Two snow and ice cores from the Rennick Glacier area, Antarctica (study area center point lat 71°15ˈS, long 162°30ˈE) were analyzed for the chemical species: chloride, sodium, reactive silicate, sulfate and nitrate. Core E10 (6.35 m) was taken from Evans Névé, a large accumulation basin at the head of Rennick Glacier. Core M1 (4.35 m) was extracted from the accumulation zone on the central plateau of the Morozumi Range, a small mountain massif bordering Rennick Glacier. Cores E10 and M1 span the time periods from 1929 to 1981 and from 1971 to 1981, respectively, as dated using seasonal variations in chloride (E10) and sulfate (M1) concentrations. An estimated net balance of 50 kg m−2 a−1 was derived for site E10 and of 182 kg m−2 a−1 for site M1. The difference in net mass balance may be explained by elevational differences (M1 1400 m a.s.1., E10 2305 m a.s.1.). Recent (1973 to 1980) increases in sodium, silicate and sulfate at site E10 may be linked to decreases in Antarctic pack-ice extent (Kukla and Gavin 1981) for the same period.

Most, if not all, Antarctic and Greenland snow and ice samples to be analyzed for toxic metals and metalloids such as Pb, Hg, Sb, Cd, Ag, Se, As, Cu and Zn become more or less contaminated by these elements on their outsides, mainly during field collection. We assess here the various procedures which have been developed to try to decontaminate the samples. They include both mechanical and rinsing techniques. It appears that it is essential, but very difficult, to establish clearly the efficiency of these procedures. This can however be done by determining the geometry of contamination of the analyzed samples and by evaluating procedural blanks carefully. Such careful evaluation has been achieved at present only for mechanical procedures and for a few metals.

Measurements of trace elements in snow and ice are frequently used to describe past atmospheric composition although there is no firm basis for assuming a direct connection. Trace-element concentrations have been measured on samples of aerosol and freshly fallen snow collected simultaneously from two sites in the Antarctic Peninsula during summer. Following improvements in contamination control, the reported concentrations and crustal enrichment factors of Cd, Cu, Pb and Zn in the aerosol are lower than any values previously reported from Antarctica. Even tighter controls will be required in the future.

For a crustal element (A1) and for the marine cations (Na, Ca and K) a consistent ratio (0.48±0.31) for the concentration in air (pg m−3)/concentration in snow (pg g−1) is obtained for simultaneously collected samples. This supports a simple model of aerosol scavenging proposed by Junge which considers aerosol removal over polar ice sheets to be dominated by in-cloud processes. Averaged data for Cd, Cu, Pb and Zn from samples collected at different times appear to behave similarly. These findings suggest that there is no preferential scavenging by snowfall of either crustal or heavy metal components in contemporary aerosol. If proved more general in Antarctica this may help to simplify the interpretation of time series data from ice cores.

The chemistry of recently deposited snow sampled in 1982–83 along a 430 km coast-interior traverse in Terre Adelie, East Antarctica, is reported. In addition, three firn samples, covering the same time period (1959 to 1969) and collected on the traverse at D 55, D 80 and Dome C stations, respectively at 200, 430 and 1070 km from the sea, are also studied. Concentrations of major soluble impurities (H+, , Na+, K+, Cl−, and ) were determined by ion chromatography (except H+ which was titrated) on more than 200 samples. Conditions of sampling and analysis were carefully controlled in order to avoid contamination problems. A balanced ionic budget was generally obtained for each of the samples. For stations occupying an intermediary position between the coastal areas and the central Antarctic plateau, our results demonstrate that the two major impurities are H2SO4 and HNO3. HCl is also present, but at a lower level of concentration; the sea-salt contribution is dominant only at the most coastal sites (within 40 km) of the sea. The degree of neutralization of the snow acidity by NH3 is always very low as indicated by the values of content. The mean concentrations of H2SO4 along the traverse are relatively constant whereas an increase of the HNO3 concentrations is observed when going inland. It decreases, however, in most central areas. These results are discussed in relation to the glaciochemical data published for other locations on the Antarctic plateau, in particular the sulphate concentrations which depend strongly on explosive volcanic activity.

An extensive investigation of the visible stratigraphy, microparticle concentration, liquid conductivity, oxygen isotopes and beta-radioactivity was conducted in pits excavated at Amundsen-Scott South Pole station. The objectives of the investigation were to assess the spatial representativeness of the geochemical and physical records preserved within the snow strata and to ascertain the temporal resolution which can be obtained from such ice-core records. Accurate interpretation of the time scale and reconstruction of climatic conditions from these time series requires (1) the analysis of as many stratigraphic parameters as possible and (2) the synthesis of data from a suite of cores in the study area. For periods of 10 a or less, regionally representative accumulation rates cannot be obtained from annual accumulation time series reconstructed at a single site. Although the microparticle concentrations, liquid conductivity and oxygen isotopic abundances all exhibit a seasonal cycle in the firn, the construction of an accurate time scale requires all three parameters in conjunction with the beta-radioactivity. Absolute dating will be impossible for cores from South Pole where entire accumulation years may be missing. Nevertheless, for East Antarctica, where accumulation rates are low (<0.1 m a−1 water equivalent), the good temporal resolution and the preservation of a distinct annual signal in some geochemical parameters makes the South Pole a very attractive site for deep ice-core drilling.

Dirt layers of tephra were found on the bare ice surface in the Meteorite Ice Field near the Yamato Mountains, Dronning Maud Land, and near the Allan Hills, Victoria Land, Antarctica. The grain-size analyses of volcanic ash fragments show that the mean grain size in the Allan Hills region is larger than that in the Yamato Mountains region. This fact indicates that the volcanic sources of the dirt layers in the Yamato Mountains region are farther away than those in the Allan Hills. Their constituent fragments are well-sorted and composed mainly of volcanic glass shards with minor amounts of crystal fragments.

Glass shards of the tephra from the Yamato Mountains region have a composition of tholeiitic andesite which is low in alkali and high in iron but not so enriched in titanium, and the associated crystal fragments consist of calcic plagioclase, subcalcic clinopyroxene, orthopyroxene and magnetite. The nature of island arc tholeiite of the tephra indicates that its source is some volcano in the South Sandwich Islands.

On the other hand, the tephra from the Allan Hills region is composed of glass shards of trachybasaltic composition and crystal fragments of titanaugite calcic plagioclase, kaersutite, olivine, rhönite and titanomagnetite. A young volcano of the McMurdo volcanic group is suggested as a possible source of this tephra.

In 1968 an ice core 2164 m long was recovered from Byrd station in West Antarctica. About 2000 tephra layers were observed in the core and have been differentiated into ash and dust bands according to the grain size and concentration of particles in the layers. Mount Takahe, a local volcano in Marie Byrd Land, Antarctica is the probable source.

Detailed examinations of the particle morphology, composition and ice chemistry associated with some of the tephra layers have led to the conclusion that the eruptions which produced the layers were probably hydrovolcanic. Melted glacier ice is considered the most likely source of the water involved in the eruptions.

Processes associated with hydrovolcanism such as particle aggregation (causing premature deposition of fine particles), rapid conversion of sulfur dioxide to sulfuric acid (in the presence of abundant moisture) and scavenging of acid droplets by the fine dust particles are inferred to have taken place. Such processes would greatly reduce the atmospheric residence time of the eruptive products and thus their atmospheric and climatic impact.

Since 1979–80, isotope studies with oxygen-18 (18O) have been carried out at several snow pits and ice cores near the German Georg-von-Neumayer station (Ekström ice shelf, Atka Bay), as well as from the Filchner-Ronne ice shelves, in the framework of the German Antarctic research programme. The investigations of snow pits on the Filchner-Ronne ice shelves yield a standard deviation for the annual average δ18O values of approximately 1‰ over the last five years, and a decrease of δ18O with distance from the ice edge of about 1‰ per 50 km. The variation of δ18O for stratigraphically matching snow layers from snow pits at the same location in different years is about 0.3‰ on the Filchner-Ronne ice shelves, and 0.8‰ at Georg-von-Neumayer station. The mean annual accumulation rate in the surroundings of Georg-von-Neumayer station was determined to be 34 g cm-2 for the years 1977–81. On the Filchner-Ronne ice shelves the mean annual accumulation rate (1979–83) decreases from 22 g cm−2 at Filchner station to 15 g cm−2 at traverse point T340, located 200 km southeast of Filchner station.

The composition of air in bubbles of ice reveals important information on the composition of the atmosphere at the time at which the ice was formed. The enclosed air has essentially an atmospheric composition if the ice formation occurred by dry sintering of snow and firn at low temperatures. The history of atmospheric CO2 concentration is of special interest due to its present increase and its climatic implications. Currently, CO2 analysis of air extracted from ice cores seems the best and most direct method of determining the natural CO2 concentration before 1950 and possible pre-industrial variations. Analyses of ice samples representing the past 40 ka show that there were significant changes in concentration of atmospheric CO2 at the end, and probably during part, of the last glaciation. δ13C measurements on CO2 extracted from ice cores can indicate possible mechanisms causing these changes in the concentration of atmospheric CO2.

Recent snow from two sites in the Antarctic Peninsula has been analyzed for Al, Cd, Cu, Pb and Zn. Measurement of full procedural blanks and of the extent of penetration of surface contamination has allowed a rigorous appraisal of both sampling and analytical methods. Whilst the particular samples of cored firn used here have been shown to be unsuitable due to penetration of surface contamination into their interiors, surface samples collected directly into acrylic tubes showed very limited penetration of contamination. The surface samples gave the following average concentrations: Al: 0.7±0.3 ng g−1, Cd: 0.26±0.09 pg g−1, Cu: 1.9±0.5 pg g−1, Pb: 6.3±3.3 pg g−1 and Zn: 3.3±1.7 pg g−1. The Pb concentration agrees well with data from other workers for recent snow from East Antarctica, while the values for Cd, Cu and Zn are about ten times lower than have been reported previously, even for ancient Antarctic ice. Although these data refer to only one site and a short time period, it is believed that they are representative of modern Antarctic snow.

The true concentrations of Cd, Cu, Pb and Zn in ancient Antarctic ice are not yet established. However the following limits are suggested for increases over natural background levels due to anthropogenic emissions: Pb 1 to 40 times, Cd 1 to 180 times, Cu 1 to 4.5 times and Zn 1 to 6 times. These results indicate that contamination may have been responsible for reports of anomalous enrichment of several heavy metals in Antarctic snow.

D 57 station in Terre Adélie lies between the coast and the central Antarctic plateau. A 200 m ice core was recovered in summer 1980–81 at this location and analyzed by an electroconductometric method to detect exceptional acid levels linked to fallout from major volcanic eruptions. Several signals were indeed found. The corresponding ice-core sections were then analyzed for mineral acids (H2SO4 and HNO3). We detected several large volcanic events, in particular two eruptions identified as Tarabora (1815) and Galunggung (1822). The background concentration of sulphate was found to be relatively low (about 0.5 μeq 1−1). On the other hand nitrate values were higher than at coastal or central Antarctic locations (except for the Sauth Pole). Two spikes were found in the nitrate profile at depths of 140 and 148 m. It is thought that they could be either linked to the 1604 and 1572 supernovae Kepler and Tycho or correspond to epochs of particularly high solar activities. With the aid of these sulphate and nitrate exceptional events, a dating of the D 57 ice core can now be proposed which corresponds to a mean snow accumulation rate of 22 cm of ice equivalent per year over the last four centuries.

Because of snow drifting, two time series of any variable derived from two adjacent ice cores will differ considerably. The size and statistical nature of this noise element is discussed for two kinds of measured substance. A theory is developed and compared to data from Greenland and Canadian Arctic ice cores. In case 1, the measured substance can diffuse and the seasonal cycle degrade with time and depth, e.g. δ(18O). In case 2, the measured substance cannot diffuse, e.g. microparticles. The case 2 time series contain drift noise proportional to that in the accumulation series. For accumulation series, the spectral power is concentrated at the high frequencies, i.e. is “blue”. Such noise can be easily reduced by taking relatively short time averages. The noise in the case 1 time series, however, starts out “blue” but quickly diffuses to have a “red” character with significant power at longer wavelengths, and many decades of such series must be averaged to reduce the noise level. Because the seasonal amplitude of any given variable is an important input to the drift noise and because the seasonal amplitudes of some variable types are latitude-dependent, some sites have inherently less drift noise than others.

We present glaciochemical data from a pilot study of two snow-pits from Quelccaya ice cap, Peruvian Andes. These are the first samples to be analyzed from Quelccaya for nitrate and sulfate by ion chromatography (IC), for nitrate-plus-nitrite, reactive silicate and reactive iron by colorimetry, and for sodium by atomic absorption spectrophotometry. The 3 m pits used in this study represent a one year record of mass accumulation and the 29 samples collected provide the first glaciochemical data from this area which can be compared with glaciochemical studies from other locations.

Reactive iron, reactive silicate and sodium, and the profiles of >0.63μm microparticles from Thompson and others (1984) are coincident, suggesting that transport and deposition into this area of each species are controlled by similar processes. The common source is probably local, resulting from crustal weathering. In general, the reactive silicate values are lower than those observed in other alpine glacier ice. The highest sulfate and nitrate values were observed in the upper few centimeters of the snow-pit. Most of the sulfate concentrations were less than 3 μM and are similar to values obtained for fresh surface snows from Bolivia (Stallard and Edmond 1981). Since biological gaseous emissions are thought to be the major source of sulfur and nitrogen to the atmosphere over the Amazon basin, the sulfate and nitrate fluctuations may be due to seasonal biological input and/or seasonal shifts in wind direction bringing material to Quelccaya.

With only one exception, the colorimetric nitrate-plus-nitrite data were higher than the IC nitrate data. Unfortunately, the IC analyses were conducted 81 d after the colorimetric analyses. The difference between the two data sets could be attributable to the following: (1) the colorimetric technique may yield erroneously high results as suggested for polar ice by Herron (1982), (2) the IC technique yields erroneously low results due, in part, to the possible exclusion of nitrite concentrations, and/or (3) nitrite was lost via biological removal during the 81 d period before the IC analyses. If the IC data are correct, the mean nitrate value is 0.4μΜ (n = 29). This value is similar to those reported from pre-industrial aged polar ice (Herron 1982). If the colorimetric mean value (1.1 μM) is correct, it is similar to colorimetrically determined values from other high-elevation alpine ice (Lyons and Mayewski 1983).

Since 1979 we have been conducting a program of glaciochemical sampling and analysis in selected portions of the Indian Himalayas. The primary purpose of this work has been the retrieval of data that are of specific use in assessing the signal expressed by the chemistry of air masses entering the Himalayas. The techniques used for this purpose provide data sets for the following: chloride, sodium, reactive iron, reactive silicate, reactive phosphate, nitrite-plus-nitrate, ammonium, pH, oxygen isotopes, deuterium, microparticles, total β-activity, density and scanning electron microscopy. The results of this work appear in a series of papers (Lyons and others 1981, Lyons and Mayewski 1983, Mayewski and others 1981, 1983, 1984 and Goss and others 1985). In summary this work demonstrates: (1) problems encountered in high-altitude ice-core recovery, (2) effects of percolation on chemical records, (3) specific requirements necessary for the retrieval of unaltered glaciochemical records from Himalayan glaciers, (4) potential spatial variability of chemical species concentrations and interpretation of this with respect to time series, (5) usefulness of various glaciochemical indicators as applied to relative dating (seasonality) and air mass tracking, (6) specific details of the chemical and physical properties in Himalayan ice, and (7) recommendations for future Himalayan ice-core studies.

Core drillings were carried out in 1979 and 1983 on the temperate Alpine glacier Vernagtferner, Ötztal Alps, Austria. In addition, studies on the variation of the water table of the firn were carried out in the bore holes. The 18O, 2H and 3H content of the cores, as well as the electrical conductivity, were measured. The isotope profiles and the electrical conductivity of the 1983 core display a strong damping in comparison with the profiles from the 1979 core for the depth from 17 to 20 m (10.5 to 13 m water equivalent) below the surface. In the 1983 core, the 3H contents are lower, and the 2H and 18O contents are more homogenized. This core section lies between the lower and upper limits of the water table fluctuations from 1979 to 1982, and it is also assumed that the transition from firn to ice takes place at this depth of the glacier. The most reasonable explanation for this damping is that within the range of the firn water table important isotopic and chemical exchanges between the meltwater and the porous medium (firn) take place. It is also possible that a small amount of meltwater is included in the firn while it is changing into ice.

We report trace element, tritium, and hydrochemical analyses for glaciers in mountain regions of western China, based on about 200 samples collected between 1980 and 1983. Conclusions are drawn about the origin of water vapour contributing to precipitation (local or otherwise) and about the contributions to glacier run-off (meltwater or precipitation), based on the tritium measurements.

Atmospheric particles originate from a variety of natural and anthropogenic sources. Estimates of man’s contribution to the total particle mass loading of the atmosphere range from 5 to 48%. Much of this is initially in the form of gases (SO2, NO2, hydrocarbons) that convert to particles while being transported to glacial receptors where they are incorporated into snow. The complex physical and chemical processes involved in the deposition of atmospheric particles to glaciers are reviewed. Both wet and dry deposition contribute to the pollutant loading of a snowfield. However, except in the case of low snowfall (<60 kg m−2 a−1) or exceptionally large particle sizes, such as might be released by volcanoes or when unrimed snowfall predominates, wet deposition is dominant. Estimates of the relative importance of wet and dry deposition based on scavenging ratios for rimed snowfall and dry deposition velocities agree well with observations in southern Greenland.

A study of pollen grain concentration in surface snow and ice cores at 15 sites in the Canadian high Arctic and one site near the tree line, together with published pollen deposition rates south of the tree line has shown long-range dispersal of pollen from the boreal forest to the limits of our area on the Arctic Ocean close to Svalbard and the North Pole. There are no discernible trends of deposition rates within the high Arctic which suggests extremely long trajectories with strong zonal components; some of the pollen may have an Eurasian source. We relate the trajectories to synoptic patterns in the mid- and high Arctic.

A simple model, which is zonally averaged, for the transport of atmospheric water vapour is presented which uses as input the zonally averaged evaporation field and the mean meridional travel distance of tropospheric water vapour as functions of latitude. The model demonstrates that for polar regions each of the 10° latitude strips poleward of 25° is of equal importance as a moisture source. The model is used to predict zonal averages of δ(18O) for the present day and 18 ka BP. Both annual average values and seasonal amplitudes are presented and compared to observations. Sea-ice cover is an important factor in determining both annual averages and seasonal amplitudes today and at 18 ka BP. An earlier model (Jouzel and others 1983) linking δ(18O), the deuterium excess, and sea-salt content in an Antarctic ice core to the relative humidity of the source region is based on a single-source atmospheric water-vapour cycle type model and is re-evaluated using the present model.