Stable isotope record, Mg/Ca ratios and sea surface reconstruction for sediment core MD06-3067

We present sea surface, upper thermocline, and benthic d18O data, as well as temperature and paleoproductivity proxy data, from the International Marine Global Change Study Program (IMAGES) Core MD06-3067 (6°31'N, 126°30'E, 1575 m water depth), located in the western equatorial Pacific Ocean within the flow path of the Mindanao Current. Our records reveal considerable glacial-interglacial and suborbital variability in the Mindanao Dome upwelling over the last 160 kyr. Dome activity generally intensified during glacial intervals resulting in cooler thermocline waters, whereas it substantially declined during interglacials, in particular in the early Holocene and early marine oxygen isotope stage (MIS) 5e, when upwelling waters did not reach the thermocline. During MIS 3 and MIS 2, enhanced surface productivity together with remarkably low SST and low upper ocean thermal contrast provide evidence for episodic glacial upwelling to the surface, whereas transient surface warming marks periodic collapses of the Mindanao Dome upwelling during Heinrich events. We attribute the high variability during MIS 3 and MIS 2 to changes in the El Niño Southern Oscillation state that affected boreal winter monsoonal winds and upper ocean circulation. Glacial upwelling intensified when a strong cyclonic gyre became established, whereas El Niño-like conditions during Heinrich events tended to suppress the cyclonic circulation, reducing Ekman transport. Thus, our findings demonstrate that variations in the Mindanao Dome upwelling are closely linked to the position and intensity of the tropical convection and also reflect far-field influences from the high latitudes.

Tropical climate and vegetation simulations during the Heinrich event 1 using an Earth System Model of Intermediate Complexity (EMIC) - the University of Victoria Earth System-Climate Model (UVic ESCM)

Abrupt climate changes from 18 to 15 thousand years before present (kyr BP) associated with Heinrich Event 1 (HE1) had a strong impact on vegetation patterns not only at high latitudes of the Northern Hemisphere, but also in the tropical regions around the Atlantic Ocean. To gain a better understanding of the linkage between high and low latitudes, we used the University of Victoria (UVic) Earth System-Climate Model (ESCM) with dynamical vegetation and land surface components to simulate four scenarios of climate-vegetation interaction: the pre-industrial era, the Last Glacial Maximum (LGM), and a Heinrich-like event with two different climate backgrounds (interglacial and glacial). We calculated mega-biomes from the plant-functional types (PFTs) generated by the model to allow for a direct comparison between model results and palynological vegetation reconstructions. Our calculated mega-biomes for the pre-industrial period and the LGM corresponded well with biome reconstructions of the modern and LGM time slices, respectively, except that our pre-industrial simulation predicted the dominance of grassland in southern Europe and our LGM simulation resulted in more forest cover in tropical and sub-tropical South America. The HE1-like simulation with a glacial climate background produced sea-surface temperature patterns and enhanced inter-hemispheric thermal gradients in accordance with the "bipolar seesaw" hypothesis. We found that the cooling of the Northern Hemisphere caused a southward shift of those PFTs that are indicative of an increased desertification and a retreat of broadleaf forests in West Africa and northern South America. The mega-biomes from our HE1 simulation agreed well with paleovegetation data from tropical Africa and northern South America. Thus, according to our model-data comparison, the reconstructed vegetation changes for the tropical regions around the Atlantic Ocean were physically consistent with the remote effects of a Heinrich event under a glacial climate background.

Results of Heinrich event 1 simulation from the University of Victoria Earth System-Climate Model (UVic ESCM) and the Community Climate System Model (CCSM3)

We investigated changes in tropical climate and vegetation cover associated with abrupt climate change during Heinrich Event 1 (HE1, ca. 17.5 ka BP) using two different global climate models: the University of Victoria Earth System-Climate Model (UVic ESCM) and the Community Climate System Model version 3 (CCSM3). Tropical South American and African pollen records suggest that the cooling of the North Atlantic Ocean during HE1 influenced the tropics through a southward shift of the rain belt. In this study, we simulated the HE1 by applying a freshwater perturbation to the North Atlantic Ocean. The resulting slowdown of the Atlantic Meridional Overturning Circulation was followed by a temperature seesaw between the Northern and Southern Hemispheres, as well as a southward shift of the tropical rain belt. The shift and the response pattern of the tropical vegetation around the Atlantic Ocean were more pronounced in the CCSM3 than in the UVic ESCM simulation. For tropical South America, opposite changes in tree and grass cover were modeled around 10° S in the CCSM3 but not in the UVic ESCM. In tropical Africa, the grass cover increased and the tree cover decreased around 15° N in the UVic ESCM and around 10° N in the CCSM3. In the CCSM3 model, the tree and grass cover in tropical Southeast Asia responded to the abrupt climate change during the HE1, which could not be found in the UVic ESCM. The biome distributions derived from both models corroborate findings from pollen records in southwestern and equatorial western Africa as well as northeastern Brazil.

Biomarker in Heinrich event layers

There are controversies regarding the origin of Heinrich layer 3 (H3), the massive ice-rafting and meltwater event in the North Atlantic during the last glacial cycle spanning a time window between 29 and 30 kyr B.P. Some argue in favor of a Laurentide Ice Sheet source similar to other Heinrich layers, while a contending view argues for the European ice sheet source. Existing geochemical proxies such as 40Ar/39Ar, 206Pb/204Pb, or epsilon-Nd, etc., could not be used to distinguish among various sources of ice-rafted debris in H3 because of their low abundances, suggesting a background glacial sediment signal. In order to circumvent this problem a biomarker-based approach is used to characterize the provenance of H layers 2, 3, and 4 and other non-Heinrich layers. The presence of hopanes and steranes and their aromatic counterparts in the H layers is incompatible with Recent sediments and is attributed to the transportation of organic matter because of the glacial erosion of source rocks. The most diagnostic and useful signatures of this ancient organic matter in the H layers are the dominance of C34 hopanoids over C33 and the occurrence of isorenieratane along with palaerenieratane. Biomarkers signatures in H layers 2 and 3 of the Labrador Sea suggest no difference in their source. Hydrocarbon distributions suggest that these sediments were derived from the Middle to Late Ordovician and Silurian source rocks of the Hudson Bay of eastern Canada. Biomarker data of the H layer 4 from the northwest Atlantic reveal that the sediments of this layer have a similar source to the H layers in the Labrador Sea.

(Appendix A) Oxygen and carbon isotope results of detrital carbonate grains of Heinrich layers from HU90-013-028 and IODP Sites 303-U1302/U1303/U1308 in the North Atlantic Ocean

We report oxygen and carbon isotope results of detrital carbonate grains from Heinrich layers at three sites in the North Atlantic located along a transect from the Labrador Sea to the eastern North Atlantic. Oxygen isotopic values of individual detrital carbonate grains from six Heinrich layers at all sites average - 5.6 ppm ± 1.5 ppm (1sigma; n = 166), reflecting values of dolomitic limestone derived from source areas in northeastern Canada. The d18O of bulk carbonate at Integrated Ocean Drilling Program (IODP) Site U1308 (re-occupation of Deep Sea Drilling Project (DSDP) Site 609) in the eastern North Atlantic records the proportion of detrital to biogenic carbonate and d18O decreases to - 5 ppm during Heinrich (H) events 1, 2, 4 and 5 relative to a background value of ~ 1 to 2 ppm for biogenic carbonate. Bulk d18O also decreases during H3 and H6 but only attains values of - 1ppm, indicating either a greater proportion of biogenic-to-detrital carbonate or a different source. Because the d18O of detrital carbonate is ~ 9 ppm lower than foraminifer carbonate, any fine-grained detrital carbonate not removed from the inner test chambers will lower foraminifer d18O. We conclude bulk carbonate d18O is a sensitive proxy for detrital carbonate and may be useful for identifying Heinrich layers in cores within and near the margins of the North Atlantic ice-rafted detritus (IRD) belt.