Millenial-scale atmospheric CO2 reconstructions using multiple Antarctic ice cores

Nehrbass-Ahles, Christoph (2017). Millenial-scale atmospheric CO2 reconstructions using multiple Antarctic ice cores (Unpublished). (Dissertation, Universität Bern, Philosophisch–naturwissenschaftliche Fakultät, Physikalisches Institut, Abteilung für Klima– und Umweltphysik)

Text nehrbass-ahles17phd.pdf - Other Restricted to registered users only Available under License BORIS Standard License. Download (20MB) | Request a copy

This thesis aims at improving ice core-derived CO2 reconstructions of the past atmosphere, far beyond modern observations. This is achieved by improving the analytical capacity of a recently developed CO2 measurement system, increasing the temporal resolution with new high-precision data in time intervals of previously only poor resolution, and reviewing processes that may bias the measurement outcome with relation to the true atmospheric value.
High-precision measurements of CO2 mole fractions ([CO2]) using Antarctic ice samples are technically challenging, especially the contamination-free and complete extraction of the air that is trapped in the ice. Chapter 2 of this thesis provides an in-depth description of a new mechanical dry-extraction unit – called the Centrifugal Ice Microtome (CIM) – that was constructed prior to this thesis and has been integrated into an automated CO2 measurement system. A major focus of this work has been the further development and refinement of this system, focusing on the long-term stability and decreasing the required maintenance frequency. These efforts resulted in a standard measurement protocol, which allows the user to carry out a continuous measurement campaign over several months. The system is now fully operational, and all potential issues connected with the continuous operation have been reduced to a minimum. In the current configuration the system allows for measuring 24 ice samples per day, with an additional 6 – 7 standard-air-over-blank-ice measurements. The latter measurements allow for the monitoring of potential system drifts and serve to provide a robust correction for extraction and analytical effects. Samples sizes can range between 3 – 14 g (typically 7.5 ± 0.5 g) and the extraction efficiencies are ≥95 % and ≥90 % for bubbly and clathrated ice, respectively. The measurement precision depends on the sample weight and the air content of the ice, but typically ranges between 0.5 – 1.2 ppm.
The aforementioned progress in the analytical system allowed for the reconstruction of [CO2] over Marine Isotope Stage (MIS) 9 – 11 (i.e. 328 – 450 ka BP; 1000 yr before present, where present is defined as 1950 CE) using ice from the European Project for Ice Coring in Antarctica (EPICA) Dome C ice core (EDC; chapter 3). In comparison to previously available data, which only allowed for investigations on orbital time scales, this record shows a higher measurement precision by a factor of three (now: 1 ppm, pooled standard deviation) and a dense sampling strategy that increased the previous sample resolution by a factor of four to six (now: ∼300 yr on average). In particular, this record revealed previously unidentified millennial-scale CO2 variability and suggests that pulse-like CO2 releases to the atmosphere have happened on sub-millennial time scales at rates of 10 ppm per century. This CO2 releasing mechanism is not unique and represents a persistent mode throughout glacial and early interglacial conditions over the past 450 ka; it may be linked to overshoots in the Atlantic Meridional Overturning Circulation (AMOC). As the first dataset over this period of such precision and resolution, it will serve the paleoclimate community well for studies of sub-millennial scale carbon cycle dynamics.
It has been known for at least two decades that past atmospheric CO2 can only reliably be determined from Antarctic ice cores, as Greenland ice is prone to in situ production of CO2, which alters the atmospheric [CO2] already within the ice. Evidence is mounting in the form of CO2 offsets that similar processes also exist at some Antarctic ice core sites, though to a smaller extent. Many previously reported offsets between different [CO2] records derived from Antarctic ice cores are apparently constant and may thus originate from differences in the applied extraction system, inconsistent standardization techniques, or post-coring effects. However, some significant offsets exist between high quality [CO2] reconstructions that are unlikely to stem from such effects. Chapter 4 provides an overview of potential CO2 offset-causing processes and presents new measurements from the EDC, Talos Dome, and Vostok ice cores during the Holocene and the Last Interglacial (LIG). In relation to the previously published record from EDC during the Holocene, the new [CO2] data of the Talos Dome ice core reveal an increasing offset over time at a rate of 0.8 ppm ka−1 from 1 – 10 ka BP. In analogy to the Holocene, the LIG does not show higher offsets as could be expected from the observed Holocene production rate. Even after 100 ka of potential production time within the ice sheet the absolute offset levels do not exceed the magnitude of the offsets found in the Holocene. These data provide another line of evidence that offset between different [CO2] reconstructions may already develop in the ice sheet. Although the responsible mechanism cannot be unambiguously identified, these offsets point to potential in situ production of surplus CO2 from inorganic or organic compounds in the Antarctic ice archive that may develop on millennial time scales after ice formation, which bias the atmospheric [CO2] to a small, but significant extent. However, their production potential seems to be limited as offsets between Talos Dome (among other ice cores) and EDC during the LIG do not exceed 8 ppm. Low-accumulation sites in the interior of Antarctica seem not to be (or only marginally) affected by such effects, including the EDC and the Vostok ice cores.
In chapter 5 we find an example of an offset between different measurement techniques applied on the same ice core (EDC). We were able to identify and correct for an offset that influenced the deepest 200 m of the EDC ice core, resulting in a revision of the CO2 reconstruction from 600 – 800 ka BP. In addition, this publication provides an updated compilation of CO2 over the past 800 ka.
The data situation over the last 100 ka is somewhat ambiguous. Many different records exist that partially show major differences with respect to each other. Chapter 6 provides a well-docuamented, updated (with respect to chapter 5), expert compilation of the greenhouse gases CO2, CH4, and N2O over the last 156 ka. The compiled data was interpolated in order to provide continuous records on temporally equidistant grids, which are typically needed for transient climate simulations of these three greenhouse gases.

Interest & Impact

Downloads

1 since deposited on 22 Feb 2024
1 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