Stable isotope analysis and temperature reconstruction data from DSDP Hole 94-609B and ODP Hole 162-984B

The Greenland ice sheet is accepted as a key factor controlling the Quaternary glacial scenario. However, the origin and mechanisms of major Arctic glaciation starting at 3.15 Ma and culminating at 2.74 Ma are still controversial. For this phase of intense cooling Ravelo et al. proposed a complex gradual forcing mechanism. In contrast, our new submillennial-scale paleoceanographic records from the Pliocene North Atlantic suggest a far more precise timing and forcing for the initiation of northern hemisphere glaciation (NHG), since it was linked to a 2-3 °C surface water warming during warm stages from 2.95 to 2.82 Ma. These records support previous models, claiming that the final closure of the Panama Isthmus (3.0- ~2.5 Ma induced an increased poleward salt and heat transport. Associated strengthening of North Atlantic Thermohaline Circulation and in turn, an intensified moisture supply to northern high latitudes resulted in the build-up of NHG, finally culminating in the great, irreversible climate crash at marine isotope stage G6 (2.74 Ma). In summary, there was a two-step threshold mechanism that marked the onset of NHG with glacial-to-interglacial cycles quasi-persistent until today.

Northwards and eastwards velocities extracted from a historical run (for the year 2000) of the UVic Earth System Climate Model of Weaver et al. (2001)

This study provides a theoretical assessment of the potential bias due to differential lateral transport on multi-proxy studies based on a range of marine microfossils. Microfossils preserved in marine sediments are at the centre of numerous proxies for paleoenvironmental reconstructions. The precision of proxies is based on the assumption that they accurately represent the overlying watercolumn properties and faunas. Here we assess the possibility of a syn-depositional bias in sediment assemblages caused by horizontal drift in the water column, due to differential settling velocities of sedimenting particles based on their shape, size and density, and due to differences in current velocities. Specifically we calculate the post-mortem lateral transport undergone by planktic foraminifera and a range of other biological proxy carriers (diatoms, radiolaria and fecal pellets transporting coccolithophores) in several regions with high current velocities. We find that lateral transport of different planktic foraminiferal species is minimal due to high settling velocities. No significant shape- or size-dependent sorting occurs before reaching the sediment, making planktic foraminiferal ideal proxy carriers. In contrast, diatoms, radiolaria and fecal pellets can be transported up to 500km in some areas. For example in the Agulhas current, transport can lead to differences of up to 2°C in temperature reconstructions between different proxies in response to settling velocities. Therefore, sediment samples are likely to contain different proportions of local and imported particles, decreasing the precision of proxies based on these groups and the accuracy of the temperature reconstruction.

Water temperature reconstruction for Terminations I and II of sediment core MD01-2378

We present centennial records of sea surface and upper thermocline temperatures in Core MD01-2378 from the Timor Sea, which provide new insights into the variability of the Indonesian outflow across the last two glacial terminations. Mg/Ca in Globigerinoides ruber (white s. s.) indicates an overall increase of 3.2 °C in sea surface temperature (SST) over Termination I. Following an early Holocene plateau at 11.3-6.4 ka, SSTs cooled by 0.6 °C during the middle to late Holocene (6.4-0.7 ka). The early Holocene warming occurred in phase with increasing northern hemisphere summer insolation, coinciding with northward displacement of the Intertropical Convergence Zone, enhanced boreal summer monsoon and expansion of the Indo-Pacific Warm Pool. Thermocline temperatures (Pulleniatina obliquiloculata Mg/Ca) gradually decreased from 24.5 to 21.5 °C since 10.3 ka, reflecting intensification of a cool thermocline throughflow. The vertical structure of the upper ocean in the Timor Sea evolved in similar fashion during the Holocene and MIS5e, although the duration of SST plateaux differed (11.3 to 6.4 ka in Termination I and from 129 to 119 ka in Termination II), which was probably due to the more intense northern hemisphere summer insolation during MIS 5e. During both terminations, SST increased simultaneously in the southern high latitudes and the tropical eastern Indian Ocean, suggesting virtually instantaneous atmospheric climate feedbacks between the high and low latitudes.