RESERVOIR-STAGE MANIPULATION AND GROUND-WATER FLOW IN THE LEDBETTER CREEK WATERSHED, CALLOWAY COUNTY, KENTUCKY

Ground-water discharge to streams and lakes has received increased attention during the past decade, in part because of ground water's role in sustaining aquatic ecosystems and because of the potential sensitivity of those ecosystems to ground-water contaminants. However, ground-water/reservoir interactions have received less attention, even though reservoirs have become the predominant aquatic feature of much of the US landscape. We are conducting a multi-year study of ground-water flow in the rural Ledbetter Creek watershed, which drains 24 km² of the northern Gulf Coastal Plain. Ledbetter Creek is a third-order tributary of the Tennessee River, which was dammed to form Kentucky Lake. The reservoir is raised ~ 1.5 m in the spring and lowered again in the fall.

Transects of piezometers were installed in the flood plain and embayment of Ledbetter Creek. Hydraulic heads and temperatures were continuously logged from December 1999 to January 2001 in the flood plain and from January 2000 to March 2002 in the embayment. TVA provided continuous reservoir-stage data. Continuous data were checked against monthly to quarterly manual measurements. Ledbetter Creek flow, flow from a spring adjoining the embayment, and hydraulic heads in domestic wells and a monitoring well were also measured monthly to quarterly. Slug tests were run for selected piezometers.

Ground water discharges both to Ledbetter Creek and to the embayment; discharge tends to be diffuse except at the monitored spring. Hydraulic conductivities are consistent with lithologies of fine sand to clay. The influence of Kentucky Lake on ground-water flow diminishes with distance from the embayment. Heads in embayment piezometers tracked reservoir stage at summer pool, whereas heads in flood-plain piezometers were more influenced by the creek. Heads in domestic wells tracked creek and spring flow, which decreased through the year from spring to winter. Reservoir-stage fluctuations seemed to have little effect on ground-water flow at watershed elevations > 0.3 m above summer pool. Steady-state simulations using MODFLOW best fit observed hydraulic heads when recharge was set to half of the annual precipitation. This result suggests that our initial assumption of no inter-basin flow is incorrect and that the flood plain and embayment are a regional discharge zone.