ELUCIDATING FLOW CHARACTERISTICS OF EPIKARST SPRINGS USING LONG-TERM RECORDS THAT ENCOMPASS EXTREME HYDROGEOLOGIC STRESSES

This study describes the results of a series of spring-discharge measurements that were conducted at the Savoy Experimental Watershed (SEW), and it offers explanations for the observed variations in springflow for varying hydrologic, temporal, and extreme stress conditions. A significant component of the study was conducted as part of the Research Experience for Undergraduates (REU) program during the summer and late fall of 2005 and the summer of 2006, and funding by the National Science Foundation (NSF) is gratefully acknowledged. This specific study is part of a longer-term assessment of hydrologic budget components at SEW, a long-term, well-instrumented karst research site in the Arkansas Ozarks.

Discharge measurements for Pond, Tree, Woodpecker, Memory, Dribble, and Red Dawg Springs were conducted twice a day, during the early morning (7:00 am during July and August; 8:00 am during December because of late sunrise), and the early afternoon, at about 2:00 pm. Data collection periods were chosen reflecting the growing season (21 days in July and August, 2005 and 2006) and the dormant season (5 days in December 2005). Volumetric measurements were made with 500 to 2000 mL graduated glass cylinders, accurate to +/- 2 mL; time measurements were made with a stopwatch readable to 0.01 seconds. The mean of 10 repetitions per spring (volume/time) for each measurement provided an accurate, reproducible assessment of discharge.

Results indicate that 1) diurnal fluctuation of these epikarst springs is dominant during the growing season and minimal during the dormant season when deciduous trees have lost their leaves and transpiration is minimal. This loss of water from the shallow ground-water system is interpreted to be evapotranspiration from the ground-water system, and ranges from 5 to 25 mL/s over the course of a diurnal cycle for each spring; 2) extreme low-flow conditions are manifest differently in different spring basins, and these provide an understanding of the flow mechanisms that may be active in karst settings; 3) the wide range of hydrogeologic response in epikarst springs to identical stresses indicates that our models of these systems are likely grossly oversimplified.