderma (sinistral)). All dates were of modern or late Holocene age except site RAPID-08-5B (9806 ± 38 uncorrected 14C years BP) and site RAPID-14-10B (11543 ± 40 uncorrected 14C years BP). The >150 um fraction was split until approximately 300 foraminifera remained and counted for number of lithic grains, benthic foraminifera, planktonic foraminifera and foraminifera fragments. In all but the shallowest sample (Greenland rise, 761m water depth) benthic foraminifera constituted less than 2% of the total >150 um fraction of the sample.