GSA 2020 Connects Online

Paper No. 213-7
Presentation Time: 3:25 PM

STABLE ISOTOPIC DELINEATION OF RECHARGE IN AN URBAN KARST BASIN, CENTRAL KENTUCKY (USA)


CURRENS, Benjamin J., ALVAREZ VILLA, Cristopher and FRYAR, Alan E., Earth and Environmental Sciences, University of Kentucky, 101 Slone Research Building, Lexington, KY 40506-0053

In karst basins, groundwater recharge can be either diffuse or focused through sinkholes and swallets (streambed sinks), which integrate surface and subsurface drainage. Delineating mechanisms of recharge is complicated in urban karst basins because of drainage modification by development, including impervious cover, channelized streams, storm and sanitary sewers, water mains, and utility trenches. We monitored deuterium and oxygen-18 from February to December 2018, generally weekly to biweekly, at four sites in the Wolf Run basin in Lexington, Kentucky. The sites included Wolf Run (WR), which has a USGS stream gauge near its outlet; an ephemeral tributary (Vaughn’s Branch, VB); a perennial spring (McConnell Springs, MC); and a sinkhole (Campbell House, CH), which ponds after storms and has been identified by dye tracing as a recharge point for the spring. We continuously logged electrical conductivity (EC) at the spring (January–October 2018) and the sinkhole (May–October 2018). Hydrograph separation was performed using digital filtering on daily discharge data. Days with and without direct runoff were classified as stormflow and baseflow, respectively, and were cross-checked against precipitation data from a nearby weather station. Statistical analyses indicated no significant differences among sites in mean or median δ2H or δ18O for stormflow or baseflow conditions. Linear regressions for each site yielded trends subparallel to the global meteoric water line, suggesting some evaporative enrichment. Excluding a depleted outlier for VB, slopes for the streams were similar (WR = 6.05, VB = 6.18); the slope was lower for the spring (5.58) and greater for the sinkhole (7.27). EC responded to storms at MC and CH; ponding at CH for 35 events lasted from 2 to 213 hr with a median of 21.5 hr. The difference in slopes for MC and CH suggests that the spring is sustained both by diffuse, more evaporated infiltration and focused recharge, whereas sinkhole drainage limits evaporative enrichment of water isotopes.