Abstract:Once meltwater reaches the base of a snowpack it can infiltrate the underlying stratum, runoff, or refreeze and form a basal ice layer. Basal ice formation is most common early in melt over saturated or very cold frozen soils. Initial meltwater becomes enriched in ion concentrations compared to the parent snow due to ion fractionation during thaw and percolation through the snowpack. If ion exclusion occurs during basal ice formation, further enrichment of initial runoff water ion concentrations might occur. The influence of basal ice formation on runoff water chemistry was examined by comparing ion concentrations in runoff water that had sustained basal ice contact, to meltwater before basal ice contact. A series of experiments, involving melting a snowpack in a large insulated box over a cold impermeable substrate in a temperature-controlled room, were carried out. A cooling system at the chemically inert base ensured formation of basal ice during snowmelt. Meltwater samples were collected throughout melt from within the snowpack using an extraction tube; runoff water was collected at the base. All samples were analysed for major anions and cations. Results showed that formation of basal ice layers can sometimes enrich the initial runoff water compared to meltwater before basal ice contact. Ion concentrations in basal ice contact runoff water were up to sixteen times greater than those in no-contact meltwater; however, on average, basal ice contact runoff water showed 1Ð5 times the ion concentrations of the no-contact meltwater. Enrichment was greatest with the rapid formation of a thick basal ice layer. The implications are that basal ice formation alters both meltwater ion pathway and concentration. When no basal ice is present, enhanced infiltration of meltwater ion load can cause relatively dilute runoff water. When basal ice is present all meltwater runs off and further ion-concentration enrichment occurs.

This article reviews several aspects of snow hydrochemistry: the chemistry of snowfall including chemical incorporation in snowfall and snowfall chemistry variability, the chemistry of cold, dry snowcovers including snow redistribution, snow–atmosphere chemical exchange and in‐pack chemical transformations, the chemistry of wet and melting snowcovers including solute leaching, particulate interactions and microbial activity, and snow‐covered basin hydrochemistry with an emphasis on nutrient chemistry. The emphasis is on the processes of chemical transformation in seasonal snowpacks and meltwaters with strong attention to the broad ecosystem view of snow chemistry rather than solely focusing on acidification effects from snowmelt. The seasonal snowcover is shown to be a dynamic hydrochemical system with strong ecological interactions. Besides wet deposition by snowfall and rain, the processes of wind redistribution, dry deposition, volatilization, crystal metamorphism, photolysis, microbial uptake and release, solute elution, and meltwater movement strongly affect the chemistry of both the snowpack and meltwaters. Snowmelt chemistry alone is rarely directly responsible for major chemical fluctuations in water bodies, but meltwater has an important role in transporting ions from soils and organic material to water bodies.

Abstract:A model is proposed in which the cumulative load of an ion infiltrating into frozen unsaturated soil can be estimated as a function of meltwater ion concentration and infiltration rate. Assumptions of the model are that the meltwater solution released to the soil surface is conservative, fully mixed within each time step, and that mass and energy are conserved. Infiltration and meltwater concentration are estimated using relationships developed by Gray and Stein respectively. The model suggests that the relationship between ion concentration and volume of infiltration is non-linear with a positive covariance. Infiltration of snowmelt ions is therefore a function of the products of the mean concentration in the meltwater and the cumulative volume of water that infiltrates, plus the covariance between instantaneous values of ion concentration and infiltration rate. This covariance effect is termed enhanced infiltration. Meteorological observations and soil parameters from four sites in western Canada were used to assess the sensitivity of the model to conditions at a prairie site, a boreal forest site, a mountain forest site, and a shrub tundra site. Model results showed the greatest cumulative infiltration of ion load for the Prairie site; the general ranking was Prairie > Mountain Forest > Boreal Forest > Tundra. However, the greatest impact of enhanced infiltration was found for the Tundra site. At this site, enhanced infiltration caused up to 50% more ion load to infiltrate within the initial third of the melt period compared to infiltration estimates not accounting for this effect. Over the whole melt period, enhanced infiltration caused 55-160% more ion load to infiltrate than estimates based solely on the mean depth of infiltration and ion concentration. Sensitivity analysis showed that enhanced infiltration varies most strongly with initial snow water equivalent, average melt rate over the whole melt period, and snowpack ion elution concentration factor (CF).

Abstract. Meltwater ion concentration and infiltration rate into frozen soil both decline rapidly as snowmelt progresses. Their temporal association is highly non-linear and a covariance term must be added in order to use time-averaged values of snowmelt ion concentration and infiltration rate to calculate chemical infiltration. The covariance is labelled enhanced ion infiltration and represents the additional ion load that infiltrates due to the timing of high meltwater concentration and infiltration rate. Previous assessment of the impact of enhanced ion infiltration has been theoretical; thus, experiments were carried out to examine whether enhanced infiltration can be recognized in controlled laboratory settings and to what extent its magnitude varies with soil moisture. Three experiments were carried out: dry soil conditions, unsaturated soil conditions, and saturated soil conditions. Chloride solutions were added to the surface of frozen soil columns; the concentration decreased exponentially over time to simulate snow meltwater. Infiltration excess water was collected and its chloride concentration and volume determined. Ion load infiltrating the frozen soil was specified by mass conservation. Results showed that infiltrating ion load increased with decreasing soil moisture as expected; however, the impact of enhanced ion infiltration increased considerably with increasing soil moisture. Enhanced infiltration caused 2.5 times more ion load to infiltrate during saturated conditions than that estimated using time-averaged ion concentrations and infiltration rates alone. For unsaturated conditions, enhanced ion infiltration was reduced to 1.45 and for dry soils Correspondence to: G. Lilbaek (gro.lilbaek@ualberta.ca) to 1.3. Reduction in infiltration excess ion load due to enhanced infiltration increased slightly (2-5%) over time, being greatest for the dry soil (45%) and least for the saturated soil (6%). The importance of timing between high ion concentrations and high infiltration rates was best illustrated in the unsaturated experiment, which showed large inter-column variation in enhanced ion infiltration due to variation in this temporal covariance.