Atmospheric trace metals over the Atlantic and South Indian Oceans: Investigation of metal concentrations and lead isotope ratios in coastal and remote marine aerosols

Abstract

Atmospheric concentrations of trace metals over the oceans are investigated through analysis of aerosol samples collected during cruises from the UK to the Falkland Islands and from South Africa to Australia. The readily soluble concentrations of Cu (4–256 pmol m⁻³), Ni (0.1–54 pmol m⁻³), Ba (0.2–60 pmol m⁻³), Zn (6–316 pmol m⁻³), Cd (0.01–0.29 pmol m⁻³) and Pb (0.4–22 pmol m⁻³) were measured in the aerosols, along with total concentrations of crustal elements (Fe, Al and Mn) to evaluate the crustal contributions. Air mass back trajectories suggested most of the aerosol samples had spent several days over the ocean prior to collection. The highest metal concentrations were observed in aerosols close to South Africa, Australia and major cities in South America, although these concentrations were lower than had been reported previously in the literature. Apart from Ba, which had a major crustal source, the trace metals were enriched relative to crustal sources in most samples, including some collected thousands of kilometers from emission sources. The mean trace metal concentrations in the remote Indian Ocean were lower than those measured in the Atlantic Ocean. Even lower concentrations are reported in the literature for the remote Pacific Ocean. In contrast to previous studies, no clear north–south gradient is observed in the concentrations of the trace metals in the aerosols.

Lead isotope measurements were also carried out on aerosol samples using a multicollector inductively coupled plasma mass spectrometer to assist in source apportionment. Clear differences were noted in the isotope ratios collected on either side of the Indian Ocean with Australian lead ore dominating over much of the eastern and mid-southern Indian Ocean. Samples collected over the western Indian Ocean and Atlantic Ocean under South African influence had lead isotopes quite different from those seen in South African cities in the past, and are closer in ratio to the coal signature of the region.

Introduction

Atmospheric input is now recognised as a major source of metals to the oceans (Duce et al., 1991). River inputs occur only at coastal locations, whereas atmospheric inputs are widely dispersed. Apart from Hg, which is mainly found in the gaseous phase in the atmosphere, trace metals are associated with aerosols in the atmosphere (Duce et al., 1991; Slemr et al., 1985). These aerosols are removed from the atmosphere through both dry and wet deposition processes. Aerosols may have natural sources such as windborne dust, sea spray and volcanoes, or may be derived from anthropogenic processes (Seinfeld and Pandis, 1998). The larger particles (1–5 μm) tend to represent mechanical sources of metals, such as soil dusts and sea spray (Duce et al., 1976; Prospero et al., 1983). Natural emission sources from desert soils are believed to dominate emissions of iron and aluminium, and this input of iron is believed to represent an important source of this micronutrient to many ocean areas (Bruland et al., 1994; Duce and Tindale, 1991; Hydes, 1983). The anthropogenic processes leading to trace metal emissions tend to be high-temperature processes such as metal processing, fuel combustion or waste incineration (Nriagu and Pacyna, 1988). Due to the high-temperature sources of these metals, they are predominantly emitted in gaseous form. These metals rapidly preferentially condense onto smaller particles with the highest availability of surface area (Blando and Turpin, 2000). Hence, in the atmosphere volatile elements such as Pb, Cd, Zn and Cu are mainly found on particles <1 μm in size (Duce et al., 1991). Larger particles are generally removed from the atmosphere more rapidly than those of a smaller size, and hence small particles may be transported over vast distances (Raes et al., 2000). Anthropogenic emissions of trace metals may thus have an important influence on remote regions. Quantification of these inputs is a prerequisite for their effective management.

Studies of aerosols in remote regions of the Pacific and Atlantic oceans have shown that many trace metals are enriched above crustal and oceanic sources (Arimoto et al., 1987; Church et al., 1990; Duce et al., 1975). Air mass back trajectories have implicated North America and Europe as the major source regions for these metal enrichments (Huang et al., 1996). The rise in the concentration of trace metals in the atmosphere since the industrial revolution, and subsequent decline since the 1970s, has been recorded in ice, sediment and peat records (e.g. Boutron et al., 1994; Candelone et al., 1995; Murozumi et al., 1969; Nriagu et al., 1979; Shotyk et al., 1998). The amount of lead present in the atmosphere has received particular attention reflecting the importance attached to the global contamination resulting from the use of Pb in car fuels (e.g. Chow and Johnstone, 1965; Flament et al., 2002; Grousset et al., 1994; Nriagu, 1990; Rosman et al., 1993; Schaule and Patterson, 1981; Véron and Church, 1997; Véron et al., 1998; Wu and Boyle, 1997). The phasing out of lead as a fuel additive allows studies of the global response to these changing emissions. In the 1980s, Patterson and Settle (1987) were able to demonstrate that lead pollution was particularly evident in the Northern Hemisphere with steep gradients away from the major European and North American source regions. The removal of lead from car fuels has rapidly changed this pattern (Huang et al., 1996), and we show here that in samples from the remote Atlantic and Indian Ocean, lead concentrations are now lower than those in the 1980s and that global gradients have been reduced, reflecting increasing Southern Hemisphere emissions from residual leaded fuel combustion and other sources.

The isotopic composition of lead has been used as an indicator of anthropogenic lead in many studies. Lead has four naturally occurring long-lived isotopes, ²⁰⁴Pb, ²⁰⁶Pb, ²⁰⁷Pb and ²⁰⁸Pb, which are formed by radioactive decay of ²³⁸U, ²³⁵U and ²³²Th. The amount of each of these isotopes in a lead ore is dependent upon its age and the initial composition of the source rock. In older lead ores, less radioactive decay occurred before the ore separated from its parent and these tend to have isotopic ratios with low ²⁰⁶Pb/²⁰⁷Pb and high ²⁰⁸Pb/²⁰⁶Pb values (Doe, 1970). Geologically old Australian lead ores have the characteristically less-enriched lead isotope signature (Grousset et al., 1994). In younger lead ores, more radioactive decay has occurred and these ores are isotopically enriched with high ²⁰⁶Pb/²⁰⁷Pb and low ²⁰⁸Pb/²⁰⁶Pb ratios. During environmental and industrial processes, this isotopic ratio remains unchanged as there is no further fractionation (Doe, 1970).

European lead added to petrol was predominantly from Australian ores, whereas lead added to North American fuels was generally from younger Mississippi lead ores (Shirahata et al., 1980). The differences in the isotopic ratios of the lead used by these regions have allowed the identification of the source of lead in remote (Hamelin et al., 1997; Rosman et al., 1993; Véron et al., 1998) and industrialised regions (Flament et al., 2002; Hopper et al., 1991; Veron et al., 1999). The extensive use of lead as a fuel additive resulted in a shift in the isotopic ratio of surface waters in oceans compared to deeper waters (Shen and Boyle, 1988), and as the proportion of particular lead ores used by various countries changed, this was reflected in the isotopic lead in remote surface ocean waters (Flegal et al., 1984). Decreased lead emissions following the phase out of leaded petrol was reflected in decreased lead in surface ocean waters (Boyle et al., 1986; Wu and Boyle, 1997). The decline in lead use in fuels and the attendant decline in lead aerosol concentrations means that minor sources with more diverse isotopic signatures, such as coal burning and smelting, and even natural crustal signatures, may now become more important (Bollhöfer and Rosman, 2000).

In this paper, we present new information on trace-metal aerosol concentrations over the Atlantic and Indian Ocean alongside lead isotope results for the same samples and use these to consider sources of metal inputs to these remote ocean areas.