EVIDENCE OF LONG-RANGE TRANSPORT OF AEROSOLS IN THE SOUTHERN HEMISPHERE FROM MEASUREMENTS OF LEAD ISOTOPES.

K J R Rosman¹, A F Bollhofer¹, G Burton¹, A L Dick²

¹Department of Applied Physics, Curtin University of Technology, GPO Box U 1987, Perth 6845, Australia. Email: rrosmank@cc.curtin.edu.au, Fax: 61 8 922662377.

²Institute of Antarctic and Southern Ocean Studies, University of Tasmania, Hobart, Australia.

ABSTRACT

The concentration and isotopic composition of lead has been determined in aerosols in boundary layer air from the Southern Ocean, in order to characterize the ambient background in the southern hemisphere. Ultraclean sampling devices and procedures were developed for this purpose and have revealed Pb concentrations to ~1pg/m³, which is comparable with the lowest measured in Antarctica. The isotopic ratios differ significantly from those found in aerosols in urban environments of the most southerly regions of South America, Africa and Australia, but show closer agreement with more northerly regions of South America and Africa. A large proportion of the samples contain small but significant amounts of highly radiogenic Pb which may arise from uranium mining in Africa or Australia.

INTRODUCTION

A recent study of Pb isotopes in aerosols in the Northern and Southern Hemispheres has shown that there is a wide variation in isotopic signatures in atmospheric emission sources (Bollhofer and Rosman, 2000a, 2000b). Since Pb is often present at detectable levels in polluted air masses it might therefore be used as an atmospheric tracer for monitoring the long-range transport of pollutants under suitable conditions. Isotopic tracers are preferred to elemental tracers because the signal cannot easily be modified by the physical and chemical processes associated with atmospheric transport. The tracer potential of Pb has been recognized for some time and has already been demonstrated in other studies (for instance, Maring et al., 1987; Hopper et al., 1991; Sturges and Barrie, 1987; Rosman et al, 1993).

In this study we have measured the concentration and isotopic composition of Pb in aerosols in order to characterize the present-day southern hemisphere background, and also to investigate the possibility of using Pb isotopes to detect long range aerosol transport over the Southern Ocean.

METHODS

Sampling location

Aerosol samples were collected at Cape Grim Baseline Air Pollution Station, located on the northwest tip of Tasmania, Australia (Figure 1). The station was set up in 1976 to monitor long-term changes in a range of atmospheric species and characteristics, including greenhouse gases, pollutant and natural aerosols, and meteorological parameters. Of primary interest are measurements made in clean, marine, 'baseline' air that is free of local pollution, as such measurements are characteristic of the atmosphere over a large portion of the Southern Hemisphere. Strong winds from the west or south west of the station typically bring such atmospheric conditions, and allow a baseline signal to be identified. Measurements of meteorological parameters and of most gaseous species are made continuously at the station, but samples of aerosols for chemical analysis are generally only collected during 'baseline' conditions. A 'baseline switch' allows sampling to be carried out only during periods when the wind direction lies between 190 and 280 degrees, and total particle counts (CN) are below 600 per cubic centimetre. These limits, calculated as one minute averages, have been determined empirically to indicate clean 'baseline' air, and occur approximately 33 % of the time at the station. Samples for Pb isotope determination were collected at the station between February 1999 and April 2000. To minimize contamination from local soil dust, baseline air was sampled from the top of a 70 m steel tower located on the 94 m high coastal cliffs approximately 100 m from the ocean. Baseline conditions had to exist for 30 minutes prior to the start of sampling, and sampling ceased immediately if either the wind direction or particle criteria went outside the limits. The sampled air volume was measured at the base of the tower on the exhaust side of the pump, using a calibrated gas meter.

Sampling

The filters were mounted inside a weather shelter consisting of a hollow cylinder constructed from polyethylene (PE) and teflon. The cylinder is divided lengthwise into two chambers by a PE plate. The filter monitor was mounted in the top chamber and air entered the monitor from below through a hole in the PE plate. During baseline sampling, air can enter the lower chamber. With non-baseline conditions the lower chamber was sealed pneumatically and pumping was stopped. Four samplers were mounted on the tower at one time to minimize the number of excursions to the top of the tower, and also to allow exposure blanks to be taken. A vacuum pump and compressor needed to operate the sampler were located at the base of the 70m tower and were switched according to the prevailing wind conditions. The samplers were cleaned and assembled in a clean laboratory supplied with HEPA filtered air then sealed in a PE bag and protective padding for transport from Curtin University (Western Australia) to Cape Grim (Tasmania).

The sampling protocol was as follows. The pumping chamber was opened 29 minutes after baseline conditions were detected. Thirty seconds later the pump started and a further thirty seconds later a valve was opened to allow air to be drawn through the filter. On detection of non-baseline conditions, the valve closed immediately, the pump switched off and the sampler closed. Tower top changes – a major undertaking – occurred at approximately 4 or 5 week intervals, and were carried out using clean room techniques to the fullest extent possible. This included wearing full clean area clothing and plastic gloves. The monitors were sealed using a "blind" PFA nut immediately after being taken from the shelter, then sealed in 2 PE bags and wrapped in bubble-wrap for protection. A complete history for exposure, including pumping times and volume pumped is logged for each filter.

Filters

The aerosols were collected on 47 mm diameter PTFE filters (0.47 mm front, 60 mm back) contained in PFA monitors (both supplied by Savillex Corporation, USA). Both monitor and filters were extensively acid cleaned and then assembled under clean-room conditions. Procedures used to prepare the filters are described by Bollhofer et al. (1999).

Sample processing.

The filters were cut in halves or quarters with an acid-cleaned stainless steel scalpel blade, then leached in 0.4 M HBr for 2 hours. Pb was extracted using BioRad AG1 X8 (100-200 mesh) anion exchange resin using HBr to load and HCl to elute. Every ion exchange column was blanked before the actual sample was loaded, so as to detect and avoid high Pb levels from resin, frits or acids. In addition, procedural blanks were measured with each batch of samples. They were typically 20 pg, most of which came from the acids. All reported concentrations and isotopic ratios were corrected for the blank.

Mass spectrometry.

Isotopic analysis was performed with a VG354 magnetic sector TIMS fitted with a 16 sample turret, 9 Faraday collectors and Daly collector that operated in current mode. The precision was determined from the reproducibility of the mean of multiple blocks of isotopic ratios. Data were corrected for bias due to isotopic fractionation by 0.12±0.05% per mass unit for the Faraday and 0.24±0.06% per mass/unit for the Daly collector. Pb determinations were made by isotope dilution (IDMS) using a ²⁰⁵Pb spike. This allowed both the isotopic composition and the concentration to be determined in the same measurement (Chisholm et al., 1995). Following chemical processing the sample was transferred to the central ionizing rhenium filament in a few mL of a silica gel/phosphoric acid mixture.

RESULTS AND DISCUSSION

Lead concentrations

Sampling began in February 1999 and continued for 12 months with filters being changed at approximately monthly intervals. The change in the concentration of Pb in the baseline air over this period is shown, together with the corresponding volumes of air collected, in figure 2.

The Pb concentrations are extremely low, with the lowest, 0.6±0.1 pg/m³, obtained in June-July, 99. This is comparable with the lowest, 1.4 ± 2.4 pg m^-3, reported by Dick (1991), on the Antarctic Peninsula at the Gipps Ice Rise in 1985.

Figure 2. The concentration of Pb in baseline air collected at Cape Grim. Error bars shown for the concentrations do not include the uncertainty in the volume pumped.

Lead Isotopes

Isotopic ratios measured in the same samples are shown in Figure 3. Here the aerosol data are superimposed on a plot of Pb isotopic ratios in aerosols collected in the Southern Hemisphere during 1997-99 (Bollhofer and Rosman, 2000a). The isotopic signatures of Pb in the Cape Grim aerosols are different from those found in urban aerosols in Australia, South Africa and southern South America. There appears to be closer agreement with more northern countries such as Indonesia, Malaysia, northern Africa and Brazil.

Of particular significance however, is the shift in the baseline aerosol data from the trend of Pb isotopic ratios normally found in emissions from urban areas. This shift indicates a relative enhancement of the ²⁰⁶Pb/²⁰⁷Pb ratio. A possible source of such material is uranium mining in the southern hemisphere. Possible sources are South Australia or the Northern Territory. However, although the distance is much greater, another possibility is Gabon in central west Africa which is extensively mined for uranium. Supporting this possibility, the Earth Probe TOMS (Total Ozone Mapping Spectrum, http://toms.gsfc.nasa.gov) aerosol index satellite data shows this region to be a source of aerosols and shows that they can travel down the west coast of Africa then westward

across the Southern Ocean towards Australia. Biomass burning in Gabon is expected to be significant source of fine dust particles containing highly radiogenic Pb.

Figure 3. Isotopic ratios of Pb in baseline aerosol collected at Cape Grim (solid circle), compared with Pb in urban aerosol from the southern hemisphere. The date sampling began is indicated.

Also, the extremely low background existing over the Southern Ocean may have allowed the radiogenic Pb to be detected, explaining our observations. The relative contribution of these and perhaps other sources has yet to be established.

ACKNOWLEDGEMENTS

This project was supported by grants from the Cape Grim Baseline Air Pollution Station and the Australian Research Council. Technical support was provided at Cape Grim by S Baly, D Evenhuis, C McCulloch, L Porter, and at Curtin by A Parker, T Margrain and P. Brooker, while the isotopic analyses were carried out in the TIMS laboratory of the Centre of Excellence in Mass Spectrometry at Curtin University.

REFERENCES

Bollhöfer A, Chisholm W, Rosman K J R (1999), Anal. Chim. Acta. 390: 227 - 235.

Bollhoker A F, Rosman K J R (2000a and 2000b), Geochim. Cosmochim. Acta (In Press))

Chisholm W, Rosman K J R, Boutron C F, Candelone J-P, Hong S (1995), Anal. Chim. Acta. 311: 141-151.

Dick A L (1991), Geochim Cosmochim. Acta 55: 1827-1836.

Heumann K G (1992), Mass Spectrom. Rev. 11: 41-67.

Hopper J F, Ross H B, Sturges W T and Barrie L A (1991), Tellus 43B: 45-60.

Maring H, Settle D M, Buat-Menard P, Dulac F, Patterson CC (1987), Nature 330: 154-156.

Rosman K J R, Chisholm W, Boutron C F, Candelone J-P, Gorlach U (1993), Nature 362: 333-334.

Sturges W.T. and Barrie L.A. (1987), Nature 329: 144-146 .