PP33A-0907 1340h
Glacial - Interglacial Changes in the Provenance and Flux of Sediment to the Cape Basin, South Atlantic: Implications for Indian-Atlantic Ocean Exchange
Deep sea sediment cores have been widely used to study changes in sea surface conditions and ocean circulation on glacial-interglacial timescales, using faunal and floral analyses, stable isotope and trace element contents of carbonate shells, and more recently subtle compositional variations in organic compounds. Radiogenic isotope ratios can be sensitive tracers of sediment provenance, and have been interpreted in terms of changes in circulation patterns. Another tool that has seen growing application for studying the past oceans is $^{230}$Th$_{xs}$$\deg$. The excess $^{230}$Th measured in the sediment is useful as a constant-flux proxy, and puts additional constraints on the lateral transport and redistribution of sediments. Until now provenance and flux measures have been applied to separate problems, and this study is unique in using the combination of these two tracers. The combined measurements of sediment provenance and flux has great potential for determining sediment transport patterns, and how they have changed through time. The southeast Atlantic is an important study site because it is a region where thermocline and intermediate waters are exchanged with the Indian Ocean by the "Agulhas Leakage". This leakage is thought to be an important source of heat and salt to the Atlantic Ocean, possibly acting as a positive feedback for the formation of North Atlantic Deep Water (NADW). Changes in the strength of NADW formation and the intensity of global thermohaline circulation (THC) have been called upon as major amplifiers for climate change during the Pleistocene glacial cycles, with the hypothesis that NADW formation was weaker during the cold periods of the glacial cycles, particularly during the Last Glacial Maximum (LGM), approximately 18,000 years ago. $^{87}$Sr/$^{86}$Sr, \epsilon $_{Nd}$ and $^{230}$Th- normalized terrigenous flux measurements from Holocene and LGM sediment samples support the Agulhas Current as a major source of sediment to the Cape Basin drift deposit. The results are consistent with 3 end- member mixing between sediment carried by the Agulhas Current, the South Atlantic or Antarctic Circumpolar Current (SAC or ACC) and a local source. A comparison of the two time- slices implies that a smaller proportion of sediment deposited in the Cape Basin is derived from the Agulhas Current during the LGM. The results are consistent with a strengthened and more sediment-laden ACC during the LGM, and possibly a reduced Agulhas leakage, and hence less interocean exchange. Further investigation is needed to constrain whether the Agulhas Current was different during the LGM.
PP33A-0908 1340h
Distinguishing Between Provenance Changes and Sorting Effects on the Rb-Sr Systematics in Glacial and Holocene South Atlantic Sediments
The results of recent studies show that the Strontium isotope composition of Holocene sediments in the South Atlantic and western Indian Oceans follow systematic geographic variations consistent with the distribution of modern surface currents. Our conclusion based on these results is that surface and intermediate, rather than deep currents are the first order means of transporting terrigenous sediments in this region, and studies of the terrigenous composition in marine sediment cores should yield insights into past ocean circulation. The two major sediment sources for the eastern South Atlantic can be identified by their Sr isotopic signatures, and appear to be transported by the Agulhas Current and the Antarctic Circumpolar Current (ACC) or South Atlantic Current (SAC). Although the Rb/Sr isotope system is a reliable measure of the provenance of terrigenous sediments, it is also known to be influenced by sorting. The goals of this study are to characterize the grain size compositions of the end-member sediment sources in both Holocene and last glacial maximum (LGM) aged core samples, and to evaluate the relative effects of provenance changes and sorting effects on the Rb/Sr system in this region during both time periods. Additionally, the silt to clay ratios of sediments may be useful in determining deep current strength at the time of deposition, due to their different hydrodynamic behaviors. Deep-sea cores from the Atlantic sector of the Southern Ocean and from the Mozambique Channel were selected to represent sediments deposited by the ACC and Agulhas Current, respectively. Samples were taken depths thought to represent Holocene and LGM. The core samples, as well as one sample taken from the Orange River in South Africa, were sieved to remove the coarse ($>$63 micron) fraction. The carbonate, oxide and opal components were leached from the remaining fine-fraction. Sub-samples of the leached fine-fraction sediments were separated into silt (2-63 micron) and clay ($<$2 micron) fractions according to Stoke's law of settling. Sr isotope ratios and Rb/Sr ratios were measured in the lithogenic $<$63 micron fraction, as well as the in silt and clay fractions of each sample. The silt/clay ratios range from 0.3 to 3.6, and are generally higher in the west and lower in the east. The silt/clay of the Orange River is 4.6, higher than any of the marine sediment samples. There is no clear pattern of change in silt/clay between the Holocene and the LGM. Sr isotope ratios are consistently higher in the clay fractions than in the silt. Sr isotope ratios of both the silt and clay fractions of the Agulhas samples are higher than those of the South Atlantic samples for both Holocene and LGM sediments. Rb/Sr isotope systematics for the slit and clay is consistent with sorting of both Holocene and LGM sediments from the Agulhas region, yielding Rb/Sr apparent ages that are higher for the silts than for the clays, and range from 256 Ma to 1.25 Ga. These results imply that most of the variability in the Rb-Sr systematics in South Atlantic sediments is due to differing provenance, rather than to systematic changes in the degree of sorting of a single sediment source. A more detailed look at the Rb-Sr systematics coupled with a study of Sm-Nd systematics in these samples may help us to distinguishing between provenance changes and sorting effects in South Atlantic Sediments.
PP33A-0909 1340h
Glacial-Interglacial Nd, Hf and Pb Isotope Compositions of North Atlantic Deep Water Extracted From the Authigenic Fraction in Pelagic Sediments
The magnitude of variation between glacial and interglacial modes of North Atlantic Deep Water (NADW) formation is under debate. A variety of proxies have been applied to the reconstruction of its flow and export to the Southern Ocean (e.g., stable and radiocarbon isotopes, 231Pa/230Th, Nd isotopes). These indicate substantial reduction in during the Last Glacial Maximum (LGM) and Younger Dryas (YD) and possibly a total shutdown of NADW formation during Heinrich event 1. We extracted the seawater-derived radiogenic isotope composition from the Fe-Mn oxyhydroxide fraction in pelagic sediments along the Blake Ridge, Western N Atlantic, spanning the interval from the LGM to the Holocene. The radiogenic isotope composition (Nd, Hf and Pb) of a water mass is a reflection of a function of the age and lithology of the continental source material and the degree of subsequent mixing with other water masses. The degree of incongruent weathering is also important, particularly for Hf. These tracers allow the three major water masses along the Blake Ridge (i.e., mid-latitudinal warm surface water, NADW and Antarctic Bottom Water) to be resolved. The radiogenic isotope composition of ambient seawater at several core locations from different water depths was determined to trace glacial-interglacial changes in the isotopic composition with depth of the Deep Western Boundary Current (DWBC), which transports NADW to the South Atlantic. The results allow a better understanding of the increased inflow of Antarctic Bottom Water during the LGM and the YD at the deepest locations along the Blake Ridge as well as the migration of the DWBC upslope during colder periods, and the location and strength of glacial NADW formation. The Hf isotope compositions can be used to trace the enhanced glacial mechanical weathering intensity in Northern Canada and Greenland during the LGM.
PP33A-0910 1340h
A 475 kyr record of extraterrestrial $^3$He and $^{230}$Th in North Atlantic sediments: caveats to derive MAR from these tracers.
Extraterrestrial $^3$He is delivered to the earth surface by interplanetary dust particles (IDPs) characterized by a very high $^3$He/$^4$He ratio (2.4$\times$10$^{-4}$), several orders of magnitude higher than ratios measured in any terrigeneous matter. Several studies in the past ten years have opened up the potential of IDP-derived extraterrestrial $^3$He as a constant flux proxy: the global average flux is around (0.8-1)$\times$10$^{-12}$ cm$^3$ STP cm$^{-2}$ kyr$^{-1}$ over at least the last 200 kyr, and probably on a longer timescale (e.g. Marcantonio et al 2001, Paleoceanography 16, 260- 267; Higgins 2001, Columbia Univ. PhD; Winckler et al 2004, QSR 23, 1873-1878). Since $^3$He is a stable isotope, this method would not be limited by radioactive decay as it is the case for $^{230}$Th. However most of the studies on extraterrestrial $^3$He in deep-sea sediments have been conducted in regions of low sedimentation rate and specially of low terrigeneous supply. About sixty samples spanning the 475 kyr SU92-10 core from the North Atlantic Ocean (44$\deg$5N, $24\deg$5W, 3108m depth) have been selected to study helium and thorium isotopes in less favourable conditions: sedimentation rate around 1cm kyr$^{-1}$, high terrigeneous input, winnowing suspected. $^3$He/$^4$He ratios between 5.7$\times$10$^{-7}$ and 7.7$\times$10$^{-9}$ have been measured: in such case terrigeneous $^3$He cannot be neglected and deconvolution of the different components of the helium signal is necessary. Another consequence of the low concentration of IDPs is the poor reproducibility of the measurements preventing the interpretation of any fine structure of the $^3$He signal. The helium carrier phase in IDPs is not precisely known but Amari and Ozima (1985, Nature 317, 520-522) found that the helium resides mainly in the magnetic fraction of the sediments. We proceeded to systematic magnetic separations of 1g of decarbonated sediment to measure separately the magnetic fraction (less than 2% in mass but up to 80% of the $^3$He) and several aliquots of the non-magnetic fraction with a standard deviation improved by a factor 2 to 10 compared to replicates of the bulk decarbonated sediment. The first results show a cyclic pattern in the helium isotopic ratio, specially of the magnetic fraction, which, if confirmed, could be linked to variations in the detritic supply. Using MAR derived from $^{18}$O stratigraphy, a low cosmic $^3$He burial rate of (2.7 $\pm$ 1.0)$\times$10$^{-13}$ cm$^3$ STP cm$^{-2}$ kyr$^{-1}$ has been calculated. The $^{230}$Th$_{xs}^{\circ}$ measurements suggest an even larger winnowing effect at this core location. However work is still going on to interpret the thorium data: with a maximum of 2.2 dpm g$^{-1}$ for $^{230}$Th, the activities measured are very low making the $^{230}$Th$_{xs}^{\circ}$ determination all the more difficult. Among the hypotheses we are working on to explain thorium anomalies, one is the loss of $^{234}$U by diffusion out of the sediments and the role of the carbonates (between 40% and 97%) towards the U-Th isotopes.
PP33A-0911 1340h
Differentiation of Cenozoic Eolian Dust Sources in the Eastern Pacific by Nd-Sr-Pb Radiogenic Isotopes
DSDP, ODP and piston core samples from the eastern Pacific provide a temporal and spatial context for mapping changing sources of eolian dust during the Cenozoic Era (65 Ma to present). We examined core sites from ODP Sites 469, 495, 853, 885/886, DSDP Sites 32, 319, 320, and piston core sites EW9709-01, 07 for Pb, Sr and Nd isotopic provenance. Isotopic analyses were performed on the $<$63 micron extracted detrital component following reductive cleaning to remove all authigenic and biogenic components. Downcore comparison of continental margin sites with those dominated by pelagic clays from north and south of the equator was facilitated by factor analysis using two programs. Multivariate factor analysis was performed with both CABFAC and Minitab. Three factors were identified and accounted for 98.4% of the variance. Factor 1 (epsilon Nd and Pb 207/206) defines the dominant late Cenozoic Asian dust source and accounts for 44.1% of the variance. Factor 2 identifies a South American dust source based on $^{87}$Sr/$^{86}$Sr and Pb 206/204, 208/204 and 207/204 and accounts for 36.5% of the variance. Factor 3 identifies a North American dust source ($^{87}$Sr/$^{86}$Sr, Pb 208/206, and Pb 207/206) and accounts for 17.8% of the variance. We conclude that in the eastern Pacific, late Cenozoic Asian dust sources can be quantitatively differentiated from both North American and South American continental sources. We are attempting to extend this record back to 65 Ma, with the ultimate goal being to reconstruct long-term Cenozoic atmospheric circulation and wind patterns across major climate transitions.
PP33A-0912 1340h
Osmium Isotope Straigraphy of Ferromanganese Crusts
To interpret the changes in isotopic compositions recorded in hydrogenetic ferromanganese (Fe-Mn) crusts over time it is essential to calibrate them in terms of time. The $^{10}$Be method is only reliable for the first 10 Myr. For older parts of the crusts the Co-constant flux method is used. Both approaches however, will fail to account for any growth hiatus or erosion in the sections older than 10 Ma. Attempts at using Sr isotope stratigraphy failed because of post-depositional exchange. For osmium (Os) isotopes on the other hand, calculations of the rate of post-depositional exchange suggest that long-term records in Fe-Mn crusts are reliable. This would allow the $^{187}$Os/$^{188}$Os profile of any hydrogeneous Fe-Mn crust to be fitted against the $^{187}$Os/$^{188}$Os seawater record established for the last 80 Myr. This stratigraphic method would determine the age of crusts at any depth and identify changes in growth rate, cessation of growth and/or intervals of crust erosion. We tested this hypothesis on the hydrogeneous crust CD29-2 from the Central Pacific Ocean which had been subject to many previous radiogenic isotope studies. CD29-2 is a 105mm thick crust with a growth rate of 2.1mm/Myr, as determined from $^{10}$Be/$^{9}$Be ratios and the Co-constant flux method. This gives a minimum age of 50 Ma for the lowermost portions of the crust. Samples were taken every 2mm through the crust which results in a time-spacing of 1Myr assuming a constant growth. For each sample the $^{187}$Os/$^{188}$Os ratio and the $^{187}$Os concentration ([$^{187}$Os]) were determined by ID-NTIMS. The [$^{187}$Re] was measured by MC-ICPMS, allowing correction for $^{187}$ Re-decay. The corrected $^{187}$Os/$^{188}$Os ratios were compared to the seawater record. Using the Be and Co time scales, the $^{187}$Os/$^{188}$Os curve obtained from the crust shows a distorted version of the established seawater record. A good match is found if three hiatuses are allowed. The first hiatus of 15 Myr is assigned to the period between 13 and 28 Ma, a second one of 3 Myr to the period between 32 and 35 Ma and a last one of 4 Myr to the period between 35 and 39 Ma. The total elapsed time in these hiatuses is 22 Myr, which increases the age of the base of the crust from 50 to 72 Ma. The crust record thus covers the K-T boundary at 65 Ma where it shows the now well known Os isotopic drop. This new dating technique is the first one to date reliably hydrogenetic Fe-Mn crusts older than 10 Ma and is only limited in time by the oldest data of the $^{187}$Os/$^{188}$Os seawater record, 80 Ma at present. It confirms the dating obtained in CD29-2 by the Be method for the first 10 Myr.
PP33A-0913 1340h
Controls on $^{231}$Pa/$^{230}$Th in the Indian Ocean: Circulation or Productivity?
($^{231}$Pa$_{xs}$/$^{230}$Th$_{xs}$)$_{0}$ ratios in marine sediments are a potential proxy of palaeoproductivity and/or past ocean circulation. Studies in the Atlantic have demonstrated the particular potential of this proxy to assess the rate of past circulation an important environmental variable which is otherwise difficult to reconstruct $^{[1, 2]}$. In this study, we present new water-column and sediment data from the western Indian Ocean to improve understanding of the controls on ($^{231}$Pa$_{xs}$/$^{230}$Th$_{xs}$)$_{0}$ and test the use of this proxy. The Indian Ocean differs from the Atlantic because deep waters forming in the Southern Ocean and transported into the basin have high rather than low Pa and Th concentrations $^{[3]}$. The ability to reconstruct past rates of flow into the Indian Ocean would be a powerful new use of Pa/Th, but no data presently exist to test this application. Water samples (10 litre), surface sediments, and a 5.5m Kasten core were collected on the Charles Darwin cruise CD154 from the area East of Madagascar. The down-core record, from10$\circ$S extends to 150 ka, and is coupled with a stable isotope stratigraphy and a sortable silt record. Water and sediment samples were analysed for isotopes of Pa, Th and U on a Nu Instruments MC-ICP-MS. The water-column dataset (26 samples) indicate that Pa concentrations are high in NADW transported into the Indian Ocean from the Atlantic, while opal scavenging has lowered Pa concentrations in deeper water masses. Sedimentary $^{230}$Th data indicate that there has been significant sediment focusing and that sediment mass fluxes are low. Preliminary sedimentary Pa/Th data show little relation to the sortable silt record. Ongoing Pa/Th and Ba data will complete this downcore record and allow comparison of a Pa/Th record with independent proxies of current speed and productivity in the same core. [1] E.F. Yu, et al, Nature 379, 689-694, 1996. [2] J.F. McManus, et al, Nature 428, 834-837, 2004. [3] M. M. Rutgers van der Loeff, and G. W. Berger. Deep-Sea Research Part I: Oceanographic Research Papers. 40(2): 339-357, 1993.
PP33A-0914 1340h
An Excess $^{226}$Ra Geochronology for Saanich Inlet
When examining past environmental and climatic change the ability to date both freshwater and marine sediments is essential. Numerous methods are currently available, however where present techniques fail, novel dating techniques are required to overcome these obstacles. In particular the use of $^{14}$C to date sediments over the Holocene period can be severely limited where there are significant hard-water and reservoir effects and where reworking of organic matter occurs. The use of unsupported or excess $^{226}$Ra offers a potential dating tool in such depositional contexts over the Holocene time period as a result of its half-life of 1600 years. Indeed, the technique has been applied to a number of environments since first being proposed by Koide et al. (1976). However, numerous question marks still surround the application of the technique, particularly with reference to the dating of sub-aqueous sediment deposits. As part of wider research into the applicability of excess $^{226}$Ra to date lake sedimentary environments, deep-sea samples from Saanich Inlet were obtained from the ODP (Leg 169S) in order to validate the dating technique. Saanich Inlet is an anoxic fjord located on the coastline of Vancouver Island, British Colombia, the sediments of which have been shown to be comprised of lamina-scale alternations between diatom-rich and diatom-poor clays and silts representing annual. Using digital sediment colour analysis Nederbragt and Thurow (2001) constructed a varve chronology for the last 6000 years. Alpha-spectrometry was successfully used to measure $^{226}$Ra, $^{230}$Th and $^{234}$U in sediment samples collected from Saanich Inlet during ODP Leg 169S. These activities were used to determine an excess $^{226}$Ra geochronology for Saanich Inlet. This radiometric chronology compares well with dates obtained from the varve record, and spans two different phases of sediment accumulation.
PP33A-0915 1340h
xs$^{231}$Pa/xs$^{230}$Th ratios in Northeastern Arabian Sea Sediments: Tracing Changes in Thermohaline Circulation over the past 50 ka?}
Patterns of millennial changes in productivity and eolian fluxes in the northeastern Arabian Sea are related to stadial-interstadial temperature variations observed in the Greenland ice cores. In an effort to constrain further the relationship between climate at low and high latitudes, we present new xs$^{231}$Pa/xs$^{230}$Th results from northeastern Arabian Sea sediments (core 93KL) spanning the past 50 ka. The unique difference in residence time between oceanic $^{231}$Pa and $^{230}$Th has allowed the xs$^{231}$Pa/xs$^{230}$Th activity ratio in sediments from ocean basins to be used as a proxy for various paleoclimatic indicators such as particle flux, particle composition, and thermohaline circulation. Over the past 50 ka, initial xs$^{231}$Pa/xs$^{230}$Th activity ratios in sediment from core 93KL are maintained at values significantly higher than the water-column production ratio of 0.093, ranging from 0.138 ($\sim$16 ka) to 0.206 ($\sim$4.3 ka). Average xs$^{231}$Pa/xs$^{230}$Th activity ratios are lower during the last glacial period than during the Holocene. In addition, there are sub-Milankovitch variations superimposed on the pattern of glacial-interglacial variability that are consistent with the timing of North Atlantic climate events. The most prominent feature of the pattern is that the lowest xs$^{231}$Pa/xs$^{230}$Th activity ratios coincide with the timing of Heinrich events 1 through 5. In contrast, the highest ratios are usually associated with interstadials. Increased productivity and boundary scavenging seem to exert some control on the variability we observe in the xs$^{231}$Pa/xs$^{230}$Th activity ratios. However, changing patterns of ocean thermohaline circulation (THC) also seem to influence these ratios. Specifically, low xs$^{231}$Pa/xs$^{230}$Th activity ratios during the Heinrich events may be caused by a slowdown in THC, which leads to a lower supply of Pa to our site in the northeastern Arabian Sea.
PP33A-0916 1340h
Do Sm/Nd- and Sr-Isotopic Signals in Lake Baikal Sediments Reflect Shifts in Atmospheric Flow Patterns of Central Eurasia During Termination I ?
Loess deposits give insights into past climatic and atmospheric conditions, as mineral dust entrainment, transport and deposition is related to humidity distribution (vegetation), wind speed and wind direction. To resolve changing atmospheric conditions and mineral aerosol transport in the northern part of Central Eurasia during the last 20 ka (LGM to present) we investigated a 2 metre long sediment core from Lake Baikal to trace provenance of the inorganic aeolian component. The location of the Lake Baikal sediment core, on a horst structure with hemipelagic sedimentation, is ideally positioned to serve as a dust trap. These sediments show a bimodal grain-size distribution, separating the silt and the clay fraction with the coarse mode falling into the range of 12-14 $\mu$m. While the mode of the coarse fraction remains rather constant with only minor variations during the investigated time interval, the silt/clay-ratio increases from the glacial to the Holocene, together with a parallel increasing proportion of minerals $>$ 20 $\mu$m. This is in contrast to Chinese loess, where the ratio of coarse to fine mode fractions is increased during the glacial periods. Historical reports and TotalOzoneMappingSpectrometer (TOMS) observations confirm that enormous dust events, evolving from the Taklimakan desert and Siberia to the West of the lake, occasionally carry a huge amount of mineral dust in the atmosphere above Lake Baikal. The grain-size distribution of aeolian deposits from the Chinese loess plateau to the Pacific shows a systematic eastward decrease in maximum particle size and thus a decreasing coarse mode of event dust with increasing source distance. Given the distance from the Gobi desert to Lake Baikal at maximum a 12-15 $\mu$m mode fraction is expected to reach the Lake Baikal core location. Modern aerosols, collected on the ice-cover of Central Lake Baikal in early spring 2001 exhibit a monomodal size distribution with a mode of 9 $\mu$m. This is in agreement with a minerogenic aerosol size distribution after several 100 kilometres of transport and supports an aeolian origin of the coarse mode fraction found in the lake sediments. To trace provenance, we used the Sm/Nd-isotopic systematics in combination with the Sr-isotopic ratios and geochemical and mineralogical information. Although trace element characteristics indicate a dominant contribution from Lake Baikal hinterland to the coarse mode fraction, climate change is well documented in the isotopic signature of both the clay and silt fraction, with the principal shift occurring during Termination I. We argue that these isotopic signals reflect a change in sediment provenance rather than a weathering related alteration of mineral associations.
PP33A-0917 1340h
High Resolution Record of Seawater Osmium Isotopes Over the Last 100,000 Years
We present a high-resolution record of the osmium isotopic composition of oceans over the past 100,000 years. The main intent of this study is to investigate the claims that the Os isotopic composition of the deep oceans varies on a glacial-interglacial timescale and consequently the mean residence time of Os in the oceans is between 8 and 10 ka (Oxburgh, 1998; 2001), about a factor of four to five less than that estimated from mass balance calculations (Sharma et al., 1997, 1999; Levasseur et al., 1999; Peucker-Ehrenbrink, 2002). At present, it is commonly believed that Os in seawater is derived from continents (rivers), submarine alteration (abyssal peridotites?), and micrometeorites. A shorter residence time would require the need to search for a yet another substantial source of Os to the deep ocean. We have measured Os isotopes in Fe-Mn fraction of core TNO57-21, which is a piston core recovered from ODP site #1089, located in the Cape Basin on the Agullhas drift ($41\deg$S, $7\deg$ E, depth = 4825 m). This site is unique in that it experienced extremely high sedimentation rates with little input of continental dust or volcanic ash. Therefore, the sediment is composed entirely of authigenic clays and carbonates. Osmium sequestered in the Fe-Mn portion of the core should yield the $^{187}$Os/$^{188}$Os ratio of ambient seawater. If Os has a residence time of 8-10 ka, we would expect to observe glacial-interglacial variations in our record reflective of variations in the intensity of continental weathering. If, on the other hand, the residence time is ~40 ka, as predicted by mass-balance equations, then the Os isotopic composition should remain essentially constant throughout the past glacial-interglacial cycle. Our data for the TNO57-21 show substantial variation during the past 100 Ka extending into the Holocene. Intriguingly, during MIS 4 and MIS 2 the isotopic composition of the oceans was more radiogenic than during much of MIS 3. Our record is relatively stable with $^{187}$Os/$^{188}$Os ratio of ~1.04 during MIS 5, 4 and 2, but shows significant variability during MIS 3: at 56 ka the $^{187}$Os/$^{188}$Os rises from 1.04 to the modern seawater composition of 1.06. It drops to a ratio of 1.0 at 40 ka and then rises to 1.04 at 36 ka. We also observe a rise in the Os isotopic composition from ~1.04 in MIS2 to ~1.06 in the most recent Holocene samples, a magnitude of change consistent with that observed by Oxburgh (1998). However, our record shows a much more recent transition to the modern seawater composition, a transition continuing into the present. It further suggests that Os is not in steady state and may be subjected to climate forcing. Indeed, Lomb-Scargle Spectrum analysis of the record shows the presence of 27 ka and 18 ka wavelengths, which have normalized spectral powers of 4 and 3.5, respectively. Further analyses are underway to understand the cause of this variation and to obtain a more complete record for the Holocene.
PP33A-0918 1340h
Deep-sea Coral Record of Paleoventilation Age of North Atlantic Deep Water During the Last Deglaciation
Wunsch (2003) posed a credible challenge to the paleoclimate community by suggesting that over zealous wiggle matching of the Greenland ice core \delta $^{18}$O records to far field proxy records grossly exaggerates the geographic relevance of the Greenland's millennial scale variability. Wunsch (2003) argues that regional wind fluctuations are a more reasonable cause of the millennial scale \delta $^{18}$O fluctuations (Charles et al., 1994) as well as other tracers trapped in Greenland ice. He asserts that the rapid fluctuations measured in Greenland ice cores need not be hemispheric or global in scope, nor be due to changes in North Atlantic Deep Water (NADW) production rates as is widely speculated. In order to test the proximal influence of the Greenland climate signal on NADW production rates, we estimate the NADW ventilation age at Orphan Knoll ($50\deg$26'N, $46\deg$22'W) using the radiocarbon tracer measured in deep-sea corals. Orphan Knoll is optimally located to record $\Delta$$^{14}$C variability on short timescales and deep-sea corals are abundant. Thus far we have calculated 400 year ventilation ages of NADW during the Bolling-Allerod interval (14.7-12.9 ka) using the $^{14}$C projection method (Adkins and Boyle, 1997), ventilation ages indistinguishable from modern estimates. These ventilation estimates are based on projections to our $^{230}$Th/$^{234}$U dated surface water coral $\Delta$$^{14}$C record from the tropics.