Estimation of Constituent Concentrations, Loads, and Yields in Streams of Johnson County, Northeast Kansas, Using Continuous Water-Quality Monitoring and Regression Models, October 2002 through December 2006

Abstract

Johnson County is one of the most rapidly developing counties in Kansas. Population growth and expanding urban land use affect the quality of county streams, which are important for human and environmental health, water supply, recreation, and aesthetic value. This report describes estimates of streamflow and constituent concentrations, loads, and yields in relation to watershed characteristics in five Johnson County streams using continuous in-stream sensor measurements. Specific conductance, pH, water temperature, turbidity, and dissolved oxygen were monitored in five watersheds from October 2002 through December 2006. These continuous data were used in conjunction with discrete water samples to develop regression models for continuously estimating concentrations of other constituents. Continuous regression-based concentrations were estimated for suspended sediment, total suspended solids, dissolved solids and selected major ions, nutrients (nitrogen and phosphorus species), and fecal-indicator bacteria. Continuous daily, monthly, seasonal, and annual loads were calculated from concentration estimates and streamflow. The data are used to describe differences in concentrations, loads, and yields and to explain these differences relative to watershed characteristics. Water quality at the five monitoring sites varied according to hydrologic conditions; contributing drainage area; land use (including degree of urbanization); relative contributions from point and nonpoint constituent sources; and human activity within each watershed. Dissolved oxygen (DO) concentrations were less than the Kansas aquatic-life-support criterion of 5.0 mg/L less than 10 percent of the time at all sites except Indian Creek, which had DO concentrations less than the criterion about 15 percent of the time. Concentrations of suspended sediment, chloride (winter only), indicator bacteria, and pesticides were substantially larger during periods of increased streamflow. Suspended-sediment concentration was nearly always largest at the Mill Creek site. The Mill Creek watershed is undergoing rapid development that likely contributed to larger sustained sediment concentrations. During most of the time, the smallest sediment concentrations occurred at the Indian Creek site, the most urban of the monitored sites, likely because most of the streamflow originates from wastewater-treatment facilities located just upstream from the monitoring site. However, estimated annual suspended-sediment load and yield were largest annually at the Indian Creek site because of substantial contributions during storm runoff. At least 90 percent of the total annual sediment load in 2005?06 at all five monitoring sites occurred in less than 2 percent of the time, generally associated with large storm runoff. About 50 percent of the 2005 sediment load at the Blue River site occurred during a single 3-day storm, the equivalent of less than 1 percent of the time. Suspended-sediment concentration is statistically related to other water-quality constituents, and these relations have potential implications for implementation of best management practices because, if sediment concentrations are decreased, concentrations of sediment-associated constituents such as suspended solids, some nutrients, and bacteria will also likely decrease. Chloride concentrations were largest at the Indian and Mill Creek sites, the two most urban stream sites which also are most affected by road-salt runoff and wastewater-treatment-facility discharges. Two chloride runoff occurrences in January?February 2005 accounted for 19 percent of the total chloride load in Indian Creek in 2005. Escherichia coli density at the Indian Creek site was nearly always largest of the five sites with a median density more than double that of any other site and 15 times the density at the Blue River site which is primarily nonurban. More than 97 percent of the fecal coliform bacteria load at the Indian Creek site and near the B