Shallow groundwater flow in the critical zone of steep headwater mountain catchments is often assumed to mimic surface topography. However, groundwater flow is influenced by other variables, such as the elevation of the water table and subsurface hydraulic conductivity, which can result in temporal variations in both magnitude and direction of flow. In this study, I investigated the temporal variability of groundwater flow in the soil zone (solum) within the critical zone of a headwater catchment at the Hubbard Brook Experimental Forest in North Woodstock, NH. Groundwater levels were continuously monitored throughout several seasons (March 2019 to Jan 2020) in a network of wells comprising three hillslope transects within the upper hillslopes of the catchment. Five clusters of three wells per cluster were screened from 0.18 – 1.1 m depth at the base of the solum. Water levels were also monitored in five deeper wells, screened from 2.4 - 6.9 m depth within glacial sediments of the C horizon. I conducted 47 slug tests across the well network to determine hydraulic properties of the aquifer materials surrounding each well. In addition, our team conducted a large-scale auger investigation mapping soil horizon depths and thicknesses.

Results show that the magnitude of hydraulic gradients and subsurface hydrologic fluxes varied at each site with respect to changing water-table elevation, having a maximum range of 0.12 m/m and 9.19 x 10-6 m/s, respectively. The direction of groundwater flow had an arithmetic mean deviating from surface topography by 2-10 degrees, and a total range that deviated from surface topography by as much as 51 degrees. During lower water table regimes, groundwater flow direction deviated from the ground surface, but under higher water table regimes, in response to recharge events, flow direction mimicked surface topography. Within most of the well clusters, there is an observable connection between the slope direction of the top of the C horizon and the direction of groundwater flow during lower water table regimes. Slug test results show the interquartile range of saturated hydraulic conductivity (Ksat) within the C horizon (1.5×10-7 to 9.8×10-7 m/s) is two orders of magnitude lower than the interquartile range of Ksat values within the solum (2.9×10-5 to 5.2×10-5 m/s). Thus, the C horizon is on average a confining unit relative to the solum that may constrict groundwater flow below the solum. Additionally, results from the larger scale auger investigation suggest that deviations in subsurface topography of the C horizon may be generalizable at the larger hillslope scale. Overall, these results indicate that 1) shallow groundwater flow direction and magnitude within this headwater catchment are dynamic and can deviate from surface topography, and 2) the subsurface topography of the C horizon can influence groundwater flow direction. These results imply that temporal dynamics of groundwater flow direction and magnitude should be considered when characterizing subsurface flow in critical zone studies. Additionally, knowledge of subsurface topography of confining units may provide constraints on the temporal variability of groundwater flow direction.