HR: 0800h
AN: C51A-0108    [Abstracts]
TI: Measuring U-Series Isotopes in Polar Ice: Toward an Absolute Ice Chronometer
AU: * Aciego, S M
EM: aciego@erdw.ethz.ch
AF: Institute for Isotope Geochemistry and Mineral Resources, ETH-Zurich Clausiusstrasse 25, NW C83.1, Zurich, 8092, Switzerland
AU: Bourdon, B
EM: bourdon@erdw.ethz.ch
AF: Institute for Isotope Geochemistry and Mineral Resources, ETH-Zurich Clausiusstrasse 25, NW C83.1, Zurich, 8092, Switzerland
AU: Schwander, J
EM: schwander@climate.unibe.ch
AF: Climate and Environmental Physics, University of Bern Sidlerstrasse 5, Bern, 3012, Switzerland
AU: Stocker, T
EM: stocker@climate.unibe.ch
AF: Climate and Environmental Physics, University of Bern Sidlerstrasse 5, Bern, 3012, Switzerland
AB: Comparison of ice records between ice sheets, alpine glaciers, and marine records currently rely on a combination of ice layer counting, matching relative time scales, and interpolation. U-series recoil from mineral aerosols (dust) into the ice matrix is one possible technique for determining the absolute age of ice, independent of any other parameters. However, the low concentrations of the U-series parents and daughters have made previous measurements difficult and the results ambiguous. We present here the first results of work we have undertaken for determining U-series recoil ages in ice cores. The primary difficulty of this technique is the extremely low concentrations of dust in polar ice samples, and therefore, of the recoil daughter products in the ice. Previous work on dust provenance indicates 0.01 to 1 mg of dust concentration per kilogram of ice from the ice cores of Greenland and Antarctica. Given these conditions, U and Th dissolved in the water fraction of the aerosol-ice system may overwhelm the total U-series budget. Constraining the possible "initial" U and Th is the first step in determining the feasibility of this dating method for ice cores. We have implemented new geochemical techniques: ultra-clean ice processing, multiple ion counter ICP-MS measurements of U and Th, and quantification of total recoveries of the aerosol and water fractions using both established USGS standards, an internal lab loess standard that best approximates the dust fraction found in ice cores, and [U]-[Th] standards SRM960 and Th105. Dissolution experiments using U and Th spikes with these standards indicate recovery of the dust and dissolved fractions are better than 99%. We present here the first concentration measurements of U from the water fraction (<0.2 microns) of freshly deposited South Pole snow (20pg/kg), as well as a series of measurements from the upper section (~128m) of the Dye 3 ice core in Greenland which thus far range from 410pg/kg to 520fg/kg U dissolved in the water fraction. The high 234U/238U activity ratio (>1) in the Greenland ice core indicates the source of the uranium within the dissolved fraction is not inconsistent with seaspray. Given these results we model, using a simple Monte Carlo simulation, the expected errors in measuring absolute ages of polar ice. Using an estimate of surface roughness factor of 160 (based on recent BET measurements of dust), a lognormal distribution of grain sizes around 2 microns, and concentrations of dust over glacial periods on the order of 0.5mg/kg, we find that our current blanks, detection limits and internal precision should allow us to measure the ages of polar ice with an accuracy of 1% for Antarctic Ice cores and 5-10% for Greenland Ice cores (2 sigma).
DE: 0700 CRYOSPHERE (4540)
DE: 0724 Ice cores (4932)
DE: 1029 Composition of aerosols and dust particles
DE: 1094 Instruments and techniques
DE: 1115 Radioisotope geochronology
SC: Cryosphere [C]
MN: 2007 Fall Meeting