Skip to main content

Modeling the chronology of deep ocean circulation changes during abrupt climate transitions

Periodic Reporting for period 1 - OCTANT (Modeling the chronology of deep ocean circulation changes during abrupt climate transitions)

Reporting period: 2016-03-19 to 2018-03-18

Ocean circulation plays an essential role in Earth’s climate and the global carbon cycle. Indeed, due to its large volume the ocean is able to store or release large quantities of heat and carbon. The rate at which these quantities are exchanged with the atmosphere is set by the rate at which interior waters are replaced by surface water, or the ventilation rate. There is a critical need to strengthen our understanding of the mechanisms controlling this rate in order to more adequately represent them in climate models, and as a result, gain confidence in future climate projections.
Addressing past ocean circulation changes could provide the means for assessing the processes at stake and the ability of ocean general circulation models (OGCMs) to reproduce them. However, our understanding of such changes in terms of transport pathways and transit times is impeded by large uncertainties in data-based reconstructions which heavily rely on radiocarbon data from deep sea cores.

In addition to measurement errors and mixing processes in the sediment, there are two important reasons why the interpretation of field data is ambiguous. The first is intrinsic to the radiocarbon cycle. Radiocarbon, whose source is in the atmosphere, is characterized by a low air-sea exchange rate. Therefore, the interplay between slow sea surface adjustment and transit pathways in the ocean interior leads to significant differences between radiocarbon-based ventilation rates and true ventilation rates. Further, radiocarbon atmospheric levels changed dramatically over the last fifty thousand years. That this evolution is not well constrained has consequences on both the accuracy of sample dating and the assessment of ventilation changes. Second, the classical methods of interpreting the measured signal display significant shortcomings. Each of them calls for implicit assumptions in terms of water mass pathways, origin and composition of source water, which may be at odds with the actual properties, especially during dramatic climate transitions.

In the OCTANT project we examined assumptions underlying methods commonly used to assess past ocean ventilation. We also investigated how deep-sea radiocarbon ages scale to the actual ventilation timescales during the transition from the last glacial maximum (26 kyr ago) to the present-day. We further developed different tools based on age theory to help interpret the results.
This lead to develop an understanding of the mechanisms by which radiocarbon ages differ from the true ventilation ages. The project also succeed in providing insight on why studies of past ocean ventilation based on deep-sea core radiocarbon measurements reach contradictory conclusions.
The new results will help develop adequate methodology and strategies for the interpretation of deep-sea cores.
In OCTANT we take advantage of the most recent version of the Max Planck Institute Earth System Model (MPI-ESM) which includes newly developed adaptive bathymetry and river routing components. We implemented in the model several sets of tracers. These include simple representations for quantities commonly measured in deep-sea sediment cores as well as dye tracers and partial ages. Dyes document the role of specific surface areas in the deep ocean mixing ratio and ventilation as reported by radiocarbon and ideal age, respectively, while partial ages provide information on ventilation pathways in the deep ocean.

A large number of numerical experiments were carried on for the purpose of optimizing the formulation of vertical mixing in the ocean so as to obtain the best representation of the modern distributions of salinity, temperature, and radiocarbon, simultaneously. These results also allowed investigating the impact of mixing formulations on deep ocean ventilation pathways. In parallel a thorough study of the reasons for discrepancies between the true ventilation timescales and those reported by radiocarbon was achieved; this work was presented at several workshops and international conferences. Papers on these two topics are under way for publications in peer reviewed journals.

In a last step several experiments of the transition from the last glacial maximum (26 ka BP) to the present-day were performed. The initial states for these experiments were prepared by running the model over several thousand years under conditions prevailing at 26 ka BP. The transient experiments were then constrained with prescribed time varying ice sheets and topography, variations of the Earth orbital elements, and reconstructed atmospheric greenhouse gas concentrations and ¹⁴C levels. We then obtain the evolution over the last 26 kyr of ventilation age, radiocarbon, and two groups of dyes (ventilation and radiocarbon) for different ice sheets, bathymetry, and mixing scenarios. Derived quantities such as benthic-planktonic and projection ages which allow a direct comparison with field studies may readily be obtained. The huge amount of results generated by these experiments is currently being assessed and further analyzed. They were presented at several international conferences and will be summarized in a publication in a peer reviewed journal.
Several processes determine the radiocarbon value at any location in the deep ocean (air-sea exchange rate, renewal time of surface water, mixing, atmospheric level evolution) rendering difficult a straightforward interpretation of radiocarbon data from deep-sea cores.
Within OCTANT we put forward that the slow equilibration with the atmosphere characterizing radiocarbon has the consequence that radiocarbon levels at depth have significant contributions from any surface regions. This is in contrast with the classical description of the deep ocean being only contributed from the North Atlantic and the Southern Ocean.
Further, numerical experiments aimed at reproducing the climate transition from the last glacial maximum illustrate that none of the classical methods used for the interpretation of deep-sea record is able to accurately predict the relative changes in ocean ventilation (see illustration). The discrepancy either stems from changes in the atmospheric radiocarbon or water mass composition.
Several aspects of the transient experiments have still to be assessed but a first outcome of our study is evidence that the ventilation rate of the deep Pacific decreased during the deglaciation as suggested by data-based studies.

The chronologies and proxy fields created in the framework of this project will be beneficial to those investigating past climates by means of ocean archives or modeling studies. Of special interest for the community is the reconstructed spatial and temporal distribution of the radiocarbon age offset between the surface ocean and the atmosphere.