PP32A-01 10:20h
Decadal and Lower Frequency South Pacific Climate Variability Since 1619 AD from Replicated Coral Records
A notable aspect of multi-century oxygen isotope (\delta$^{18}$O) time-series generated from modern Pacific Porites corals in the convergence zone is the presence of a trend component of progressively lower \delta$^{18}$O in the top (younger) sections of most published \delta$^{18}$O series. The climatic significance of this trend has remained controversial in part due to the paucity of multi-century long coral records. Here we present sub-annually resolved and replicated Porites \delta$^{18}$O records from Fiji (17S, 179E) (1619-2001AD, replicated from 1780) and Rarotonga (21.5S, 160W) (1726-1997AD; replicated from 1874) in the southwestern Pacific to evaluate the significance of the \delta$^{18}$O trend in this region. As part of this study we also analyzed bulk skeletal Sr/Ca from each of the 5 coral cores (2 subannual, 3 annually averaged). The coral \delta$^{18}$O series from Fiji and Rarotonga document site reproducible trends (similar timing within each site, and total magnitudes of 0.30 to 0.35 per mil since 1850AD) towards progressively lower and unprecedented \delta$^{18}$O in the late 20th century. However, the Sr/Ca records from the same cores do not replicate completely before 1950AD, indicating that at these sites, and in some corals, Porites bulk skeletal Sr/Ca is not strictly a function of SST or external-to-the-coral environmental variability. For skeletal \delta$^{18}$O, the overall reproducibility of the \delta$^{18}$O trend in different age corals at each site, supports a locally consistent, primarily environmental origin for this mode of \delta$^{18}$O variability. Comparison of the \delta$^{18}$O trend modes at Fiji and Rarotonga to each other and to instrumental SST and precipitation data suggests that the coral \delta$^{18}$O trend at each site is due to regionally variable but progressive warming and salinity reduction. If this interpretation is correct, the second half of the 20th century was the warmest and least saline at both sites since the early 1600s. Since Fiji and Rarotonga are both influenced to varying degrees by the South Pacific Convergence Zone (SPCZ) one explanation for the \delta$^{18}$O trend is that the SPCZ has been intensifying over the last 200 years with increasing cloud cover and rainfall as the surface ocean warmed. On decadal-interdecadal time-scales, comparison of the Fiji and Rarotonga coral \delta$^{18}$O series to other coral \delta$^{18}$O records from New Caledonia and the Great Barrier Reef indicates that some interdecadal climate shifts apparently were related to changes in the SPCZ and others apparently were unrelated to the SPCZ. This observation suggests the possibility that decadal-interdecadal climate variability in the South Pacific has multiple sources, and may at times be related to higher latitude South Pacific processes.
PP32A-02 10:35h
Select, High-Resolution Windows Into Sub-Centennial-Scale Climate Variability in the Western Pacific Warm Pool Between 7 and 12 ka
Post-glacial, coral-based climate records from the Western Pacific Warm Pool (WPWP), a region that serves as a major heat and moisture source to the ocean-atmosphere system, provide sub-annually resolved windows into climate variability on interannual to multi-decadal timescales from this climatologically significant region. Paleoclimate reconstructions based on fossil corals require that the skeletal geochemistry be unaffected by diagenesis and that secular changes in seawater chemistry be known. Global seawater \delta$^{18}$O variations can be constrained using knowledge of past variations in ice volume, whereas much less is known about global seawater Sr/Ca variations, if they occur in the post-glacial interval. Our paleoclimate reconstructions are based on monthly resolved \delta$^{18}$O and Sr/Ca records in fossil Porites corals from the Western Solomon Islands ($\sim$$8\deg$ S, $\sim$$157\deg$ E; Tetepare and Rendova). Post-depositional alteration of our fossil coral samples is minimal based on mineralogic (XRD), petrographic (SEM) and geochemical criteria (preservation of modern marine initial \delta$^{234}$U values). Four of these fossil coral time series are of particular interest: 99RND (age, 7,992$\pm$42; ~45 years), 01T-B (age, 7,647$\pm$73; ~65 years), 01T-AQ (age, 10,208$\pm$44; ~30 years), and 99TET-B (age, 11,987$\pm$69; ~ 30 years). We apply a model that uses simultaneous variations in coral \delta$^{18}$O and Sr/Ca in combination with estimates of post-glacial changes in seawater chemistry to reconstruct mean climate state during the early Holocene and the Younger Dryas. Model results indicate that on average SSTs in the WPWP were within $1\deg$C of modern and that surface waters were more saline than modern during each of the four time intervals during which our fossil corals grew.
PP32A-03 10:50h
Radiocarbon Calibration Between 30,000 and 50,000 Years Before Present Using Fossil Corals
Early attempts to calibrate the radiocarbon time scale using lake sediments, deep sea cores, and cave flowstones relied on assignment of calendar ages based on correlations to other time series or modeled ages (Kitagawa and van der Plicht, 2000; Voelker et al., 2000; Hughen et al., 2004) or modeled ground water chemistry (Beck et al., 2001). These calibrations are too disparate to be adopted for routine radiocarbon conversion. Fossil corals are considered an excellent archive for extending the calibration beyond the tree ring records because corals can be directly dated by both U-series and radiocarbon methods. Coral data sets, however, are not consistent between 30,000 and 50,000 years before present (Yokoyama et al., 2000; Cutler, 2000) and show considerable scatter. Our experiments show that much of the scatter in corals can be explained by minor amounts of diagenetic calcite and trace amounts of organics, thereby significantly biasing the radiocarbon dates toward younger ages. A greater than 50% etching of coral samples immediately preceding radiocarbon measurements is the accepted pretreatment for radiocarbon dating (Yokoyama et al., 2000). In fossil corals that were exposed to the vados or phreatic freshwater environments and contain minor amounts of diagenetic calcite, the acid etching pretreatment enriches the fraction of secondary calcite by 200% to 400%. Our experiments and simple model demonstrate the enrichment of secondary calcite shifts the age of the fossil coral hundreds to thousands of years younger. Previous studies set the lower limit of acceptable calcite in corals between 1% and 5% (Yokoyama et al., 2000; Cutler, 2000; Paterne et al., 2004) or report a detection limit of 1% (Bard et al., 1998). However, our model and experiments show that only corals with $<$0.2% calcite fall within the age uncertainty of radiocarbon dating after progressive etching. We demonstrate a detection limit of 0.2% calcite in aragonite can be routinely achieved by X-ray diffraction analysis, and adopt $<$0.2% calcite as an essential criteria for inclusion of coral U-series and radiocarbon data in our calibration data set. Another significant source of radiocarbon error in fossil corals that have been exposed to vados environments during lower sea level or due to uplift is trace organic matter that is not removed by the standard acid etching procedures. Our experiments show progressive whitening of coral samples over a 3-day period by ultrasonic cleaning in 30% peroxide and a corresponding increase in radiocarbon age due to the oxidation of acid resistant organic matter.
PP32A-04 11:05h
New constraints on Quaternary sea level oscillations provided by U-series measurements of a submerged speleothem from the Italian coastline
Speleothems have become increasingly important as tools to place long paleoenvironmental records into a temporal context through the use of U-Th dating, which is one of the most precise and accurate geochronometers available for the late Quaternary. This investigation was designed to provide U-series ages for a stalagmite collected from a submerged cave on Argentarola Island to constrain the timing, duration, and magnitude of Quaternary sea-level highstands. Argentarola cave has been alternately submerged and subaerially exposed as evinced by the presence of marine encrustations of {\it Serpulid} calcite that alternate with dense, microcrystalline speleothem calcite that forms during subaerial exposure. As such, speleothems from Argentarola cave provide a unique archive of the relative height of Quaternary sea level oscillations along a portion of the Italian coastline that appears to have been tectonically stable on the time scales considered here. Moreover, the application of U-series dating to these speleothems can provide an estimate of the absolute timing and relative duration of Quaternary marine transgressions. We have studied a stalagmite that was collected 18 meters below present sea level. Preliminary results indicate speleothem growth during stage 8, in addition to stages 2, 6 and 7.2 that have previously been identified, and define marine layers that correspond to the Holocene transgression and highstands associated with marine isotope stages (MIS) 5, 7.1, and 7.3. Because few relative sea level (RSL) indicators exist for glacial cycles prior to the last interglacial, data from Argentarola speleothems provide important benchmarks for future RSL models and also provide an important test for existing models. Our data are generally in good agreement with RSL curves for this time period and indicate that the correlation between RSL and foraminiferal \delta$^{18}$O over the last glacial cycle is a robust predictive tool for RSL estimates extending back to the initiation of MIS 8.
PP32A-05 11:20h
Structure of the Penultimate Deglaciation Along the California Margin and Implications for Milankovitch Theory
Several studies have suggested that the penultimate deglaciation occurred earlier than Milankovitch theory predicts therefore calling into question the orbital theory of ice ages. Furthermore, the detailed structure and timing of Termination II is insufficiently defined yet critical for understanding mechanisms responsible for abrupt regional and global climate change. High-resolution climate records and a means to accurately date them independent of orbital tuning are necessary to address these issues. Here we present planktonic and benthic foraminiferal oxygen isotope records at unprecedented resolution encompassing late MIS 6 and Termination II ($\sim$150 to 120 ka) from Santa Barbara Basin (ODP Site 893) supported by additional southern California margin records (Sites 1017 and 1014), a region highly sensitive to global millennial-scale climate oscillations during the last deglaciation. These records reveal millennial- and centennial-scale climate variability throughout the interval including a prominent interstadial immediately preceding the deglaciation, a brief warm event near the beginning of Termination II, and a B$\o$lling-$\AA$ller$\o$d/Younger Dryas-like climate oscillation midway through the deglaciation. Previous work suggests the presence of millennial-scale variability over broad regions during TII including climate instability during late MIS 6 and a pause or climate oscillation during TII. Low-resolution Mg/Ca- and alkenone-derived temperature records from these sites indicate the relative timing within the sequence of local salinity and global ice volume changes. Few options exist for dating marine records during Termination II. They are typically tuned to orbital insolation, which precludes testing deglacial mechanisms. However, the distinct millennial-scale features of TII documented by the California margin records represent several widespread climate events that potentially could be correlated with well-dated high-resolution records elsewhere. For example, the $^{230}$Th-dated stalagmite oxygen isotope record from Dongge Cave, China appears to exhibit the same millennial-scale climate events as the California margin records during late MIS 6 and Termination II. This allows the adoption of the stalagmite radiometric chronology to the California margin records. This chronology supports the Milankovitch theory of deglaciation. The suborbital history of climate variability during Termination II may account for records of early deglaciation.
PP32A-06 11:35h
Testing the Silicic Acid Leakage Hypothesis: A Paleoproductivity Reconstruction in the Eastern Equatorial Pacific From the Last Glacial Maximum to the Present
Records of Biogenic opal, Uranium, Thorium and Protactinium were measured in four cores in the Eastern Equatorial Pacific in order to reconstruct paleoproductivity through the Last Glacial Maximum (LGM) and to test the Silicic Acid Leakage Hypothesis (SALH). The SALH suggests that during the LGM unused silicic acid escaped the Southern Ocean through intermediate and mode waters, and was transported to the equatorial ocean. Previous work indicates that in the presence of excess silicic acid, diatoms are better able to utilize nitrate and other nutrients than cocolithophorids, therefore giving diatoms an ecological advantage. The implications of this include an ecological shift from cocolithophorids to diatoms, and more importantly a drawdown of CO2 from the atmosphere due to calcium carbonate compensation. Furthermore, the SALH predicts a rise in opal burial during the LGM in the equatorial oceans as a result of increased diatom productivity. Preserved biogenic opal was measured in four cores near the Peru margin to test this hypothesis. Additionally, 230Th was used to correct for the effects of sediment focusing, authigenic Uranium was used to indicate periods of reducing conditions characteristic of high-productivity areas, and 231Pa/230Th ratios were used as an indicator of opal and total particle flux. Alkenone data from previous studies were also included. These data do not support the SALH, as there is no evidence for greater opal burial during the LGM in any of the four cores. Furthermore, glacial alkenone fluxes remained high and 231Pa/230Th ratios and Uauth concentrations do not show significant shifts. This suggests that the SALH may not be a sufficient explanation for any drop in atmospheric CO2 levels during the LGM.
PP32A-07 11:50h
Paleoceanographic Reconstruction of Productivity and Oxygenation Changes Across the Mediterranean Sea During the Deposition of two Late Pliocene Sapropels
The Mediterranean Sea is a marginal basin located at low latitudes that is extremely sensitive to climatic and environmental variations. In particular, climate oscillations associated with precessional cycles are amplified within this basin and drive dramatic oceanographic and environmental changes that result in the deposition of discrete layers of dark sediment at a 21-kyr periodicity. Sapropels are layers of organic-rich sediment deposited over the whole basin in correspondence with precessional minima, which are periods of increased humidity and seasonality over the Mediterranean area. Sapropel layers are interspersed in the light-colored, organic-poor calcareous marls that constitute the majority of the sedimentary record of the basin, which is 'normally' oligotrophic and fully oxygenated. We investigated the geochemistry of complete and continuous sequences containing two late Pliocene sapropels and intervening 'normal' sediment from six sites selected to cover an ample spectrum of oceanographic conditions from both eastern and western parts of the basin. The orbital signal that drives the formation of sapropels is modulated by local factors creating gradients and differences in the expression of sapropels across the basin. The geochemical composition of sapropels indicates they formed under oceanographic and environmental conditions extremely different from today. Results from bulk organic carbon and nitrogen isotopes in fact reveal that sapropels reflect an increase in the surface productivity coupled with extremely high rates of nitrogen fixation. Moreover, enrichments of chalcophile and redox-sensitive elements suggest the presence of anoxic conditions in the deep waters. Sapropels provide an ideal natural experiment to study the mechanisms and dynamics associated with climate changes because they allow the investigation of the same processes acting repetitively through time on a basin-wide scale. The use of proxies diagnostic of surface processes (carbon and nitrogen stable isotopes), such as productivity changes, and of bottom processes (trace and redox-sensitive elements), such as changes in the oxygenation state of deep waters, allows the evaluation of either a surface or bottom control on sapropel deposition. In addition, investigation of the concordant sapropel layers across the basin permits assessment of the spatial variability of sapropel expression caused by the interaction of local factors with the global signal and of the relative importance of water depth, proximity to land, and location in different sub-basins in modulating the orbital forcing. Finally, the presence of interruptions within sapropel layers suggests the existence of a delicate balance driving the formation of sapropels and allows the investigation of sub-orbital variations in the climate of the Mediterranean Sea.
PP32A-08 12:05h
Warmer Tropics During the mid Pliocene: Evidence From Alkenone Paleothermometry and a Fully Coupled Ocean-Atmosphere GCM
Traditional reconstructions of sea-surface temperatures (SSTs) produced by the PRISM Group (Pliocene Research Interpretations and Synoptic Mapping) indicate that mid Pliocene surface ocean temperatures were unchanged relative to modern at the tropics and low-latitudes and significantly warmer at higher latitudes, particularly in the North Atlantic. This change in the latitudinal pattern of SSTs has been attributed to enhanced meridional ocean heat transport generated by more vigorous surface ocean gyres and/or thermohaline circulation. This study assesses established SST reconstructions for the mid Pliocene through the application of combined data/modeling techniques. New SST estimates were derived using alkenone paleothermometry at three tropical/subtropical Pacific sites. Published alkenone SST estimates for the Atlantic Ocean were also utilized. The new SST data were combined with predicted SSTs derived from a fully-coupled mid Pliocene ocean-atmosphere general circulation model. Significant differences are noted between absolute PRISM, alkenone and model-predicted SSTs. These differences are generated by errors in the PRISM and alkenone paleothermometry estimates as well as limitations of the climate model itself. However, alkenone and model-based SST estimates are consistent in that both predict a pattern of SST warming during the mid Pliocene in tropical and low-latitude regions, which contrasts with PRISM's estimates of unchanged SST in these regions. The latitudinal pattern of SSTs, produced by alkenone estimates and modeling, is not characteristic of that produced by enhanced meridional ocean heat transport or thermohaline circulation. Instead the pattern is similar to that which might be expected as a result of higher concentrations of atmospheric CO2, which would act to warm the oceans globally.