Direct radioisotope dating of ancient ice

The polar ice sheets constitute a very valuable archive of the past, in particular reflecting paleo-atmospheric and paleo-climatic conditions on Earth. Ice core research therefore became one of the major fields in studying the past, reaching currently back to about 800,000 years. Older ice may still exist, e.g. in the Dry Valleys of Antarctica, where indirect dating indicates several million years. Although several indirect methods have been applied to date such old ice, direct dating of ice in this time range and beyond remains elusive. The current project proposes to use the atom ratio of two cosmogenic radioisotopes, 10Be (t = 1.386 Ma) and 26Al (t = 0.717 Ma) as a chronometer to date old ice. In a closed system, which ice might be, the initial 26Al/ 10Be atom ratio decreases with age with an effective half-life of 1.49 Ma. The ratio of two radioisotopes with similar properties both in cosmogenic production and atmospheric transport is likely to be more reliable for dating than a single radioisotope. The basic requirements for this method to work is: i) an input ratio which is globally constant both in spatial distribution on Earth and in time, ii) a robustness against fractionation of the two radioisotopes after deposition in the ice. Our goal is to prove that direct dating of ancient ice in the range from 0.5 to 5 million years with this method is feasible. In a previous FWF project (P17442-N02, "Study of Cosmogenic 26Al in Atmospheric and Climate Research"), detailed studies about the little-explored meteoric 26Al and first measurements of 26Al/ 10Be ratios in the atmosphere and in deep ice were performed, with encouraging results (Auer et al., Earth Planet. Sci. Lett., 2009, online). Here we propose: i) to improve the analytical aspects of the dating method with the 26Al/ 10Be atom ratio well beyond our previous project, i.e. to substantially lower amounts of ice needed, and to extend its application to mineral loaded ice, ii) to understand excursions of the 26Al/ 10Be ratio observed in deep ice samples (fractionation), iii) to apply the adequately improved dating tool at available basal ice core sections as well as to million-year-old Antarctic rock glacier remnant. An important part of this project is a close collaboration with the Ice Group of the Institut für Umweltphysik, Universität Heidelberg, Germany. This will help for general advise in all matters concerning ice collection and treatment, and for providing contact to drill sites, where basal ice may be available. Concerning access to ancient ice from the Dry Valleys in Antarctica, we collaborate with the Cosmogenic/Stable Isotope Geochemistry Group at the Faculty of Earth and Space Sciences of the University of Washington in Seattle, USA. Another potential collaborator with great interest in the project is the Antarctic Research Centre of the Victoria University in Wellington, New Zealand, which has ongoing programs on the ancient history of Antarctica. Sample preparation including all tests with "dirty" ice, and the AMS measurements of 10Be and 26Al, will be performed at the VERA Laboratory, Fakultät für Physik, Universität Wien, Austria, which has extended experience in this kind of measurements.

The research project P 22401-N21 Direct radioisotope dating of ancient ice is the follow-up of the project P17442-N02 Cosmogenic Al-26 in atmospheric and climate research. In the previous project aerosol measurements showed that the atmospheric 26Al/10Be atom ratio of the cosmogenic radionuclides 26Al (t1/2= 0.72 x106 a) und 10Be (t1/2= 1.386 x 106 a) exhibits little geographical variation. Further it was found that the 26Al/10Be atom ratio in surface ice is in good agreement with the atmospheric value of (1.89 ??0.07) x 10-3. For an age determination in ice cores this ratio must only change according to the different half-lives, and not due to other processes in the ice causing depletion or enrichment (fractionation) of 26Al relative 10Be (or vice versa). The aim of the present project was to answer open questions which arose from the results of the previous project and to investigate if changes of the meteoric 26Al/10Be ratio (effective half-life: 1.49 x106 a) in ice cores are suited to establish a dating method for old ice in the age range from ~0.5 to ~5 Million years. In a pilot study Auer et al. determined the 26Al/10Be ratioin ice (drill chips) samples from ~2500 to 2760 m depth of the Antarctic EDML ice core (M. Auer, PhD thesis 2009). A first age estimate of ~0.6 x106 a was obtained from the measurement of the 26Al/10Be ratio for the deepest ice sample. However, measurements on several ice samples located ~170 m above the dated sample were less encouraging, resulting in 26Al/10Be values significantly higher than the atmospheric ratio. The cause of this anomaly was left unclear. In the current project we investigated several ice samples (drill chips) originating from a similar depth in the EDML core. In order to verify whether enhanced 26Al/10Be ratios found in the previous investigation are caused by fractionation or contamination in the course of the chemical sample preparation, we applied two different melting methods. Both methods also yielded 26Al/10Be values significantly above the atmospheric value. Thus we conclude that the effect is unlikely due to the used chemical methods. This assumption is corroborated by the results of two test samples from the near surface part of the Greenland NEEM ice core. For one of the two samples also a 26Al/10Be value above the atmospheric one was measured. In contrast to the dirty drill chip samples from the EDML core, the NEEM samples came from a pure and clean part of the ice core. This finding suggests that the anomalies in the 26Al/10Be atom ratios are probably caused by effects inherent to the ice cores. From the current state of affairs we therefore conclude that according to our results a radiometric dating method of old ice based on the 26Al/10Be atom ratio does not seem to be promising.