A better understanding of the ground-water and surface-water resources of the Rock River Valley in southwestern Minnesota was needed due to concerns surrounding future reliable sources of water for public supply. The Rock River Valley aquifer consists of a surficial sand and gravel unit that underlies the entire Rock River Valley and a buried sand and gravel unit that is present only in the vicinity of the Luverne Municipal and Airport well fields. The surficial and buried units of the aquifer are separated by a clay and till layer ranging in thickness from 1 to 38 feet. The combined maximum saturated thickness of the aquifer is 52 feet, with a median of 22 feet. The thickness of the buried unit ranges from 3 to 17 feet. Recharge to the Rock River Valley aquifer occurs primarily by infiltration of precipitation to the saturated zone (areal recharge) and by induced infiltration from the Rock River due to withdrawals from supply wells near the river. Discharge from the aquifer occurs primarily as leakage to streams and ground-water evapotranspiration.

The water budget for the calibrated steady-state simulation indicated that areal recharge accounts for 38 percent of the sources of water to the Rock River Valley aquifer and leakage from streams contributes 58.7 percent. The largest discharge from the aquifer is leakage to streams, (71.1 percent). The net stream-aquifer leakage is approximately 5 cubic feet per second from the aquifer to the streams. The simulated contributing areas for the wells in the three well fields generally extend to the aquifer boundaries on the west and are generally truncated at the Rock River. The simulated transient water budget for 1996 indicated that the principal sources of water to the aquifer were as follows: (1) winter, spring, and late summer stress periods— leakage from streams and water released from storage and (2) early summer and fall stress periods—areal recharge and leakage from streams. The principal discharges from the aquifer were leakage to streams for all stress periods, ground-water evapotranspiration for the early and late summer stress periods, and addition to storage for the early summer and fall stress periods.

The herbicides atrazine, alachlor, metolachlor, acetachlor, and cyanazine, and metabolites of these herbicides, occurred in concentrations of 0.05 to 11.5 micrograms per liter in the Rock River at Luverne during major runoff events following application of herbicides in the spring. Atrazine and metabolites, alachlor ESA (a metabolite of alachlor), metolachlor and metabolites, metolachlor ESA and metolachlor OA, and acetochlor metabolites acetochlor ESA and acetochlor OA, were detected at concentrations of 0.05 to 2.8 micrograms per liter in municipal supply wells less than 500 feet from the river during November 1995 through August 1997. The Rock River is the major source of the herbicides and metabolites. However, concentrations of atrazine and metabolites, alachlor ESA, metolachlor ESA, and metolachlor OA in supply wells may also reflect sources of these herbicides and metabolites in the ground-water contributing areas to the supply wells. Nitrite plus nitrate nitrogen concentrations in supply wells and in the ground-water contributing area to the Luverne Municipal well field were generally less than 1.5 milligrams per liter. Nitrite plus nitrate nitrogen concentrations of 2.4 to 8.5 milligrams per liter in the Rock River in the Rock County Rural Water well field and 14 to 18 milligrams per liter in the ground-water contributing area to the Rock County Rural Water supply wells are not having a substantial affect on nitrite plus nitrate nitrogen concentrations in most supply wells. Isotopic mixing calculations indicate that proportions of river water withdrawn from supply wells less than 500 feet from the river range from 5 to 60 percent of total withdrawals.

The Rock River is a gaining stream in most reaches, but is losing water to the aquifer in the vicinity of the Luverne Municipal and Rock County Rural Water well fields, located 150 to 1,500 feet from the river. Simulated streamflow losses due to ground-water withdrawals in the well fields were approximately 2.1 cubic feet per second. Because an average of about 1.5 cubic feet per second of the water pumped by Luverne is returned to the Rock River as wastewater discharge, the net steady-state simulated streamflow loss for the study area is 0.6 cubic feet per second. The streamflow losses as a result of ground-water withdrawals are insignificant in comparison to typical streamflow, and are likely to have a measurable effect on streamflow only during low-flow conditions of less than approximately 10 cubic feet per second. 

Model results indicate that the additional water withdrawn by wells due to anticipated increased ground-water withdrawals was derived from a decrease in net leakage of ground water from the aquifer to the streams. The simulations indicated that the increased ground-water withdrawals and normal precipitation resulted in an increase in induced infiltration from the Rock River of 0.1 cubic feet per second for the Luverne Municipal well field and 0.3 cubic feet per second for the Rock County Rural Water well field. Maximum drawdowns ranged from 0.5 to 1.4 feet near the three well fields. For drought conditions, the simulated streamflow losses constituted approximately 30 percent and nearly 65 percent of the flows in the Rock River for the Luverne Municipal and Rock County Rural Water well fields, respectively. Maximum drawdowns ranged from 3.8 to 7.0 feet near the three well fields. Transient simulations with anticipated increased ground-water withdrawals and drought conditions indicated declines in hydraulic heads ranging from 0.2 to 0.4 feet per year in the vicinity of the three well fields, except for near the Rock River.