Results from cryogenic column experiments are compared with the geochemical data collected in the Canadian and Fennoscandian Shields over the past 25 years to investigate the relative influence of the glacial-interglacial cycle; specifically, the impact of continental glaciers, permafrost, and methane hydrate, on the evolution of groundwater from crystalline shield environments. Several different geochemical indicators of freezing processes (either glacial or permafrost-related) were utilized: comparisons of Na/Cl and Br/Cl ratios, delta O-18 and delta H-2 values, and delta O-18 values and Cl- concentration. During freezing, fluids with different dominant cations follow distinctly different linear trends when Na/Cl and Br/Cl ratios are compared. Significantly, none of the freezing trends follows the trend hypothesized by Herut et al. (1990) for the evolution of seawater chemistry during freezing. Intrusion of glacial meltwater and in situ freezing (i.e., permafrost formation) result in a similar end-member when comparing delta O-18 values and Cl- concentration. The geochemical influence of a freezing process on fresh, brackish, and some saline fluids was identified at some, but not all Canadian Shield sites, regardless of site location with respect to modern-day permafrost. Appreciably, physical and geochemical data do not support the formation of brines through any freezing process in the Canadian and Fennoscandian Shields, as hypothesized by Starinsky and Katz (2003). Rather, on all diagnostic freezing plots, brines are an end-member, indicating a different evolutionary pathway. Significant depletions in O-18 with respect to modern precipitation, an indication of either glacial meltwater or a freezing process, were identified at depths of up to 1 km at some sites in the Canadian Shield, and to shallower depths in the Fennoscandian Shield. The potential of this fluid to reach such depths could be attributable to artificial gradients and mixing, glacial recharge, permafrost or paleo-permafrost formation, or methane hydrate or paleo-methane hydrate formation. At most locations it was not possible to distinguish between the different scenarios using the current geochemical database. (C) 2011 Elsevier Ltd. All rights reserved.
