Variability in the Chemical Composition of in Situ Subglacial Meltwaters

Meltwaters collected from boreholes drilled to the base of the Haut Glacier d'Arolla, Switzerland have chemical compositions that can be classified into three main groups. The first group is dilute, whereas the second group is similar to, though generally less concentrated in major ions, than contemporaneous bulk glacial runoff. The third group is more concentrated than any observed bulk runoff, including periods of flow recession. Waters of the first group are believed to represent supraglacial meltwater and ice melted during drilling. Limited solutes may be derived from interactions with debris in the borehole. The spatial pattern of borehole water levels and borehole water column stratification, combined with the chemical composition of the different groups, suggest that the second group represent samples of subglacial waters that exchange with channel water on a diurnal basis, and that the third group represent samples of water draining through a ‘distributed’ subglacial hydraulic system. High NO−3 concentrations in the third group suggest that snowmelt may provide a significant proportion of the waters and that the residence time of the waters at the bed in this particular section of the distributed system is of the order of a few months. The high NO−3 concentrations also suggest that some snowmelt is routed along different subglacial flowpaths to those used by icemelt. The average SO2−4: (HCO−3 + SO2−4) ratio of the third group of meltwaters is 0.3, suggesting that sulphide oxidation and carbonate dissolution (which gives rise to a ratio of 0.5) cannot provide all the HCO−3 to solution. Hence, carbonate hydrolysis may be occurring before sulphide oxidation, or there may be subglacial sources of CO2, perhaps arising from microbial oxidation of organic C in bedrock, air bubbles in glacier ice or pockets of air trapped in subglacial cavities. The channel marginal zone is identified as an area that may influence the composition of bulk meltwater during periods of recession flow and low diurnal discharge regimes. © 1997 by John Wiley & Sons, Ltd.