The city of Independence, Missouri, operates a well field in the Missouri River alluvial aquifer. Steady-state ground-water flow simulation, particle tracking, and the use of chemical and isotopic composition of river water, ground water, and well-field pumpage in a two-component mixing equation were used to determine the source contributions of induced inflow from the Missouri River and recharge to ground water from precipitation in well-field pumpage.

Steady-state flow-budget analysis for the simulation-defined zone of contribution to the Independence well field indicates that 86.7 percent of well-field pumpage is from induced inflow from the river, and 6.7 percent is from ground-water recharge from precipitation. The 6.6 percent of flow from outside the simulation-defined zone of contribution is a measure of the uncertainty of the estimation, and occurs because model cells are too large to uniquely define the actual zone of contribution. Flow-budget calculations indicate that the largest source of water to most wells is the Missouri River.

Particle-tracking techniques indicate that the Missouri River supplies 82.3 percent of the water to the Independence well field, ground-water recharge from precipitation supplies 9.7 percent, and flow from outside defined zones of contribution supplies 8.0 percent. Particle tracking was used to determine the relative amounts of source water to total well-field pumpage as a function of traveltime from the source. Well-field pumpage that traveled 1 year or less from the source was 8.8 percent, with 0.6 percent from the Missouri River, none from precipitation, and 8.2 percent between starting cells. Well-field pumpage that traveled 2 years or less from the source was 10.3 percent, with 1.8 percent from the Missouri River, 0.2 percent from precipitation, and 8.3 percent between starting cells. Well-field pumpage that traveled 5 years or less from the source was 36.5 percent, with 27.1 percent from the Missouri River, 1.1 percent from precipitation, and 8.3 percent between starting cells. Well-field pumpage that traveled 10 years or less from the source was 42.7 percent, with 32.6 percent from the Missouri River, 1.8 percent from precipitation, and 8.3 percent between starting cells. Well-field pumpage that traveled 25 years or less from the source was 71.9 percent, with 58.9 percent from the Missouri River, 4.7 percent from precipitation, and 8.3 percent between starting cells.

Results of chemical (calcium, sodium, iron, and fluoride) and isotopic (oxygen and hydrogen) analyses of water samples collected from the Missouri River, selected monitoring wells around the Independence well field, and combined well-field pumpage were used in a two component mixing equation to estimate the relative amount of Missouri River water in total well-field pumpage. The relative amounts of induced inflow from the Missouri River in well-field pumpage ranged from 49 percent for sodium to 80 percent for calcium, and sensitivities ranged from 0 percent for iron to plus or minus 35 percent for naturally occurring stable isotope (¹⁸O). The average of all mixing equation results indicated that 61 percent of well-field pumpage was from induced inflow from the Missouri River.

All methods used in the study indicate that more than one-half of the water in well-field pumpage was inflow from the Missouri River. River inflow estimates from ground-water simulation methods are larger and error values are smaller than those using chemical and isotopic data in the mixing equation, although substantial uncertainties exist for both estimation methods. Because of the complex hydrology of the aquifer near the Independence well field, the source estimates using particle tracking probably are the most reliable of the ground-water simulation methods. Mixing equation results are less reliable than those of the ground-water simulation for this study. However, more reliable results can be obtained from the mixing equation by increasing the number of samples and collecting samples for a longer period of time, and during different flow conditions. In the absence of a calibrated ground-water flow simulation, the mixing equation can provide a reasonable estimate of the sources of water to a well field at relatively low cost, if sources of error are clearly understood.