Skip to main content

Carbonate chemistry, carbon cycle and climate change (a multidisciplinary view)


Trace elements in marine biogenic carbonates may be used as proxies for past ocean chemistry provided that a number of conditions are fulfilled; 1) There is an established relationship between the trace element proxy and a parameter of interest. 2) This relationship is preserved within the biogenic carbonate after burial. 3) The trace element can be determined at the precision and accuracy necessary to produce reliable data. For example, the relationship between Cd and P in the oceans has enabled Cd/Ca in foraminifera to be used as a proxy for deepwater phosphate and deep ocean circulation. The relationship between Ba and alkalinity permits the use of Ba/Ca ratios as a proxy for alkalinity and Zn/Ca ratios combined with Cd/Ca may be used as proxy for deep-water carbonate ion concentration. During recent years Mg/Ca in foraminiferal calcite has become established as an important palaeotracer of ocean temperature. Result description: The instrumental method for the rapid and precise determination of multiple element-calcium ratios in foraminiferal calcite has been established on the ICP-MS instrument at the University of Cambridge. Ten element/Ca ratios, Li/Ca, B/Ca, Mg/Ca, Al/Ca, Mn/Ca, Zn/Ca, Sr/Ca, Cd/Ca, Ba/Ca and U/Ca, are determined in each sample by this quadrupole ICP-MS procedure. The technique is efficient in generating elemental ratios of palaeoceanographic interest simultaneously from a single purified foraminiferal carbonate sample. The long-term reproducibility of the method has yielded precisions of Li/Ca = 2.4%, B/Ca = 4.2%, Mg/Ca = 1.4%, Al/Ca = 14%, Mn/Ca = 0.9%, Zn/Ca = 2.8% (1.2 - 7.8 mmol/mol) and 5.1% (0.5 - 1.2 mmol/mol), Sr/Ca = 0.9%, Cd/Ca = 2.4% (0.07 - 0.24 mmol/mol) and 4.8% (0.01 - 0.07 mmol/mol), and U/Ca = 2.5% for foraminiferal samples as small as 60 mg. Key innovative features: The method takes advantage of the fast scan capability of quadrupole ICP-MS to determine element/calcium ratios directly from drift-corrected intensity ratios. Matrix effects are overcome by the measurement of element intensity ratios after dilution of samples to a constant Ca concentration and matrix matched element ratio standards ensure accurate instrument calibration. Dissemination and use: The published method is of interest to other researchers both within geochemistry and palaeoceanography and in the wider analytical area. Collaboration between the University of Cambridge and other geochemical laboratories is being used to verify the technique and to maintain and improve consistency of results between laboratories. Expected benefits: The simultaneous determination of multiple trace element proxies on a single sample reduces the overall sample size and the analytical time required, thereby improving the correlation between proxies. The faster analysis time combined with small sample size enables higher resolution multi-proxy records to be more readily obtained which in turn will lead to better understanding of geochemical processes in past oceans.
Background and Developments Ice-core records have documented a close relationship between atmospheric CO2 and climate for the past few hundreds of thousand years, with peak cold periods being characterised by about 80-100ppmv (parts per million by volume) lower concentrations than peak interglacials. Although there is agreement that the ocean must play a major role in driving the observed variations, the exact mechanisms remain unclear. A reduction of the carbonate-carbon to organic-carbon ratio in the biogenic export production (the "rain ratio") during glacial times has been advanced to explain the observed variations. This hypothesis is tested and implications for the dynamics of sedimentary carbonate preservation and dissolution are explored with MBM ("Multi-Box Model"), a ten-box model of the ocean carbon cycle, fully coupled to the sediment model MEDUSA ("Model of Early Diagenesis in the Upper Sediment (A)"). MEDUSA is a new transient one-dimensional advection-diffusion-reaction model describing the coupled early diagenesis processes of carbonates, opal and organic matter in the surface sediment. Sediment is represented as a two-phase porous medium (solids and pore-water). In the vertical, it is subdivided into two different zones. Solids raining down from the surface of the ocean are collected by the reactive mixed-layer at the top. Here, solids are transported by bioturbation and advection, solutes by molecular diffusion. Chemical reactions are restricted to this zone. Solids that get transported deeper than the bottom boundary of the reactive mixed-layer enter the second zone, the historical zone, where sediment accumulates. The historical zone is made up by a pile of layers representing a synthetic sediment core. The two zones may exchange material in a fully bi-directional way, i.e., the model can take chemical erosion into account when dissolution in the top reactive layer exceeds the supply of material from above. Results Description A transient scenario with a peak reduction of the rain ratio by 40% at the Last Glacial Maximum (LGM) was found to produce a net atmospheric pCO2 reduction of about 40ppm with the coupled MBM-MEDUSA. Changing shelf carbonate accumulation rates and continental weathering inputs produced a 55-60ppm reduction. The combination of the two mechanisms generates a pCO2 change of 90-95ppm, which compares well with the observed data. However, the resulting model sedimentary record is in phase opposition with available data. The calcite transition zone (CTZ), i.e., the reconstructed separation between the Calcite Saturation Horizon (CSH) and the Carbonate Compensation Depth (CCD), gets about 1.5km thicker during glacial times than at present-day (interglacial). Sedimentary evidence from the Equatorial Pacific witnesses an extended transition zone during interglacial and a thin one during glacial times. Although the actual figures might be biased by the box-model approach adopted here, we do not expect that more realistic models using similar scenarios will provide a response that is opposite in phase. Realistic changes in the aragonite fraction of the carbonate rain were found to have only a minimal impact on atmospheric pCO2. Finally, chemical erosion of deep-sea sediment was shown to reduce the amplitude of variation of the sedimentary CCD by about 10-20%. It may provide a mechanism to improve the model-data agreement. Related additional conclusions and issues that deserve further study, preferably with higher-resolution models, are given in the section on the "Potential Offered for Further Dissemination and Use". The sediment model MEDUSA has not only been coupled to the box-model MBM but also to a 3D ocean carbon cycle model available in our research group. Several 30000-year steady-state as well as 20000-year (Last Glacial Maximum to present-day) and 120000-year (a complete climatic cycle) transient simulation experiments have been carried out. Key Innovative Features The newly developed sediment model MEDUSA offers a flexible, fast and robust representation of time-dependent sedimentary exchange process, with excellent mass conservation properties; is fully bi-directional, i.e., can consistently take chemical erosion into account; can be coupled to carbon cycle models of any complexity; can be used locally at high vertical resolution. Expected Benefits MEDUSA allows realistic long-term simulation experiments with 3D models in open system set-ups, where riverine carbon and alkalinity inputs to the ocean drive carbonate burial fluxes in the deep sea. It helps to overcome a major shortcoming of the commonly used unrealistic closed-system model set-ups where riverine inputs are only used to balance the model-generated burial flux. MEDUSA also enables us to verify paleoceanographic hypotheses in a quantitative way.
Resonant laser secondary neutral mass spectrometry (r-Laser-SNMS) is an emerging analytical technique for efficiently measuring the isotopic composition of trace elements. R-Laser-SNMS is similar to the well-known ToF-SIMS (time-of-flight secondary ion mass spectrometry) technique. Both techniques use a focused energetic ion beam, which can be focused down to 50nm in diameter, for bombarding a solid sample and a ToF mass spectrometer for analysing the sputtered particles. But instead of using only the small fraction of sputtered secondary ions, as SIMS does, r-laser-SNMS uses tuneable lasers to resonantly ionise the majority of the sputtered neutral atoms, which are less affected by the chemical composition of the surface, thus leading to much greater accuracy than SIMS. The energy spectrum of discrete excited states is unique to each element, so that the selection of particular excited states for resonant multiphoton post-ionisation (RMPI) analysis provides extremely high selectivity, thus making the technique especially valuable for high precision isotope ratio measurements or detecting ultra-trace elements in complex samples. This post-ionisation technique also offers a high sensitivity: a useful yield, defined as the number of observed counts at the detector divided by the number of analyte atoms sputtered from the surface, of 3% to 8% has been demonstrated for many elements including boron and detection limits of sub-ppb was shown for boron in silicon. Key Innovations As the number of detectable neutrals is crucial for determining accurate isotope ratios it was important to investigate the sputtering process of atomic boron neutrals from different chemical environments under different analysis conditions. The experimental conditions have been optimised for a high and continuous flux of boron neutrals from calcite surface of single foraminiferal shells over the entire analysis time. For this purpose r-laser-SNMS analyses of different sample systems with various boron concentrations were performed under different analysis conditions. In addition, the influence of the experimental set-up on the measured value of the boron isotope ratio was investigated and optimised for time-stable isotope ratio analysis of single foraminiferal shells resulting in a significantly improved accuracy of isotope ratio measurements of boron containing samples. Consecutive measurements on a test sample, a boron containing metallic glass alloy, were performed with a standard deviation smaller than two per mill. Although, this accuracy could not yet be reached on single foraminiferal shells, the values for the standard deviation correlate with the size of the counting statistical error. Thus, advancement in accuracy can be achieved by enhancing the number of sputtered boron atoms, by increasing the analysis time, by increasing the repetition rates of the analysis cycle and/or by increasing the primary ion current by using special designed high current ion guns. Expected Benefits The results of this project suggest that a microprobe instrument based upon the RMPI technology could significantly advance capabilities in earth sciences for isotope ratio measurements and nanoanalysis. The advantages of the r-laser-SNMS technique include excellent selectivity, sensitivity and efficiency, comprehensive elemental and isotopic applicability, reduced fractionation and matrix effects, nanoanalysis capabilities, and freedom of isobaric interferences. In addition, the integration of RMPI with TOF-SIMS will also allow performing imaging of nanostructure samples where ultra-trace elements can be detected and quantified by r-laser-SNMS, while quasi-simultaneous chemical images of other elements and molecules by ToF-SIMS will provide useful and complementary information of the samples. Dissemination and use The close collaboration between the University of Munster and the company ION-TOF GmbH guarantees that the RMPI technology becomes commercially available. Such an analytical tool would have widespread applications in the earth sciences, greatly extending the resolution in geochemistry, cosmo-chemistry, and paleoceanography brought about by the ion microprobe. It will also be significantly less expensive than other complex geoscience microprobe instruments and therefore very competitive. R-laser-SNMS will also find important applications in chemical and biomedical applications, geological exploration, materials and environmental sciences, and intelligence operations. The demand for analytical instrumentation with spatial resolution in the nanometre range is rapidly increasing in these research fields because reduction in sample sizes and structures and increases in materials purity and compositional fidelity are straining available characterization techniques. A substantial market for such instrumentation can be anticipated.
Among the most important challenges remaining to be addressed by Quaternary paleoceanographers is the mechanism responsible for lowering pCO2 during the Last Glacial Maximum (LGM) and possible feedback mechanisms with climate change (e.g. Sigman and Boyle, 2000). During 6C we used a multi-proxy approach to reconstruct the oceanic carbonate chemistry and understand the natural relationship with the global carbon cycle. One important component was the development of new tools for reconstructing environmental parameters (see TIP results 1-6). The other was to combine analytical records of the sedimentary archive (combining several proxies) in combination with numerical models in order to 1) distinguish the mechanisms that control the operation of the oceanic carbon cycle, 2) identify water masses as sinks or sources of atmospheric CO2 and hence, 3) better constrain the role and the impact of the carbon cycle on climate oscillations. Knowledge of the nature and amplitude of natural fluctuations in the past are a precondition to assess the stability of modern subsystems and their potential range of variations in the future. With regard to the second objective we have combined multiple proxies (boron isotopes, B/Ca, Mg/Ca, Cd/Ca and alkenones) to study sediments from the northern Arabian Sea. We could demonstrate that the magnitude of CO2 degassing from this area increased significantly at ~18 ka, and may thus have played an important role in initiating the rise in atmospheric CO2 levels at the start of the last deglaciation. It is generally accepted that the oceans were instrumental in regulating glacial-interglacial changes in atmospheric CO2, but there is uncertainty over past changes in the location and magnitude of oceanic sources and sinks of CO2. Our reconstructions indicate that the northern Arabian Sea has been a source of CO2 to the atmosphere since 30 ka. The delta11B and B/Ca proxies further suggest that this source (and the intensity of upwelling) increased in intensity from the last glacial maximum to the Holocene. This hypothesis accords with findings from most other studies of the region that suggest the summer monsoon was less intense in the LGM as the Tibetan plateau was heated less strongly at this time. Around ~18 ka the change from relatively low pCO2 values to higher pCO2 values is coincident with the start of the rise in atmospheric CO2 during the last deglaciation. In this context it is noteworthy that it has been observed that the mean effective moisture levels from the Asian monsoon margin started to increase between 18.5 and 17 ka, suggesting that this may represent onset of the summer monsoon after the LGM. Hence, intensification of upwelling in the Arabian Sea and degassing of CO2-rich surface waters may well have played a role in the increase in atmospheric CO2 that was further enhanced by increased La Nina-type conditions in the equatorial Pacific between ~14-16 ka. Although much progress has been achieved with regard to all 3 objectives stated above, we cannot answer objective 1 and 3 as detailed as we had hoped for at the beginning of the project. To address these objectives requires to fully reconstructing the carbonate chemistry of the global ocean. Hereto, planktonic foraminiferal proxies would constrain the chemistry of the surface water and benthic foraminifera would be used to determine the bottom water chemistry. However, several proxies turned out to be influenced by more environmental parameters than just the target parameter. For instance, we proposed to use "size normalized weight" (SNW) of planktonic foraminifera as an estimator for bottom water carbonate ion concentrations. Unfortunately, we had to conclude that "cryptic speciation" in planktonic foraminifera masks the true bottom water values and that SNW is therefore not a reliable proxie. In addition, we didn't reach the desired precision and accuracy for single shell boron isotope analysis required to reliably approximate bottom water pH. These drawbacks made it impossible to reconstruct the bottom water carbonate chemistry. Since "down core" bottom water carbonate chemistry estimates along depth transects of critical ocean areas are the only means to distinguish between the processes that gave rise to the observed changes (physico-chemical carbon uptake or release versus the organic carbon pump or the alkalinity pump) and since, it is also the only way to reconstruct lysocline dynamics, we are currently unable to fully address two of the main objectives that remain the "Holy Grail" for future climate change predictions (but it should be stressed that we have contributed significantly to the objectives).
During the Quarternary several rapid and short-lived climatic shifts have taken place in the North Atlantic Ocean (the so called Heinrich events) due to the release of massive icebergs from high latitude ice sheets. Few studies have considered the continental organic matter that may be transported in conjunction by ice rafting to the North Atlantic Ocean floor and which mixes with locally produced marine organic matter. Usually total organic carbon (TOC) and d13C of TOC (d13Corg) are used to infer the origin of organic matter but as a rule they give poor estimates of soil organic matter contributions. We analysed glycerol dialkyl glycerol tetraethers (GDGTs) in a sediment core from the North Atlantic spanning the last 30 ky to investigate organic matter deposition due to ice rafting. TOC content was low in sediments representing glacial times (0.2 to 0.4%) and even lower in Holocene sediments (<0.2%). d13Corg values varied from -240 in glacial times to -200 at the start of the Holocene (8 ky cal BP; calibrated, Before Present) with negative excursions to -260 during Heinrich events. The d13Corg values correlated with the percentage of ice rafted debris in the sediments, suggestive of supply of continental organic matter by ice rafting. GDGT analysis revealed varying amounts of soil-derived branched GDGTs and the marine isoprenoid GDGT, crenarchaeol, which is expressed in the Branched Isoprenoid Tetraether (BIT) index. This BIT index was relatively high (0.3) in sediments deposited during the glacial compared to those laid down at the start of the Holocene (0.1), suggesting enhanced delivery of terrestrial organic matter to the North Atlantic by ice rafting, in agreement with the d13Corg results. This was confirmed by analyses of 14C-contents of TOC, which indicated substantially older ages than the inferred sediment age. BIT indices and d13Corg show phase offsets during Heinrich events, suggesting differences either in timing of supply or in changing contributions of source areas. The use of the BIT index thus leads to a substantial improvement in estimating the contribution of soil derived organic matter transported by ice-rafted debris into the oceans. This will improve estimates of terrestrial organic carbon burial into the deep ocean. Key innovation Our results show that the BIT index provides a good tracer to infer the presence and relative abundance of soil organic matter in ice rafted debris. GDGTs are relatively newly discovered biomarkers and thus have been rarely used as proxies and therefore this result is highly innovative. Current status, dissemination and use This result has been written up as a scientific publication and is currently under review at the journal Organic Geochemistry. This technique may be used by paleoclimatologists and paleoceanographers to reconstruct past fluxes of organic matter from the continent to the deep sea. We are presently using this tool to reconstruct past transport of soil organic matter into marine environments. We have plans to apply it to periods further back in time (i.e. older than the Quarternary) where the presence of ice rafted debris has been poorly documented.
The concentration of carbon dioxide (CO2) in the atmosphere plays a major role in defining the Earth's climate, and studies have shown that atmospheric CO2 levels were lower in glacial times than in inter-glacial periods. Since we also know that the oceans play a key role in controlling the level of CO2 in the atmosphere today, a major goal of scientific research is to understand how this role may have varied in the past. This result shows that the equatorial Pacific Ocean was a major source of CO2 to the atmosphere between 14 and 16 thousand years ago - a time when the concentration of CO2 in the atmosphere was rising rapidly and the Earth's climate was warming. The planktonic foraminifer, Globigerinoides sacculifer, faithfully records the boron isotope composition (d11B) of dissolved B(OH)4- in the seawater from which the foraminifer grew its shell, and this is directly related to the pH of the seawater. Once the pH of the seawater is known the pCO2 of the waters can be calculated if alkalinity concentrations are also known. We measured the d11B of samples of G. sacculifer from the western equatorial Pacific covering the period 0.4 - 23.2 thousand years ago (ka). The results show that for most of this time the surface waters of this area of the ocean were in equilibrium with atmospheric CO2. However, during the period 13.8 - 15.6 ka the pCO2 values of the surface waters were ~100ppmv higher than atmospheric levels. In the modern western equatorial Pacific the physical structure of the water column inhibits upwelling of nutrient-rich water into the euphotic zone, leading to the formation of a warm pool that has relatively low levels of biological productivity and surface water pCO2 levels that are in equilibrium with the atmosphere. To the east, upwelling brings nutrient and CO2-rich waters to the surface. The low iron concentration of these waters means that the nutrients are not fully utilized (i.e., this is a high nutrient-low chlorophyll (HNLC) area) which results in this area of the ocean being the largest natural source of CO2 to the atmosphere. The longitudinal divide between the warm pool (from which the studied core was recovered) and HNLC areas moves to the east during El Nino events, such that the extent and intensity of the HNLC is reduced and the thermocline shallows in the western equatorial Pacific. In general, surface water pCO2 values vary from close to atmospheric levels in "normal" and El Nino years, but during La Nina (increased upwelling) periods the warm pool retreats to the west and the surface water pCO2 reaches ~80 ppmv at the sample site. Hence, we hypothesise that the pCO2 record reflects a period of more frequent La Nina conditions between 13.8-15.7 ka. This conclusion is supported by several other studies that have also suggested that interstadials are characterised by La Nina conditions. For example, increased wind driven upwelling along the Oman margin and higher productivity in the Cariaco Basin have both been linked with La Nina-type conditions. Finally, the period of high pCO2 observed here is approximately coincident with the deglaciation carbon isotope minimum event (that reached its greatest intensity at 15.9±0.2 ka), and is also apparent in our data. This observation has been ascribed to increased upwelling of CO2-rich sub-Antarctic Mode Water as a consequence of the reestablishment of circum-polar deep water that itself resulted from melt-back of Antarctic sea ice. Our data suggest that a significant portion of this upwelling occurred in the equatorial Pacific. Current status & use Data from this and other studies are compatible with the hypothesis that there was an increase in the intensity of upwelling in the eastern equatorial Pacific at a time that is coincident with the steepest rise in atmospheric CO2 levels during the last deglaciation. The extent to which such upwelling played a role in the increase in atmospheric CO2 during this time requires more precise dating of the relative timing of the various paleo-records as well as modelling of the physical controls of ocean-atmosphere CO2 exchange (e.g., wind stress). Nevertheless, our study shows that d11B studies of planktonic foraminifers are a powerful tool with which to investigate Pleistocene variations in ocean-atmosphere CO2 exchange. Expected benefits: Together with other records this method will help in -identifying sources and sinks of atmospheric CO2 in the past and thereby -improving our understanding of how sensitive the atmosphere is to changes in CO2 - understanding what role the ocean plays in controlling atmospheric CO2. Dissemination and use potential It is planned to develop a faster and easier method to measure d11B using multi-collector ICPMS. This would also strongly increase the value of this method to study CO2 fluctuations in the past oceans for other palaeoclimatologists.