Exploration for lead deposits has occurred in a mature karst area of southeast Missouri that is highly valued for its scenic beauty and recreational opportunities. The area contains the two largest springs in Missouri (Big Spring and Greer Spring), both of which flow into federally designated scenic rivers. Concerns about potential mining effects on the area ground water and aquatic biota prompted an investigation of Big Spring.

Water-level measurements made during 2000 helped define the recharge area of Big Spring, Greer Spring, Mammoth Spring, and Boze Mill Spring. The data infer two distinct potentiometric surfaces. The shallow potentiometric surface, where the depth-to-water is less than about 250 feet, tends to mimic topographic features and is strongly controlled by streams. The deep potentiometric surface, where the depth-to-water is greater than about 250 feet represents ground-water hydraulic heads within the more mature karst areas. A highly permeable zone extends about 20 mile west of Big Spring toward the upper Hurricane Creek Basin. Deeper flowing water in the Big Spring recharge area is directed toward this permeable zone. The estimated sizes of the spring recharge areas are 426 square miles for Big Spring, 352 square miles for Greer Spring, 290 square miles for Mammoth Spring, and 54 square miles for Boze Mill Spring.

A discharge accumulation curve using Big Spring daily mean discharge data shows no substantial change in the discharge pattern of Big Spring during the period of record (water years 1922 through 2004). The extended periods when the spring flow deviated from the trend line can be attributed to prolonged departures from normal precipitation. The maximum possible instantaneous flow from Big Spring has not been adequately defined because of backwater effects from the Current River during high-flow conditions. Physical constraints within the spring conduit system may restrict its maximum flow. The largest discharge measured at Big Spring during the period of record (water years 1922 through 2004) was 1,170 cubic feet per second on December 7, 1982.

The daily mean water temperature of Big Spring was monitored during water years 2001 through 2004 and showed little variability, ranging from 13 to 15° C (degree Celsius). Water temperatures generally vary less than 1° C throughout the year. The warmest temperatures occur during October and November and decrease until April, indicating Big Spring water temperature does show a slight seasonal variation.

The use of the traditional hydrograph separation program HYSEP to determine the base flow and quick flow or runoff components at Big Spring failed to yield base-flow and quick-flow discharge curves that matched observations of spring characteristics. Big Spring discharge data were used in combination with specific conductance data to develop an improved hydrograph separation method for the spring. The estimated annual mean quick flow ranged from 15 to 48 cubic feet per second for the HYSEP analysis and ranged from 26 to 154 cubic feet per second for the discharge and specific conductance method for water years 2001 to 2004.

Using the discharge and specific conductance method, the estimated base-flow component rises abruptly as the spring hydrograph rises, attains a peak value on the same day as the discharge peak, and then declines abruptly from its peak value. Several days later, base flow begins to increase again at an approximately linear trend, coinciding with the time at which the percentage of quick flow has reached a maximum after each recharge-induced discharge peak. The interval between the discharge peak and the peak in percentage quick flow ranges from 8 to 11 days for seven hydrograph peaks, consistent with quick-flow traveltime estimates by dye-trace tests from the mature karst Hurricane Creek Basin in the central part of the recharge area.

Concentrations of environmental tracers chlorofluorocarbons (CFCs: CFC-11, CFC-12, CFC-113), and sulfur hexafluoride in discharge from Big Spring vary approximately linearly with percent quick flow from about 5 to 45 percent of discharge. Linear extrapolation to 100 percent quick flow implies CFC and SF₆ concentrations nearly identical to those in the 2002 atmosphere and indicates a modern age for the quick-flow component. Tracer concentrations for less than about 5 percent quick flow are increasingly lower than those expected from linear extrapolation to zero percent quick flow, indicating that the reservoir of older water in the Big Spring watershed may be a series of water mixtures with piston-flow ages greater than those obtained by extrapolation to zero percent quick flow. Each sample point with a low percentage of quick flow (less than 5 percent) may be a unique mixture.

Environmental tracer data from Big Spring plot intermediate to the simple binary mixing of modern and old, pre-tracer water and results from the exponential mixture model. The mean ages of waters in the base-flow component approximately range from 30 to 200 years. The mean age of the base-flow component is youngest (30 to 40 years) in samples containing the highest quick-flow component (45 percent quick flow) and increases to 200 years or more as the fraction of quick flow decreases to less than 5 percent. Tritium data are consistent with a model of dilution of a modern component with an old, pre-tracer component and indicates that the old fraction is mostly pre-1960s in age with mean residence time of more than several hundred years. All of the samples from Big Spring and Greer Spring have water temperatures warmer than their nitrogen-argon recharge temperature, which range from approximately 10.5 to 14° C, suggesting recharge to the Big Spring watershed occurs primarily in late winter to early spring. The water temperatures at Big Spring are consistent with relatively shallow circulation (less than about 600 feet), and the water does not appear to be warmed by deep circulation along a geothermal gradient.

Specific conductance values and concentrations of most inorganic constituents in water samples from Big Spring generally decrease with increasing discharge because of dilution with quick-flow water of lower ionic strength. Concentrations of some constituents such as chloride and nitrite plus nitrate, and fecal coliform densities, however, did not decrease with increasing discharge, indicating that quick flow probably is a more important source of these constituents compared to base flow. Water samples from Big Spring plot along the line of dolomite dissolution by carbonic acid, are at equilibrium with dolomite and calcite, and have a molar ratio of Ca:Mg of near 1, indicating dissolution of the mineral dolomite as the primary control on concentrations of calcium, magnesium, and bicarbonate. The flux of calcium and magnesium from Big Spring represents the dissolution of about 1,950 cubic feet of dolomite per day. The suspended sediment load of Big Spring was estimated to range from about 1 to about 70 tons per day, and the sediment load during base-flow periods ranged from about 1 to about 7 tons per day.