PP51B-0476
A Monte Carlo approach to understanding the relative importance of Greenlandic vs Antarctic forcings of global sea level during the last glacial cycle
The relative importance of Northern Hemisphere versus Southern Hemisphere forcing(s) of global sea level during the last glacial period is hotly debated. There is enticing evidence for both cases but the question remains open. As an exercise, we investigate something approaching the simplest possible situation of global ice volume that can be ‘forced' by Greenland or Antarctic temperature records or a mixture of both. An advantage of using such a simple approach or model is that we are able to run tens of thousands of possible variations in order to optimise the potential of any given temperature record to simulate the global sea-level curve for the last glacial cycle. The analysis provides insight into not only possible drivers of former sea level but also the records as paleoclimate proxies. The results indicate that neither the Antarctic nor the Greenland record alone can simulate the complete stratigraphy of the global ice-volume record. Instead, during different phases of the last glacial cycle, the Antarctic and Greenland signals play either leading or secondary roles in terms of simulating the stratigraphy. We suggest that the analysis implies that the transitions between ‘Greenland and Antarctic dominance' are fundamental drivers of transitions between glacial and interglacial climate states.
PP51B-0477
Elucidating the Holocene Relative Sea-Level History of the US Atlantic Coast
There is an urgent need for a re-assessment of the quality of Holocene relative sea-level (RSL) observations from the Atlantic and Gulf coasts of the United States. Holocene RSL data provide an essential benchmark to compare against other records of RSL change in the last 100-150 years. Only high quality sea-level data reveal spatial and temporal variations in crustal movements since the Last Glacial Maximum and provide a vital constraint upon dynamical models of the Glacial Isostatic Adjustment (GIA) process. We require an accurate model of the GIA process to inform the global data set currently being produced on the time dependence of the gravitational field of the planet by the Gravity Recovery and Climate Experiment (GRACE). We have compiled database of 1400 possible sea-level index points for the Atlantic coast of the United States. These unvalidated data contain much scatter due to concepts inherent in their original interpretation. They rarely allow for other factors such as sediment compaction and tidal range variation. Following validation, results from Delaware and North Carolina show a rapid rise in RSL from the early to mid Holocene, with no evidence of sea- level above present. They also differ somewhat from GIA model predictions. Rates of RSL change during the late Holocene divide into three broad zones: (1) Maine to Boston, MA ~ RSL rise of less than 1 mm yr-1; (2) Cape Cod, MA to Maryland ~ RSL rise between 1 and 3 mm yr-1; and (3) North Carolina to South Carolina ~ RSL rise less than 1 mm yr-1. Comparison with tide-gauge records demonstrate an increase of at least 1 mm yr-1 at all sites since AD 1900.
PP51B-0478
Revised stratigraphy, timing, and amplitude of the middle Miocene eustatic sea-level fall recorded on the Marion Plateau, Northeastern Australia (ODP Leg 194)
Carbonate sequences of the Marion Plateau of northeast Australia (ODP Leg 194) have been critical in reconstructing the amplitude of the middle Miocene sea-level fall. Here, we present a revised stratigraphic framework for the Marion Plateau based on correlating drilling sites along the slope using core and downhole log data, and tying the platform to the slope sequences using recently published Sr isotope data. The new stratigraphic framework is significantly different from the shipboard results, which prompts us to reevaluate the estimate of the magnitude of the eustatic sea-level fall based on backstripping of the Marion Plateau sequences. Because correlation between the upper slope and distal slope sites is significantly improved, we can now use the relatively well-constrained age model of the distal slope to date critical sequences on the upper slope. Preliminary results show that the lowstand interval is characterized by condensed depositional facies, probably triggered by a reduction in the carbonate factory during eustatic lowstands. Furthermore, we show that the middle Miocene "lowstand" recorded by the Marion Plateau is a composite of several phases of successive sea-level falls, spanning the 13.5-11.7 Ma interval (oxygen isotope events Mi3, Mi4, and Mi5).
PP51B-0479
Oxygen isotopes of Pacific seawater, 0-40 kyr, based on d18O and Mg/Ca of benthic and planktic foraminifera: relation to deglacial sealevel rise.
Combining high resolution (~200 year sample resolution) oxygen and carbon isotope measurements and Mg/Ca analyses in planktic (G. ruber and N. dutertrei) and benthic foraminifera (Uvigerina sp.; use of infaunal benthics minimizes likely CO3= effects) from the mid-depth eastern Pacific provides for detailed estimates of changing d18O of seawater over the past 40 kyr at ODP Site 1242. The key to this analysis is improved precision of the Mg/Ca analyses based on a new generation of high precision flow-through time-resolved analysis (FT-TRA) (e.g., average internal precision for Mg/Ca is now +/-0.005 benthic, and +/-0.015 planktic). This method is relatively insensitive to mild dissolution of heterogeneous shells, and thus requires no corrections for preservation. The change in seawater d18O (at 1364 m depth) between the Holocene and Last Glacial Maximum is 1.2 +/- 0.04 permil when averaged over the stages, consistent with pore-water diffusion estimates; however, millennial scale events (which the pore-water data cannot detect) bring the total range up to about 1.6 permil. Are such short-term events related to sealevel change, or do they reflect changing watermasses? Measurements of d13C data (C. wuellerstorfi), sensitive to modern subsurface watermass gradients, are not highly correlated to short- term changes in d18Oseawater suggesting a transient response to ice volume changes. Glacial weakening of AAIW (salty, high d18O, high d13C) relative to north Pacific watermasses (fresher, lower d18O, low d13C) suggest that the benthic d18Oseawater may underestimate total local changes related to ice volume (with a caveat regarding proper scaling of benthic Mg/Ca to temperature). Benthic d18Oseawater falls through the deglaciation in steps, starting at 18 cal ka, with maximum rates of change at 14-15 ka, and with secondary rapid steps at 16-17 ka and 10-11 kar. Planktic foraminifera yield smaller glacial-interglacial d18Oseawater values, with Holocene-to-LGM stage- average changes of 1.04 +/- .06 for surface-dwelling G. ruber (with significant temperature changes recorded by Mg/Ca) and 1.12 +/-0.03 for thermocline-dwelling N. dutertrei (with essentially no temperature changes recorded by Mg/Ca). Changes lower than the benthic seawater estimates suggest slight glacial freshening of the Eastern Pacific Warm Pool. Millennial scale changes are also superimposed on the planktic data, with variations on the scale of 4-5 kyr in both Holocene and LGM time.
PP51B-0480
Mechanisms for Sea Level Change During Marine Isotope Stage 3
A number of climate proxies indicate that a ~7-kyr oscillation occurred during Marine Isotope Stage (MIS) 3, of which change in the Atlantic meridional overturning circulation (AMOC) and attendant change in cross-equatorial ocean heat transport played an integral role. The timing of MIS-3 sea-level changes is clearly linked to this climate oscillation. We applied the GENESIS (V2.2, GEN2) AGCM in a series of sensitivity tests to evaluate the response of the mass balances of the Northern Hemisphere ice sheets to changes in the tropical Pacific and North Atlantic SSTs that are part of this climate oscillation. To evaluate the sensitivity of ice-sheet mass balance to changes in SSTs during MIS 3, we use the Byrd temperature record as a proxy for establishing the timing of tropical SST changes, and the GISP2 ƒ¿ƒ¡ƒ¨ record as a proxy for establishing the timing of well-established changes in North Atlantic SSTs. On this basis, we identify four primary combinations of tropical and North Atlantic SSTs during each of the MIS 3 oscillations that we used to prescribe global SST fields in our simulations: (i) cold tropics, cold North Atlantic, (ii) warm tropics, cold North Atlantic, (iii) cold tropics, warm North Atlantic, and (iv) warm tropics, warm North Atlantic. Our modeled sea-level history is characterized by four fluctuations that are remarkably similar to those inferred from several other proxies of sea level change, including the New Guinea coral-reef record, benthic ƒ¡ƒ¨ƒ¿ records, and the Siddall-03 Red Sea ƒ¡ƒ¨ƒ¿ record. Our record also shares a similar structure, within dating uncertainties, to the Arz-07 Red Sea sea-level reconstruction. Our modeled sea-level changes (order of 10 m) are similar to those derived from the New Guinea coral record (10-15 m), are comparable, within error, to those inferred from the Red Sea records, but are substantially less than needed to explain the benthic ƒ¡ƒ¨ƒ¿ records, suggesting either a greater ice-sheet contribution than we have modeled or the need for parallel changes in North Atlantic deep water and Antarctic Intermediate Water temperature.
PP51B-0481
Calibration of the carbonate `clumped isotope' paleotemperature proxy using mollusc shells and benthic foraminiferal tests
It has recently been shown that the carbonate `clumped isotope' thermometer can provide temperature constraints that depend only on the isotopic composition of carbonate (in particular, on the proportion of 13C and 18O that form bonds with each other), and that do not require assumptions about the isotopic composition of the water in which the carbonate formed (Ghosh et al., 2006). Furthermore, this novel method permits the calculation of seawater δ18O based on the clumped isotope temperature estimates and the simultaneously obtained δ18O of carbonate, thereby enabling the extraction of global ice volume estimates for both the recent and distant geologic past. Here we present clumped isotope analyses of several naturally occurring marine carbonates that calcified at known temperatures in the modern ocean. First, we analyzed benthic foraminiferal tests from six high-quality multicore tops collected in the Florida Strait, spanning a temperature range of 9.3-20.2 degrees C. Second, we analyzed shallow-water mollusc shells from a variety of different climate regimes, spanning a temperature range of 2.5-26.0 degrees C. We find that the calcitic foraminiferal species Cibicidoides spp. agrees well with the inorganic calcite precipitation experiments of Ghosh et al. (2006), while the aragonitic species Hoeglundina elegans is significantly offset. Similarly, clumped isotope results obtained from aragonitic mollusc shells also reveal an offset from the Ghosh et al. (2006) trend, although the offset observed in mollusc aragonite is quite different in nature from that observed in foraminiferal aragonite. Assuming our estimates of the growth temperatures of these naturally occurring organisms are correct, these results suggest that there are vital effects associated with the stable isotope compositions of the aragonite-precipitating organisms examined in this study; further work will be required to determine their cause. Nevertheless, the internal coherence of trends for these materials indicates that clumped isotope analyses of aragonitic molluscs and the tests of the benthic foraminiferal species Cibicidoides spp. and Hoeglundina elegans can be used to reconstruct bottom water temperatures, seawater δ18O values, and global ice volumes.
PP51B-0482
Age differences between Atlantic and Pacific benthic d18O change at terminations
Because a large fraction of benthic δ18O change is due to global ice volume change, benthic δ18O is often used as stratigraphic tool to place marine records on a common age model and as a proxy for the timing of ice volume/sea level change. These applications require the assumptions that δ18O change is rapidly transmitted throughout the deep ocean and that the effects of hydrographic changes are in phase with ice volume. Recently, Skinner and Shackleton [2005] found that the timing of benthic δ18O change at the last termination differed by 4500 years between two sites in the Atlantic and Pacific. Based on Mg/Ca paleothermometry, they argued that these age discrepancies resulted from a late temperature increase in the Pacific and millennial-scale circulation changes in the Atlantic. Do these results imply that benthic δ18O change may not accurately record the timing of terminations? We compare benthic δ18O records from 34 sites in the Atlantic and Pacific to evaluate the impact of ocean mixing rates and deep water changes on the relative timing of terminations recorded in benthic δ18O. Statistical analysis of sedimentation rates derived from the alignment of benthic δ18O suggests an Atlantic lead over Pacific benthic δ18O change for all terminations of the last 600 kyr. The magnitude of sedimentation rate change suggests an average termination age difference of 1000-1500 years between the Atlantic and Pacific, consistent with or slightly greater than the delay expected due to ocean mixing rates, given that most glacial meltwater probably enters the North Atlantic.