Snow and Ice Core Records of Airborne Trace Metals

In spite of the abovementioned difficulties, snow and ice cores represent one of the most valuable tools for reconstructing climate history and changes in atmospheric contamination. Other depositional environments such as peat bogs and lake sediments have been widely used for recording past changes in atmospheric aerosol chemistry (e.g. Shotyk et al. 1998; Chillrud et al. 1999; Weiss et al. 1999). These approaches may show limitations in temporal resolution due to physical mixing, low accumulation rates or bioturbation, and the interpretation of data can be complicated by the relative magnitude of other sources with respect to the atmosphere. The skeletons of annually banded corals have also been used to record surface ocean concentrations of metals, which largely reflect atmospheric sources. However, the dynamics of the upper ocean and the residence time of elements in the marine environment filter the atmospheric input signal and generally limit temporal resolution to a couple of years (Sherrell et al. 2000). Other long-term archives of atmospheric metal contamination such as tree rings or herbarium collections have been used; however, with respect to all these environmental matrices, the chemical composition of permanently frozen snow and ice is more directly related to that of the overlying atmosphere. Furthermore, snow and ice cores provide detailed records of atmospheric deposition (including individual precipitation events) over timescales of years to hundreds of millennia. For these reasons, in spite of difficulties in the quantitative evaluation of airborne contaminant concentrations, studies on snow and ice deposits have provided most of our knowledge on the history of climate (see Chap. 2) and atmospheric metal pollution. In addition to studies on snow and ice cores from Greenland and Antarctica to assess hemispheric or global-scale deposition of persistent atmospheric contaminants, during the last two decades research has also been performed on ice cores from the Alps (e.g. Wagenbach 1989) and high-altitude tropical glaciers. Studies in the latter regions developed after the pioneering work of Thompson and co-workers (1985), which showed the importance of data from ice cores collected near sources of moisture and atmospheric contaminants for reconstructing palaeoclimate and atmospheric contamination processes (e.g. Thompson 2000).

5.3.1 Lead as a Paradigm of Hemispheric-Scale Anthropogenic Impact

Concern for the possible effects of trace metal atmospheric pollution on global and/or hemispheric scales was raised when Murozumi et al. (1969) found evidence of extensive Pb contamination in ice cores from Greenland, due to anthropogenic sources in the Northern Hemisphere. This paper reported data from three polar locations, and meteorological and environmental differences among the sites may have affected time-series records; however, the finding remains substantially unchallenged and constitutes a landmark for research on atmosphere pollution. Although extremely low contaminant concentrations (usually pg g-1 or less) severely challenged environmental research on polar ice sheets for many years, during the last decade sample collection methods, contamination control and the sensitivity of analytical techniques have improved greatly. Table 2 reports literature data on Pb concentrations in snow deposited during the 1970s and 1980s in Antarctica (Gorlach and Boutron 1992; Barbante et al. 1997; Planchon et al. 2001), Greenland (Boutron et al. 1991), the Andes (Correia et al. 2003) and the Alps (Ros-man et al. 2000). Although different environmental conditions and large variations in Pb concentrations do not allow reliable comparisons between values measured in different regions, the data in Table 2 clearly show that, during the periods of maximum leaded-gasoline consumption, Pb concentrations in Antarctic snow were much lower than those in snow from all other remote areas of the world.

Historically, most anthropogenic sources of Pb and other airborne metals were located in the Northern Hemisphere, and evidence suggests that the impact of human activities began in ancient Greek and Roman times (e.g. Candelone et al. 1995). The analysis of ice cores and other environmental archives such as peat bogs, sediments or tree rings usually show that in more recent times, the atmospheric deposition of Pb in Europe peaked during the 18th century (from coal burning, ferrous and non-ferrous smelting, and open

Table 2. Range of Pb concentrations (pg g-1) in snow deposited in the 1970s and 1980s in Antarctica, Greenland, the Andes and the Alps



Elevation (m)


Pb range (Pg g"1)



Coats Land


34'S,25° 22'W




Planchón et al. (2001)

Hercules Névé


06'S, 165° 28'E




Barbante et al. (1997)

Adélie Land


OO'S, 137° 46'E




Gôrlach and Boutron (1992)




35'N, 37° 38'W




Boutron et al. (1991)


Nevado Illimani


37'S, 67° 46'W




Correia et al. (2003)


Mont Blanc


50'N, 6° 48'E




Rosman et al. (2000)

waste incineration during the Industrial Revolution), and between the 1960s and 1980s, mainly due to the combustion of leaded gasoline by vehicles. During the last two decades, concentrations of Pb in Greenland snow have decreased in response to the remarkable reduction in automotive emissions (e.g. Candelone et al. 1995). These authors suggested that a 6.5-fold reduction in atmospheric Pb concentrations occurred between the 1970s and 1992, and concluded that current values are below those recorded at the beginning of the Industrial Revolution. However, high-resolution (subseasonal) studies on recent Greenland snow (Boyle et al. 1994; Cheam et al. 1998; Sherrell et al. 2000) show that, owing to high spatio-temporal variability in the deposition flux, which depends on the meteorological and environmental features of the sampling site, it is difficult to accurately quantify the magnitude of the decrease in Pb deposition. Sherrel et al. (2000), for instance, found that Pb and Cd concentrations in snow samples deposited in the period 1981-1990 at Summit (Greenland) show order-of-magnitude seasonal variability, with maxima in the spring of each year. The seasonal and inter-annual variability is so large as to complicate the assessment of decadal-scale trends and the effective reduction in Pb concentrations resulting from the phasing-out of leaded gasoline. During the 1981-1990 decade, a small decrease in Pb (<5 %) was estimated and no significant trend for Cd was found (Sherrel et al. 2000). In Greenland snow both metals were still dominated by anthropogenic sources, and Pb isotopic ratios (206Pb/207Pb and 208Pb/207Pb) indicated seasonally distinct source regions - from eastern Europe and the former Soviet Union during spring maxima, and a mixture of US and European sources for the seasonal Pb minima.

Although the interpretation of data from snow and ice cores collected in European glaciers is further complicated by periodic melting or percolation of meltwater, 1993-1996 records of atmospheric Pb deposition and Pb isotopes in snow at Jungfraujoch (Switzerland, about 3,500 m a.s.l.; Döring et al. 1997) indicated that the emission of Pb from traffic had decreased significantly, but was still detectable and in the same range as that from other anthropogenic sources such as waste incineration. Rosman et al. (2000) performed a detailed analysis of Pb concentrations and Pb isotopes in snow deposited at Mont Blanc and found large seasonal variations, especially in winter when a low-altitude inversion establishes in the area. 206Pb/207Pb ratios in snow from Mont Blanc and Greenland were considerably different, particularly in the period 1969-1980: while Mont Blanc samples were dominated by Australian Pb used in petrol in the Piedmont region of northwest Italy, those from Greenland were dominated by the highly radiogenic Mississippi valleytype Pb from the USA.

The low Pb concentrations in Antarctic snow make it difficult to carry out analyses and complicate the differentiation between natural and anthropogenic sources. Average annual precipitation on the Antarctic plateau is very low (usually 2-4 g cm-2 year-1), with the advantage that a 1-m core in Antarc tic ice covers many years; however, this makes it difficult to use stratigraphic methods, because a single severe storm could blow away an entire annual layer (Hammer 1982). As a result, trace metal concentrations in snow and firn samples usually show large intra- and inter-annual variations. In eight snow-ice sections of two well-dated cores from a small, coastal ice cap (Law Dome, Wilkes Land) with high snow-accumulation rates (64-116 g cm-2 year-1), Pb and Cd concentrations in winter were two- to fourfold greater than in spring-summer (Hong et al. 1998). A very high metal concentration variability was also found in samples from two snow pits in Coats Land, covering the 1920-1990 period (Planchon et al. 2002b); the highest measured Pb, Cr, Mn and U concentrations, for instance, were approximately 100 times higher than the lowest ones. Thus, while Antarctic ice cores allow the assessment of average Pb concentrations during a climatic cycle, with low values during the Holocene (about 0.4 pg g-1) and relatively high values in the cold terminal stage of the last ice age (about 14 pg g-1; Boutron et al. 1987), metal deposition trends during past decades or centuries are difficult to assess.

Was this article helpful?

0 0

Post a comment