Abstract

Increasingly complex water-management decisions require water-quality monitoring programs that provide data for multiple purposes, including trend analyses, to detect improvement or deterioration in water quality with time. Understanding surface-water-quality trends assists resource managers in identifying emerging water-quality concerns, planning remediation efforts, and evaluating the effectiveness of the remediation. This report presents the results of a study conducted by the U.S. Geological Survey, in cooperation with the Kentucky Energy and Environment Cabinet–Kentucky Division of Water, to analyze and summarize long-term water-quality trends of selected properties and water-quality constituents in selected streams in Kentucky’s ambient stream water-quality monitoring network.

Trends in surface-water quality for 15 properties and water-quality constituents were analyzed at 37 stations with drainage basins ranging in size from 62 to 6,431 square miles. Analyses of selected physical properties (temperature, specific conductance, pH, dissolved oxygen, hardness, and suspended solids), for major ions (chloride and sulfate), for selected metals (iron and manganese), for nutrients (total phosphorus, total nitrogen, total Kjeldahl nitrogen, nitrite plus nitrate), and for fecal coliform were compiled from the Commonwealth’s ambient water-quality monitoring network. Trend analyses were completed using the S-Plus statistical software program S-Estimate Trend (S-ESTREND), which detects trends in water-quality data. The trend-detection techniques supplied by this software include the Seasonal Kendall nonparametric methods for use with uncensored data or data censored with only one reporting limit and the Tobit-regression parametric method for use with data censored with multiple reporting limits. One of these tests was selected for each property and water-quality constituent and applied to all station records so that results of the trend procedure could be compared among stations. Flow-adjustment procedures were used with these techniques at all stations to remove the effects of streamflow on water-quality variability. Flow adjustments were used for all constituents, except temperature. A decreasing trend indicates a decrease in concentration of a particular constituent; whereas, an increasing trend indicates an increase in concentration and potential degradation in water quality.

Trend results varied statewide by station and by physical property and water-quality constituent. The results for all stations and all physical properties and water-quality constituents examined had at least one statistically significant (p-value ‹0.05) increasing or decreasing trend during the specified period of record. Water temperature and concentrations of dissolved oxygen had no significant decreasing trends at any station. Water temperature had one significant increasing trend at the South Fork Cumberland River near Blue Heron station. Specific conductance and concentrations of hardness had one significant decreasing trend at the South Fork Cumberland River near Blue Heron station. pH also had a significant decreasing trend at the Mud River near Gus station. Concentrations of total suspended solids had 1 increasing trend at the Kentucky River at High Bridge station and 10 decreasing trends with 5 of those stations located in the Cumberland River Basin.

Major ions analyzed for trends included chloride and sulfate. Concentrations of chloride at the 37 stations had increasing trends at 15 stations, decreasing trends at 3 stations, and no significant trend in concentration over time at 19 stations. Most of the increasing trends in concentrations of chloride are located in the northern part of Kentucky, possibly indicating an increase in the use of road salts for road deicing and (or) the result of resource extraction (oil, gas, and coal). Increasing trends of sulfate concentrations were detected at seven stations, all located in the Appalachian Region of eastern Kentucky, where water quality in streams is potentially affected by surface mining. Two stations with the largest median concentrations of sulfate had no trend.

Concentrations of total iron had statistically significant increasing (one station) and decreasing (four stations) trends scattered in the eastern part of Kentucky, where high concentrations of total iron are common. The Tygarts Creek near Lynn station had a significant increasing trend in total iron; although, an explanation for this increase is beyond the scope of this report. Five stations had median concentrations of total iron greater than 1,000 micrograms per liter, but none had a statistically significant trend. Concentrations of total manganese had increasing trends at 2 stations, decreasing trends at 13 stations, and no significant trend in concentration over time at 22 stations. All six monitoring stations in the Cumberland River Basin had decreasing concentrations of total manganese, possibly because of a decline in coal extraction. The median concentration of total manganese exceeded the U.S. Environmental Protection Agency’s secondary drinking-water standard (50 micrograms per liter) for total manganese at 29 stations. Eleven of these stations had decreasing trends, 2 stations had increasing trends, and 16 stations had no trend in concentrations of total manganese.

Trend analysis for all nitrogen constituents analyzed indicated no trends at 25 stations. Concentrations of total nitrogen had two significant increasing trends at the Little Sandy at Argillite station and at the Little River near Cadiz station. Trend analysis for concentrations of total phosphorus had 14 decreasing trends and 4 increasing trends. Three of the four stations with increasing trends in total phosphorus were located in the Kentucky River Basin; the other station, Green River near Woodbury, was located in the Green River Basin. Trends for concentrations of ammonia were not performed because concentrations of ammonia were highly censored (›50 percent) for the majority of stations.

Trend analysis for concentrations of fecal coliform showed 16 significant decreasing trends and 3 significant increasing trends, statewide. Every major drainage basin had at least one station with a significant decreasing trend in concentration of fecal coliform, except the Middle Ohio–Little Miami River Basin. This basin only contains one ambient water-quality monitoring network station, (Kinniconick Creek near Tannery.

Posted April 3, 2009