Abstract: Spectrofluorimetric and spectrophotometric continuous flow analysis (CFA) methods were developed and applied to the determination of aluminium and iron in EPICA Dome C (East Antarctica) ice-core samples (6-585 m depth). The methods are able to measure the fraction of Al and Fe which can be detected once the sample is filtered on a 5.0 ?m membrane and acidified to pH 2. Both the methods present high sensitivity (detection limit of 10 ng L-1 for Al and 50 ng L-1 for Fe) and reproducibility (5% at sub-ppb level). The Fe and Al profiles show sharp decreases in concentrations in the last glacial/interglacial transition, reflecting the decreasing dust aerosol load. The two elements show a different pattern during the Antarctic Cold Reversal (ACR) climatic change, with high iron concentrations (similar to the glacial period) and low but increasing Al content during the ACR minimum. In order to interpret the Al and Fe data obtained by CFA, a comparison with Al and Fe composition, as measured by inductively coupled plasma sector field mass spectrometry (ICP-SFMS), was performed for Holocene, ACR and glacial periods. The percentage of CFA-Al with respect to ICP-SFMS-Al in the three periods shows a lower variability than CFA-Fe (3% in the glacial period and 64% in the ACR). This pattern may be explained by the different dominant iron sources in the different climatic periods. During the Last Glacial Maximum, Fe is proposed to arise mainly from insoluble continental dust, while a variety of ocean-recycled Fe, mainly distributed in fine particles and as more soluble species, shows a higher contribution in the ACR and, to a lesser extent, in the Holocene.

Abstract: Total-dissolvable iron has been measured in sections of three ice cores from Law Dome, East Antarctica, and the results used to calculate atmospheric iron deposition over this region during the late Holocene and to provide a preliminary est. of aerosol iron deposition during the Last Glacial Maximum (LGM). Ice-core sections dating from 56-2730 BP (late Holocene) and ~18 000 BP (LGM) were decontaminated using trace-metal clean techniques, and total-dissolvable iron was determined in the acidified meltwaters by flow injection analysis. Our results suggest that the atmospheric iron flux onto the Law Dome region has varied significantly over time-scales ranging from seasonal to glacial-interglacial. The iron concentrations in ice-core sections from the past century suggest (1) a 2-4-fold variation in the atmospheric iron flux over a single annual cycle, with the highest flux occurring during the spring and summer, and (2) a nearly 7-fold variation in the annual max. atmospheric iron flux over a 14 year period. The average estimated atmospheric iron flux calculated from our late-Holocene samples is 0.056-0.14 mg m-2 a-1, which agrees well with Holocene flux estimates derived from aluminum measurements in inland Antarctic ice cores and a recent order-of-magnitude estimate of present-day atmospheric iron deposition over the Southern Ocean. The iron concentration. of an ice-core section dating from the LGM was more than 50 times higher than in the late-Holocene ice samples. Using a snow-accumulation rate estimate of 130 kg m-2 a-1 for this period, we calculated 0.87 mg m-2 a-1 as a preliminary estimate of atmospheric iron deposition during the LGM, which is 6-16 times greater than our average Late-Holocene iron flux. Our data are consistent with the suggestion that there was a significantly greater flux of atmospheric iron onto the Southern Ocean during the LGM than during the Holocene.