PP43D-01 INVITED
The polar ocean and glacial cycles in atmospheric carbon dioxide: Approaching the sawtooth
Ice age reduction in Antarctic overturning, often referred to here as Antarctic "stratification", is one of the few plausible mechanisms for glacial cycles in atmospheric carbon dioxide. However, this change alone cannot clearly explain the full amplitude of carbon dioxide variation, nor does it provide an inherent mechanism for the time structure of the observed cycles in carbon dioxide and climate. We review here the evidence and ideas behind the Antarctic stratification hypothesis and, with recent data and model results, provide a suggestion for how Antarctic stratification interacted with the global ocean overturning to yield the "sawtooth" structure of the climate and carbon dioxide records: their multi-step progression to glacial maxima, followed by their rapid termination into interglacials. One common thread is the dynamical connection between Antarctic and North Atlantic overturning, through the deep ocean's constant hunger for dense water.
PP43D-02 INVITED
The role of ocean biology in setting atmospheric pCO2: the preformed nutrient theory
The sensitivity of atmospheric pCO2 to changes in (1) ocean circulation and (2) high latitude ocean biology has great implications for understanding the lower atmospheric CO2 content of the glacial atmosphere. Here we present a simple theory for how atmospheric pCO2 varies with preformed PO4 inventory and the ocean carbon storage due to the soft tissue pump. Under conditions of perfect equilibrium between atmosphere and ocean and assuming that total ocean nutrient is conserved, atmospheric pCO2 can be written as a sum of exponential functions of the preformed nutrient inventory. We validate our theory against simulations in a suite of realistic ocean general circulation models (GCMs). We then proceed to discuss two important points as follows: (1) Changes in physical forcing can drive large changes in atmospheric pCO2, even with minimal changes in surface nutrient concentration. If Southern Ocean deepwater formation strengthens, the preformed nutrient inventory and thus atmospheric pCO2 increase. (2) As shown by previous studies, an increase in high latitude nutrient utilization can decrease atmospheric pCO2 significantly. Here we show that the response of atmospheric pCO2 to changes in surface nutrients depends on the oceanic circulation in the models. Increasing deep ocean ventilation by increasing vertical mixing or Southern Ocean winds increases the atmospheric sensitivity to surface nutrient forcing. Conversely, stratifying the Southern Ocean decreases the atmospheric CO2 sensitivity to surface nutrient forcing. In conclusion, atmospheric pCO2 is more sensitive to changes in surface biology changes in climates characterized by higher deep ocean ventilation.
PP43D-03 INVITED
Biological Pump Driven by Temperature Through Remineralization
In the soft tissue carbon pump (hereafter biological pump), sinking particulate organic carbon (POC) transports carbon from the surface ocean to the deep ocean. This vertical transport is able to effectively sequester carbon in the deep ocean and thus maintain an atmosphere with relatively low CO2 concentration. Previous studies have called for a stronger biological pump to explain the low atmospheric CO2 during past glacial periods as well as to help reduce future accumulation of fossil fuel CO2 in the atmosphere. Traditionally an increase in export production driven by increased availability of nutrients is invoked to strengthen the biological pump. Here I argue instead that the biological pump can be strengthened by cooling, which slows the remineralization rate of POC, allowing it to reach greater depths. In other words, the biological pump can be effectively decoupled from production. Temperature has an opposite effect on CO2 when applied to production rate, but this effect is not as significant because production is limited more by nutrients and light than temperature. Also, by changing the depth scale of POC remineralization, the new temperature effect on remineralization can cause a redistribution of nutrients. In a global ocean model, remineralization rates retarded by a 5°C cooling cause the atmospheric CO2 to be reduced by more than 30 ppm and shifts nutrients away from the Atlantic to the Pacific Ocean. Warming has the opposite effect on atmospheric CO2 and nutrient redistribution and has implications for the future.
PP43D-04
Changes in CaCO3 Burial Trump the Biological Pump
The dramatic increases in atmospheric CO2 at the ends of ice ages are usually attributed to a one-two punch coming from the ocean. First, a weakened biological pump vents organically cycled CO2 from the deep ocean via changes in the ventilation of the deep ocean around Antarctica. The initial CO2 increase is then augmented by an enhancement of CaCO3 burial due to a process called CaCO3 compensation (after Broecker, W. S and T.-H. Peng, Global Biogeochem. Cycles, 1, 15-29, 1987). Here, we argue that the importance of the biological pump has been exaggerated. The main effect comes from circulation-induced changes in the burial of CaCO3. As shown in a recent paper by Andreas Schmittner and co-authors (Schmittner, A., E. Brook and J. Ahn, Impact of the ocean's overturning circulation on atmospheric CO2, in Ocean Circulation: Mechanisms and Impacts, Geophys. Monogr. 173, A. Schmittner, J. Chiang, and S. Hemming, eds., pp. 209-246, AGU, 2007) changes in the ventilation of the deep ocean around Antarctica gave rise to 20-30 ppm increases in atmospheric CO2 every 5,000-7,000 years during isotope stages 3 and 4 (30,000 to 70,000 years ago). None of these venting events gave rise to a compensation response. Meanwhile, Jaccard et al. (Science, 308, 1003-1006, 2005) show that all the big CO2 increases during terminations through stage 11 were accompanied by huge increases in CaCO3 burial. This suggests that the enhanced burial of CaCO3 is obligatory rather than compensatory with respect to the dramatic CO2 increases. Broecker and Peng's compensation idea is based on an assumption that the rain of CaCO3 to the sea floor is the same everywhere. More specifically, it assumes that there is no spatial correlation between the production of CaCO3 at the surface and the burial on the sea floor. We find instead that the production and burial of CaCO3 tend to be co-located in regional "hot spots" and that burial in the hot spots balances the input of Ca++ and HCO3- ions in rivers. The hot spots can also move from place to place in response to changes in circulation. The main hot spots today are the eastern Atlantic and southern Indian; the main hot spot during the last glacial was the equatorial Pacific. Renewed deep-water formation in the Atlantic at the end of the last ice age shifted the locus of CaCO3 burial back to the Atlantic and southern Indian and led to a huge drawdown in global alkalinity, which is ongoing today and accounts for most of the deglacial rise in atmospheric CO2.
PP43D-05 INVITED
Models Are Not Just For Christmas
The Quest for an understanding of the state of the ocean chemistry system and the concentration of atmospheric CO2 in the past, and the controls thereon, has created an unholy alliance of models and proxies. Traditionally, models of the Earth system or of its sub-systems are utilized in quantifying the implications of proxy reconstructions, but only once measurements have been fully translated into environmental change. However, the increasingly complexity of new proxies in conjunction with the divergence with time in the past of ocean composition with respect to modern, means that proxies of the ocean carbonate system may be forced to entrain 'guestimates' that can create a considerable degree of overall uncertainty. Computer models in a variety of disguises can be employed in close association with proxy measurements to address some of these uncertainties. This talk will review this closer integration of models, focussing on proxies for glacial-interglacial changes in carbonate preservation and boron isotope based reconstruction of Paleocene-Eocene ocean acidification.
PP43D-06
Deep sea carbonate ion concentrations reconstructed using foraminifer faunas and the modern analog technique
Reconstructions of the glacial-interglacial cycles in seawater carbonate chemistry are useful in understanding past as well as future changes in atmospheric carbon dioxide. The deep sea carbonate ion concentration between 1500-4000 m can be reconstructed using the modern analog technique with an uncertainty of +/- 10 micromoles/kg. The method exploits the observation that the degree of carbonate ion saturation is second only to the temperature of the plankton growth environment in determining the foraminifer assemblages preserved through time. This approach is crude compared to other proxies, but offers the advantage that uncertainty can be quantified and some competing influences (such as changes in the plankton growth environment) can be excluded. Here we expand our previous results to compare the MAT method with other proxies, extend the reconstruction to the sea surface (using other proxies), and examine the changes through time. When combined with surface reconstructions, the results indicate a steepening of the vertical gradient in the carbonate ion concentration between the surface and 1500 m during the last glacial maximum, with little change between 1500 and 4000 m. Time series reconstructions at 1500 m with century-scale resolution do not show the preservation and dissolution spikes expected from whole-ocean changes in chemistry. Instead, these data point to the combined influence of circulation and the biological pump altering the gradients within the upper 1500 m throughout the glacial- interglacial cycles.
PP43D-07
Deep ocean carbonate ion proxies and application to the North Atlantic Ocean and Caribbean Basin reconstructions
The pH and carbonate ion concentrations ([CO32-]) of the deep ocean are crucial for understanding the role of the oceans in the regulation of atmospheric CO2 on timescales of thousands of years and longer. However, proxy-based estimates of glacial-interglacial [CO32-] changes conflict, highlighting the need for new methods. Measurements of B/Ca in core-top samples show that B/Ca in benthic foraminifera serves as a novel, quantitative proxy for past deep water [CO32-] reconstructions. Boron isotopic systematics indicates that benthic foraminiferal δ11B should reflect deep water pH and hence [CO32-]. We applied the benthic B/Ca method to reconstruct deep water [CO32-] changes of the North Atlantic Ocean (50-60°N) water column since the last glacial. Records from six cores over 1 to 4-km reveal that the carbonate chemistry of the glacial North Atlantic was more stratified than the modern ocean, with higher [CO32-] by ~20-30 μmol kg-1, and pH by ~0.15, at intermediate depths and lower concentrations by ~20 μmol kg-1, and pH by ~0.05, at sites deeper than 3.5 km, producing an 800 m glacial shoaling of calcite saturation horizon. Comparison with benthic foraminiferal δ13C shows that the deep glacial waters with low [CO32-] are consistent with a Southern Ocean source, while those with high concentrations comprised two endemic water sources, one being the Norwegian-Greenland Sea. Application of benthic B/Ca and δ11B to a Caribbean Basin core (VM28-122, 3.62 km) shows that deep water [CO32-] varied inversely with atmospheric CO2 during the last 160 kyr. Compared with interglacial periods, glacial intervals show higher [CO32-] by ~35 μmol kg-1 and pH by ~0.2. Applications to the North Atlantic and Caribbean Basin cores demonstrate that benthic B/Ca and δ11B are useful proxies for deep water [CO32-] reconstructions, providing important clues about roles of deep ocean carbonate chemistry on past atmospheric CO2 changes. When combined with benthic foraminiferal δ13C, they can be used together as effective tools for past hydrographic change studies.
PP43D-08
Atmospheric Carbon Dioxide and Climate Following the Middle Miocene Expansion of the East Antarctic Ice Sheet
The Middle Miocene Climate Transition (MMCT) represents a major step, as represented by global oxygen isotope records, from the greenhouse world of the Cretaceous to the icehouse world of the present day. The transition, which is recorded by a δ18O shift of >1‰, has long been associated with a major build-up of ice on East Antarctica. Major changes in the carbon system have been recorded in both the lead-up to the MMCT (the "Monterey Excursion"), and also at the MMCT itself, culminating with a >1.2‰ transient shift in inorganic δ13C immediately following ice build-up ("CM6"). Traditional interpretations suggest that the carbon isotope excursions represent increased burial of organic carbon, which led to a drawdown of carbon dioxide from the atmosphere. More recently, it has been proposed that a reduction in silicate weathering following the growth of the ice sheet led to increased atmospheric carbon dioxide over CM6. We present high resolution coupled multi-proxy records of the 1.1 million years following the MMCT and use alkenone paleobarometry to reconstruct atmospheric CO2 across CM6 to test these hypotheses. Our records shed new light on the cause of this major positive carbon isotope excursion and may have implications as to what controls climate at this point in the Cenozoic.