Modelling studies on the paleoceanographic potential of neodymium as a novel circulation proxy

Rempfer, Johannes (2012). Modelling studies on the paleoceanographic potential of neodymium as a novel circulation proxy (Unpublished). (Dissertation, Universität Bern, Philosophisch–naturwissenschaftliche Fakultät, Physikalisches Institut, Abteilung für Klima– und Umweltphysik)

Text rempfer12phd.pdf - Other Restricted to registered users only Available under License BORIS Standard License. Download (12MB)

The ocean’s Meridional Overturning Circulation (MOC) is an important component of the climate system and its role in past and future climate change is subject of current research. A way to gain deeper insight into variations in MOC in general, and its driving mechanisms in particular, is the investigation of past changes by using paleocirculation proxies. However, to date, an unequivocal reconstruction is hampered by limited understanding, uncertainties and sparse reconstructions.
A novel and promising proxy of water mass sources and mixing, which is increasingly being used in paleoceanographic studies, is the isotopic composition of neodymium, εNd. Although it is accepted that all the Nd is introduced into seawater by weathering of the continental rocks surrounding the ocean basins, the nature and magnitude of its sources and sinks is still a matter of debate.
In this thesis a new approach for the simulation of Nd isotopes is presented and the potential of εNd for the reconstruction of past changes in MOC is explored in a systematic and comprehensive way. Results of εNd from simulations over glacial-interglacial time scales are presented and discussed with regard to available reconstructions, thus for the first time providing direct constraints on simulated glacial-interglacial changes in MOC.
The introduction in chapter 1 is intended to provide a brief overview on driving mechanisms of present-day MOC, as well as on projections of future and reconstructions of past changes in MOC. The role of MOC for marine biogeochemical cycles and concentrations of atmospheric greenhouse gases is illustrated using the example of carbon. Some details on important characteristics of four paleocirculation proxies are outlined. Additionally, a brief overview is given on archives from which reconstructions are possible. Finally, millenial-scale and glacial-interglacial changes in MOC during the last glacial cycle, as reconstructed from paleo archives and as simulated with models of different complexity are briefly reviewed.
By introducing a novel and comprehensive approach for the simulation of Nd isotopes the study presented in chapter 2, and published in Geochimica et Cosmochimica Acta in 2011, forms the basis for subsequent chapters 3-5. Robust knowledge of the marine Nd cycle is of great interest in the oceanographic and paleoceanographic community. This is illustrated by the fact that Nd is one of the key parameters in a large international scientific program (GEOTRACES), “which aims to improve the understanding of biogeochemical cycles and large-scale distribution of trace elements and their isotopes in the marine environment”.
Due to its computational efficiency the Bern3D Earth System Model of Intermediate Complexity, which is used throughout this thesis, is effectively unrestricted by computational costs and therefore permits a careful parametrisation and a thorough investigation of sensitivities. It is shown that the model is able to simulate both Nd dissolved concentration ([N d]d) and εNd in reasonable agreement with observations. In fact, despite its low resolution and due to new parametrisations and a comprehensive adjustment of model parameters, agreement between simulated and observed [N d]d and εNd is even better than in a previous study which used a more sophisticated and better resolved ocean general circulation model. As a consequence, the study presented in chapter 2 puts tighter constraints on some characteristics of the marine Nd cycle. For example, the mean residence time of Nd is about 700 yr, hence
similar but slightly smaller than the mean ocean mixing time of the model (about 830 yr). Finally, the potential of N d as a quasi-conservative proxy of water mass mixing is indicated
by the fact that salinity and εNd show a close covariation in the Atlantic.
Chapter 3 presents a study in which the effect of millenial-scale changes in MOC on εNd is examined in a systematic way (submitted to Paleoceanography on March 2, 2012). Large variations in εNd (∆εNd, up to 5 εNd -units) result from periodic weakening and strengthening of the formation of North Atlantic Deep Water (NADW) and Antarctic Bottom Water (AABW). Main findings are, first, that no unequivocal relationship is found between εNd and the formation of NADW, thus seriously complicating quantitative reconstructions. In contrast, the relationship is more pronounced between εNd and the formation of AABW. Second, changes in εNd of end-members are relatively small, this being an important prerequisite for the interpretation of variations in εNd further downstream as changes in MOC. Third, results indicate that inferences on the origin of changes (North Atlantic versus Southern Ocean) are possible based on two different characteristics of the pattern of variations in εNd at the seafloor: (i) Atlantic patterns of ∆ εNd differ between experiments where either NADW or AABW are affected; (ii) the sign of changes in εNd in the global ocean is uniform in case of variations in NADW, but is opposite in the Atlantic and in the Pacific when regarding variations in AABW. Fourth, it is shown that although absolute values of εNd are somewhat affected by variations in the magnitude and the composition of particle export fluxes, the pattern of changes is not. This reduces uncertainties associated with the interpretation of variations in εNd as circulation signals.
One of the major findings of a previous modelling study2 is being revisited in chapter 4 (submitted to Journal of Geophysical Research on April 26, 2012). The authors argue that past changes in the sources of Nd could hamper the use of εNd as a proxy for past changes in the distribution of deep water masses. This in fact is a serious concern and needs to be examined in more detail with a more complex model.
For illustrative purposes rather extreme changes in either the magnitude of source fluxes, or their isotopic composition, or both are applied. On the one hand, results show that changes in Nd sources have the potential to affect εNd in seawater. On the other hand, results show that substantial changes are required to generate large-scale changes in εNd in deep water that are similar in magnitude to those that have been reconstructed from sediment cores on glacial-interglacial time-scales, or that result from changes in MOC in model experiments as presented in chapter 3. It is therefore concluded that a shift in Nd comparable to glacial-interglacial variations is difficult to obtain by changes in Nd sources alone, but that smaller variations could be caused by such changes and thus require careful interpretation or additional constraints on changes in Nd sources.
Results from first glacial-interglacial simulations of εNd are presented and compared to available reconstructions in chapter 5. Simulated εNd at sites in the Northwest and South Atlantic
are in qualitative agreement with reconstructions and support previous interpretations. With regard to glacial-interglacial changes in AMOC results agree with the scenario of increased overturning in the North Atlantic during MIS 5.4-5.1, a marked decrease at the transition from MIS 5.1 to glacial MIS 4, and a weaker and more shallow overturning cell in the North Atlantic during parts of MIS 4. Consequently, a larger fraction of the Atlantic basin is filled by southern source water during MIS 4. At the transition to MIS 3 penetration depth and strength of overturning in the North Atlantic increase again. Another marked weakening and shoaling occurs at the transition from MIS 3 to glacial MIS 2, accompanied by an increase in overturning in the deep Southern Ocean, again flushing the deep Atlantic with
southern source water. During the Last Glacial Maximum overturning is weak (about 50% of its modern strength) and confined to depths shallower than 2 km in the North Atlantic and is stronger in the deep South Atlantic (about a factor of 4, compared to its modern strength). During the deglaciation AMOC increases within a few thousand years and reaches its modern strength of 14 Sv at about 6.5 kyr BP. Covariation of εNd and salinity in the Atlantic, indicating the water mass property of εNd, becomes less pronounced in the course of the glacial due to a general increase in salinity and a decrease in its inter-basin gradient. The relationship is least pronounced during glacial stages 4 and 2.
Results presented in chapter 5 provide additional confidence in the performance of the Nd module as well as in the performance of the Bern3D model and permit constraints on glacial-interglacial changes in AMOC. However, they also point to uncertainties which are, for example, associated with the rigid-lid character of the model and the changes in εNd in the North Atlantic end-member. In conclusion, these uncertainties together with the small number of available reconstructions prevent a rigorous quantitative interpretation of the simulated patterns, particularly in the South Atlantic and the Indian Oceans.
Finally, an outlook on how this work can and will be continued in the near and more distant future is given in chapter 6.

Interest & Impact

Downloads

2 since deposited on 04 Oct 2013
2 in the past 12 months

Citations

5 Citations in Web of Science ®
6 Citations in Scopus

Search

in Google Scholar™

Services

Actions (login required)

Edit item

Actions (login required)

Edit item