CHANGES IN CHLORIDE CONCENTRATIONS THROUGH TIME

Continuously measured specific conductivity can be a surrogate for chloride concentrations. A linear regression model was used to estimate the relation between the continuously measured specific conductance and chloride concentration in five monitoring wells in McHenry County, Illinois. The wells were selected based on their elevated concentrations of chloride (107 to 456 milligrams per liter (mg/L)). The range in chloride concentrations throughout the period of record for three out of four wells was near or above the U.S. Environmental Protection Agency (USEPA) aquatic life chronic toxicity criteria (230 mg/L) and the USEPA secondary maximum contaminant level (SMCL) drinking water standard (250 mg/L). Four of the wells are adjacent to roads or in urbanized areas, and one is a background well. This type of data can enable water managers and the general public to track chloride concentration in groundwater on a continuous real-time basis and can be used to better understand temporal (daily, seasonal, multi-year) variations in the concentration of chloride entering surface water as baseflow, identify sources of chloride, and assess the effectiveness of current best management practices (BMPs). Continuous water-quality data shows chloride concentrations vary in response to seasons, precipitation, and geology. Chloride concentrations were greater during winter and spring seasons and declined during the summer seasons. Following a wet winter, spring melt and precipitation caused an increase in chloride concentrations. However, during a drought period, the chloride concentrations at some wells increased. Following a drought period, spring rains occurred and a decrease in chloride concentrations was observed. Chloride concentrations rebounded to previous levels shortly after spring rains had ceased. The continuous data reveals how these variables might affect concentrations obtained from discrete samples and could help to explain the variability in discrete chloride concentration data. This information can be used to design a more effective sampling regimen.