Spatial and temporal variability of streambed hydraulic conductivity in West Bear Creek, North Carolina, USA

SummaryThe hydraulic conductivity (K) of the streambed is an important variable influencing water and solute exchange between streams and surrounding groundwater systems. However, there are few detailed data on spatial variability in streambed K and almost none on temporal variability. The spatial and temporal variability of streambed K in a North Carolina stream were investigated with 487 field measurements of K over a 1-year period. Measurements were made bimonthly from December 2005 to December 2006 at 46 measurement locations in a 262.5 m reach (the “large reach”). To give a more detailed picture of spatial variability, closely-spaced one-time measurements were made in two 62.5 m reaches (the “small reaches”, one investigated in July 2006 and the other in August 2006) that were part of the large reach. Arithmetic mean K for the large reach was ∼16 m/day (range was 0.01 to 66 m/day). Neither K nor lnK was normally distributed, and K distributions appeared somewhat bimodal. There was significant spatial variability over horizontal length scales of a few m. Perhaps the clearest feature within this variability was the generally higher K in the center of the channel. This feature may be an important control on water and chemical fluxes through the streambed (e.g., other measurements show generally higher water seepage velocity, but lower porewater nitrate concentration, in the center of the streambed). Grain size analysis of streambed cores showed that layers of elevated fines (silt + clay) content were less common in the center of the channel (overall, the streambed was about 94% sand). Results also suggest a modest but discernable difference in average streambed K upstream and downstream of a small beaver dam: K was about 23% lower upstream, when the dam was present during the first few months of the study. This upstream/downstream difference in K disappeared after the dam collapsed, perhaps in response to re-mobilization of fine sediments or leaf matter that had accumulated in quiet waters ponded on the upstream side of the dam. Temporal variability was significant and followed a variety of different patterns at the 46 measurement locations in the large reach. Temperature data show that variation in streambed and groundwater temperature was not an important cause of the observed temporal variability in K. Measurements of changes in the elevation of the streambed surface suggest erosion and deposition played an important role in causing the observed temporal variability in streambed K (of which the change described above following collapse of the beaver dam was a special case), though other potentially time-varying factors (e.g., gas content, bioturbation, or biofilms in the streambed) were not explicitly addressed and cannot be ruled out as contributors to the temporal variability in streambed K. Temporal variability in streambed K merits additional study as a potentially important control on temporal variability in the magnitudes and spatial patterns of water and solute fluxes between groundwater and surface water. From the data available it seems appropriate to view streambed K as a dynamic attribute, variable in both space and time.