Streambank Salt Plume Intrusion, Concentration, and Residence Times Along a Dynamic, Urban Headwater Stream: Investigation of Development and Remediation Mechanisms

Previous research in the Boone Creek watershed, Boone, North Carolina, USA, has documented road-salt-induced groundwater contamination through bank storage processes. Long-term monitoring of Hardin Creek, a tributary of Boone Creek, has shown similar rapid stage changes, acute (1-hour average) spikes in chloride content over 2500 mg/L, and chronic (4-day average) chloride concentrations over 300 mg/L. This project uses new stage and conductivity data to investigate the relationship between runoff event salinity, duration, volume, and frequency to the salinity, size, and persistence of the stored salt plume along Hardin Creek.

We collected three years of 15-minute resolution water level and conductivity data from both the stream and the riparian aquifer. Grain-size distribution analyses were used to determine best estimates for aquifer hydraulic properties. A cross-sectional survey of the stream was conducted to determine channel shape. These data were used to construct and calibrate a base case cross-sectional model of Hardin Creek using FEFLOW, the finite element groundwater flow and solute transport modeling software. The model was then run with synthetic boundary conditions simulating scenarios in which 1) runoff chloride concentrations were reduced by half and 2) stage fluctuations were reduced by half. All scenarios were run five years to allow salt plume development, and then an additional 25 years with continued stage fluctuation but no additional salt to determine residence time of the salt plume in the aquifer.

The base case scenario produced a salt plume 25% larger than both the reduced salt and reduced flashiness scenarios. Peak salt concentration of the base case was twice that of the reduced salt scenario, but was matched by the reduced flashiness scenario. However, peak concentration diminished more rapidly under reduced flashiness than under base conditions during the salt-free recovery period. This indicates that reducing stream flashiness through runoff control, an approach perhaps more feasible in cold climates than dramatic salt use reduction, could be useful in reducing residence time of salt in the banks of Hardin Creek and similar urban headwater streams. Future modelling incorporating channel morphology could extend these findings to other streams.