The applicability of the Gravity Recovery and Climate Experiment (GRACE) to adequately represent broad-scale patterns of groundwater storage (GWS) variations and observed trends in groundwater-monitoring well levels (GWWL) is examined in the Canadian province of Alberta. GWS variations are derived over Alberta for the period 2002–2014 using the Release 05 (RL05) monthly GRACE gravity models and the Global Land Data Assimilation System (GLDAS) land-surface models. Twelve mean monthly GWS variation maps are generated from the 139 monthly GWS variation grids to characterize the annual GWS variation pattern. These maps show that, overall, GWS increases from February to June, and decreases from July to October, and slightly increases from November to December. For 2002–2014, the GWS showed a positive trend which increases from west to east with a mean value of 12 mm/year over the province. The resulting GWS variations are validated using GWWLs in the province. For the purpose of validation, a GRACE total water storage (TWS)-based correlation criterion is introduced to identify groundwater wells which adequately represent the regional GWS variations. GWWLs at 36 wells were found to correlate with both the GRACE TWS and GWS variations. A factor f is defined to up-scale the GWWL variations at the identified wells to the GRACE-scale GWS variations. It is concluded that the GWS variations can be mapped by GRACE and the GLDAS models in some situations, thus demonstrating the conditions where GWS variations can be detected by GRACE in Alberta.

Groundwater recharge sets a constraint on aquifer water balance in the context of water management. Historical data on groundwater and other relevant hydrological processes can be used to understand the effects of climatic variability on recharge, but such data sets are rare. The climate of the Canadian prairies is characterized by large inter-annual and inter-decadal variability in precipitation, which provides opportunities to examine the response of groundwater recharge to changes in meteorological conditions. A decadal study was conducted in a small (250 km²) prairie watershed in Alberta, Canada. Relative magnitude of annual recharge, indicated by water-level rise, was significantly correlated with a combination of growing-season precipitation and snowmelt runoff, which drives depression-focussed infiltration of meltwater. Annual precipitation was greater than vapour flux at an experimental site in some years and smaller in other years. On average precipitation minus vapour flux was 10 mm y⁻¹, which was comparable to the magnitude of watershed-scale groundwater recharge estimated from creek baseflow. Average baseflow showed a distinct shift from a low value (4 mm y⁻¹) in 1982–1995 to a high value (15 mm y⁻¹) in 2003–2013, indicating the sensitivity of groundwater recharge to a decadal-scale variability of meteorological conditions.

Variability in baseline groundwater methane concentrations and isotopic compositions was assessed while comparing free and dissolved gas sampling approaches for a groundwater monitoring well in Alberta (Canada) over an 8-year period. Methane concentrations in dissolved gas samples (n = 12) were on average 4,380 ± 2,452 μg/L, yielding a coefficient of variation (CV) >50 %. Methane concentrations in free gas samples (n = 12) were on average 228,756 ± 62,498 ppm by volume, yielding a CV of 27 %. Quantification of combined sampling, sample handling and analytical uncertainties was assessed via triplicate sampling (CV of 19 % and 12 % for free gas and dissolved gas methane concentrations, respectively). Free and dissolved gas samples yielded comparable methane concentration patterns and there was evidence that sampling operations and pumping rates had a marked influence on the obtained methane concentrations in free gas. δ¹³CCH₄ and δ²HCH₄ values of methane were essentially constant (−78.6 ± 1.3 and −300 ± 3 ‰, respectively) throughout the observation period, suggesting that methane was derived from the same biogenic source irrespective of methane concentration variations. The isotopic composition of methane constitutes a robust and highly valuable baseline parameter and increasing δ¹³CCH₄ and δ²HCH₄ values during repeat sampling may indicate influx of thermogenic methane. Careful sampling and analytical procedures with identical and repeatable approaches are required in baseline-monitoring programs to generate methane concentration and isotope data for groundwater that can be reliably compared to repeat measurements once potential impact from oil and gas development, for example, may occur.