An Examination of Short‐Term Variations in Water Quality at a Karst Spring in Kentucky

Abstract

AbstractWater quality at many karst springs undergoes very high amplitude but relatively brief degradation following influxes of runoff. Accurately recording transient variations requires more rigorous sampling strategies than traditional methods. A pilot study to determine the usefulness of high‐frequency, flow‐dependent sampling strategies, combined with coincidental quantitative dye tracer tests, was implemented in the Big Spring Ground‐Water Basin in Mammoth Cave National Park, Kentucky.Data recorded following two separate runoff events showed that the concentrations of two nonpoint source pollutants, fecal coliform bacteria and suspended sediment, greatly exceeded prerunoff event values for very short periods of time. A phreatic conduit segment, calculated at 17 million liters in volume, instantaneously propagated head changes, caused by direct runoff entering the aquifer, from the ground‐water inputs to Big Spring. A significant delay between the initial increases in discharge and the arrival of direct runoff, as indicated by a steady decrease in specific conductance, represented the time required to displace this volume of phreatic water. The delay showed that sampling a karst spring only during peak discharge would be an unreliable sampling method.Runoff from two different subcatchments was tagged with tracer dye and the timing of the passage of the resultant dye clouds through Big Spring were compared to water quality variations. Distinct lag times between the arrival of direct runoff at Big Spring and the bacteria and suspended sediment waveforms were shown through the concurrent quantitative tracer tests to be related to the areal distribution of land‐cover type within the basin.