I. Blackwood

The composition of snowfall, snowpack and meltwater in the Scottish Highlands ― evidence for preferential elution

Abstract Acidic snows in a small, remote, high-altitude snowpack in the Cairngorms, Scotland, give rise to meltwaters which are proportionally rich in sulphate and nitrate. As a consequence, the within-pack snows become proportionally rich in chloride, even though depleted in solute. Preferential elution appears to be a major process in the chemical evolution of snowfall and snowpack.

The Removal of Soluble Ions from Melting Snowpacks

Field and laboratory observations have indicated that the “fractionation” mechanism, whereby the bulk of the solute in seasonal snowcover is removed in the early meltwater fractions, is a differential process. Some soluble ions may be removed at faster rates than others. This process is known as “preferential elution”. Since the process is not yet universally accepted amongst scientists who engage in research into snowmelt chemistry, in this chapter we review the pertinent experimental work. Varying experimental conditions in real snowpacks often militate against reproducible elution sequences, although sodium and chloride are consistently found to be the least mobile ions upon melt. Laboratory experiments have provided more controlled conditions under which to examine the relative removal rates of soluble ions in meltwater. Those ions which are more soluble in ice are seen to be less mobile upon melt, whereas the ions which are eluted more rapidly may be regarded as being more strongly partitioned into the solute-rich surficial liquid, or quasi-liquid, layer. In real melting snowpacks, preferential elution of some ions by the meltwater can be masked by snowpack-scale hydrology.

Spatial variability in the chemical composition of snowcover in a small, remote, scottish catchment

Abstract Fresh snowcover in a remote upland catchment is chemically heterogeneous. Coefficients of variation (Cof) for individual major ions range from 5 to 113%, assuming a normal distribution, or 1 to 144 % assuming a lognormal distribution. Aged snowcover is slightly more chemically variable, giving rise to Cof of 19–140 %, assuming a normal distribution, or 2–200 %, assuming a lognormal distribution. In general, the distribution of the concentration of major ions in snowcover is better described by the lognormal distribution. Both vertical and horizontal variability is observed in the chemical content of within-pack snows. Such heterogeneity has important implications for snowpack sampling. We suggest that rigorous methodologies should be devised for the sampling of fresh snowcover. The sampling of within-pack snows should also be performed under tight statistical constraints if representative estimates of the solute content of snowcover are to be obtained.

GEOCHEMICAL STUDIES IN A REMOTE SCOTTISH UPLAND CATCHMENT II. STREAMWATER CHEMISTRY DURING SNOW-MELT

Abstract‘Acid-flush’ events, monitored in an upland catchment in the Cairngorm Mountains (Scotland) at the time of the spring-thaw, are associated with an increase in stream discharge and raised concentrations of both major ions (Ca, Mg, Na, Cl, N03, and SO,) and trace-elements (Al, Cd, Cu, Fe, Mn, and Pb), in addition to H+. The streamwater chemistry is determined by the hydrological pathways which are operative in the catchment during these periods of snowmelt, and reflects both the meltwater composition and the influence of the soils within the catchment. Aluminium, in particular, is leached from the soils and high concentrations (up to 330 μg L−1) occur in the streamwaters. The presence of frozen soils, which result largely due to the influence of meteorological conditions prior to the accumulation of the snowpack, is likely to have a large impact on the Al concentrations in the streamwaters. The low concentrations of Ca monitored in the stream during the periods of snow-melt (<0.2 mg L−1) may promote subsequent toxic effects of the Al to aquatic life forms.[/p]

Chemical Composition of Snow in the Remote Scottish Highlands

Altitudinal surveys suggest an increase in ionic concentration in snow with increasing fall-path, although such studies are complicated by the marked spatial variability in the chemical composition of fresh snow-cover. This variability is characterised and the implications for snowfall/snowcover sampling are addressed. Nitrate concentrations in snow are relatively enhanced compared to concentrations in rainfall. The NO3 -/SO4 2- ratios in snowfall appear to be dependent on back-trajectory type and associated source region. We consider the relevance of the observations for studies of the chemical evolution of air masses. The Cairngorm range of mountains experiences, not uncommonly, deeply-coloured snowfalls with a large black carbon content which appears to be directly related to snowfall acidity. The events can represent relatively large contributions to the total impurity deposition in the mountains. The synoptic conditions which lead to the black deposition events have been identified.

Trace-element studies in a remote scottish upland catchment

Fresh and aged within-pack snow samples were collected and analyzed for 6 trace-elements during the winters of 1984 and 1985 at a remote site located within the Cairngorm Mountains, Scotland. All of the fresh snow samples were acidic and highly variable in composition. The variability of pH and trace-element concentrations of snowfall are demonstrated to be associated with different air trajectory categories, and the study catchment is susceptible to episodic pollution events due to long range atmospheric transport. Partial melt experiments have shown that both fractionation and preferential elution of trace-elements occur during melting, the concentrations being 1.3 to 5.4 times higher in the first 10% meltwater fraction than in the bulk snow. Cadmium and Mn appear to be preferentially eluted from the melting snow with respect to the remaining elements. Upon snowpack melting, the trace-elements may be mobilized and redistributed within the snow profile. The ions concentrate at depth from where they can be quickly removed with the early water runoff during the spring.

The character and causes of a pronounced snowmelt-induced ‘acidic episode’ in a stream in a Scottish subarctic catchment

Specific objectives were: (1) to characterize the major ‘episode’ of the year in a catchment in a region which is regarded as being sensitive to pollutant deposition; (2) to examine downstream compositional gradients and to explain them in terms of downstream changes in flowpath regime and differences in soil depth or type; (3) to test the hypothesis that both direct pollution inputs to the stream and the sea-salt effect are reflected in streamwater composition. Streamwater was collected, during two distinct snowcover melt events at two points, 400 m apart, with high time resolution. Snowcover and soilwater samples were also collected. There were distinct differences in the response of the stream at the two sites to the melt events; this was a consequence of spatially variable snowmelt regimes as well as of differences in soil and sub-snowcover hydrology. There were a number of unusual features apparent. There was no evidence of the almost ubiquitous base cation dilution associated with acidic episodes. There was also a most pronounced direct effect of pollution input at the uppermost stream sampling site because of the accumulation of pollution on top of a deep snowcover, the melt from which followed rapid flowpaths to the stream. The direct effect was not apparent at the lower sampling site, although evidence for the sea-salt mechanism was very clear, unlike the case for the top site. The evidence for an ion exchange control on streamwater Al3+ was strong at the lower site, but weak at the top site. The observations are interpreted via a schematic flowpath model.

CHANGES IN STREAMWATER CHEMISTRY DURING SNOWMELT

The ionic composition of precipitation in a small, remote Scottish catchment can be treated as a mixture of sea salt and acidic components. Snow is proportionally enriched in nitrate with respect to rain, which was sampled at five locations in open bulk collectors. In contrast to the variable composition of the major ion matrix of precipitation, streamwater draining the catchment shows only limited variations in the proportional composition of cations (Na+, Mg2+, H+) and anions (Cl-, SO2- 4, NO- 3). This is attributed to buffering by chemical reactions within the soil. The proportional anionic composition of streamwaters is more variable than the proportional cationic composition. The composition of streamwater is most influenced by meltwater at the onset of snowmelt, when the proportion of nitrate is highest. The composition of streamwater seems to be inrluenced by the distribution of solute within the snowpack and the flow routing through the pack. The effects of the input of concentrated meltwater to the stream can be readily observed, but the differential removal (preferential elution) of ions from the snowpack does not appear to be reflected in the composition of the streamwater.

The effect of polluted and leached snow melt waters on the soil bacterial community-quantitative response.

The effect of polluted snow melt waters on the number of soil bacteria was determined using soil cores extracted from an upland catchment in the Cairngorm Mountains, Scotland. Total numbers of viable heterotrophic bacteria and bacterial denitrifiers were determined using plate and MPN counts. Separate soil cores were treated with simulated melt waters representative of either the composition of the first melt fraction from polluted or leached snowpacks. The number of bacteria in the Ah soil horizon (Hodgson, 1974) treated with polluted snow melt (PSM) water decreased significantly by 28-fold, but increased by 11-fold in the BC horizon. Denitrifier numbers decreased by 8-fold in the Ah horizon, but increased by over 2-fold lower down the profile. Overall the bacterial community exposed to simulated leached snow melt (LSM) waters showed little change in the Ah horizon. In the BC horizon (Hodgson, 1974), total viable bacterial numbers decreased by 20-fold, but denitrifiers numbers were unaffected.