U.S. Geological Survey
Water-Resources Investigations Report 00-4096

Characterization and Simulation of Ground-Water Flow in the Kansas River Valley at Fort Riley,
Kansas, 1990-98

Abstract

Hydrologic data and a ground-water flow model were used to characterize ground-water flow in the Kansas River alluvial aquifer at Fort Riley in northeast Kansas. The ground-water flow model was developed as a tool to project ground-water flow and potential contaminant-transport paths in the alluvial aquifer on the basis of past hydrologic conditions. The model also was used to estimate historical and hypothetical ground-water flow paths with respect to a private- and several public-supply wells.

The ground-water flow model area extends from the Smoky Hill and Republican Rivers downstream to about 2.5 miles downstream from the city of Ogden. The Kansas River Valley has low relief and, except for the area within the Fort Riley Military Reservation, is used primarily for crop production.

Sedimentary deposits in the Kansas River Valley, formed after the ancestral Kansas River eroded into bedrock, primarily are alluvial sediment deposited by the river during Quaternary time. The alluvial sediment consists of as much as about 75 feet of poorly sorted, coarse-to-fine sand, silt, and clay, 55 feet of which can be saturated with ground water. The alluvial aquifer is unconfined and is bounded on the sides and bottom by Permian-age shale and limestone bedrock.

Hydrologic data indicate that ground water in the Kansas River Valley generally flows in a downstream direction, but flow direction can be quite variable near the Kansas River due to changes in river stage. Ground-water-level changes caused by infiltration of precipitation are difficult to detect because they are masked by larger changes caused by fluctuation in Kansas River stage.

Ratios of strontium isotopes Sr⁸⁷ and Sr⁸⁶ in water collected from wells in the Camp Funston Area indicate that the ground water along the northern valley wall originates, in part, from upland areas north of the river valley. Water from Threemile Creek, which flows out of the uplands north of the river valley, had Sr⁸⁷:Sr⁸⁶ ratios similar to those in ground water from wells in the northern Camp Funston Area. In addition, comparison of observed water levels from wells CF90-06, CF97-101, and CF97-401 in the Camp Funston Area and ground-water levels simulated for these wells using floodwave-response analysis indicates that ground-water inflow from bedrock is a hydraulic stress that, in addition to the changing stage in the Kansas River, acts on the aquifer. This hydraulic stress seems to be located near the northern valley wall because the effect of this stress is greater for well CF97-101, which is the well closest to the valley wall.

Ground-water flow was simulated using a modular, three-dimensional, finite-difference ground-water flow model (MODFLOW). Particle tracking, used to visualize ground-water flow paths in the alluvial aquifer, was accomplished using MODPATH. Forward-in-time particle tracking indicated that, in general, particles released near the Kansas River followed much more variable paths than particles released near the valley wall. Although particle tracking does not simulate solute transport, this increased path variability indicates that, near the river, ground-water contaminants could follow many possible paths towards the river, whereas more distant from the river, ground-water contaminants likely would follow a narrower corridor. Particle tracks in the Camp Funston Area indicate that, for the 1990-98 simulation period, contaminants from the ground-water study sites in the Camp Funston Area would be unlikely to move into the vicinity of Ogden's supply wells. Backward-in-time particle tracking indicated that the flow-path and recharge areas for model cells corresponding to Ogden's supply wells lie near the northern valley wall and extend into the northern Camp Funston Area. The flow-path and recharge areas for model cells corresponding to Morris County Rural Water District wells lie within Clarks Creek Valley and probably extend outside the model area.

Three hypothetical simulations, in which pumpage from Ogden's supply wells was increased by 2, 5, or 10 times 1997 pumping rates, indicate that further development of ground-water resources near Ogden could degrade the quality of water pumped from the city wells. Two hypothetical simulations in which hypothetical wells southeast of Ogden were pumped at 100 and 900 gallons per minute indicate that these hypothetical wells would obtain much of their water from the Camp Funston Area, but none of the flow-path areas for model cells corresponding to Ogden's wells or the hypothetical wells intersected the Southwest Funston Landfill. Although the hypothetical simulations indicate the general direction that ground-water contaminants would move under the simulated pumping and climatic conditions, the simulations do not indicate whether contaminants would actually reach pumping wells because the concentrations of many ground-water contaminants decrease over time as a result of naturally occurring processes such as chemical degradation, mechanical and chemical dispersion, and bacterial metabolic action.

Myers, N.C., 2000, Characterization and Simulation of Ground-Water Flow in the Kansas River Valley at Fort Riley, Kansas, 1990-98: U.S. Geological Survey Water-Resources Investigations Report 00-4096, 122 p.

To request a paper copy of this report, email: GS-W-KS_info@usgs.gov