Watershed characteristics and water-quality trends and loads in 12 watersheds in Gwinnett County, Georgia

Abstract

The U.S. Geological Survey, in cooperation with Gwinnett County Department of Water Resources, established a Long-Term Trend Monitoring (LTTM) program in 1996. The LTTM program is a comprehensive, long-term, water-quantity and water-quality monitoring program designed to document and analyze the hydrologic and water-quality conditions of selected watersheds of Gwinnett County, Georgia. Water-quality monitoring initially began in six watersheds and was expanded to another six watersheds in 2001. As part of the LTTM program, streamflow, precipitation, water temperature, specific conductance, and turbidity were measured continuously at the 12 watershed monitoring stations for water years 2004–09. In addition, discrete water-quality samples were collected seasonally from May through October (summer) and November through April (winter), including one base-flow and three stormflow event composite samples, during the study period. Samples were analyzed for nutrients (nitrogen and phosphorus), total organic carbon, trace elements (total lead and total zinc), total dissolved solids, and total suspended sediment (total suspended solids and suspended-sediment concentrations). The sampling scheme was designed to identify variations in water quality both hydrologically and seasonally. The 12 watersheds were characterized for basin slope, population density, land use for 2009, and the percentage of impervious area from 2000 to 2009. Precipitation in water years 2004–09 was about 18 percent below average, and the county experienced exceptional drought conditions and below average runoff in water years 2007 and 2008. Watershed water yields, the percentage of precipitation that results in runoff, typically are lower in low precipitation years and are higher for watersheds with the highest percentages of impervious areas. A comparison of base-flow and stormflow water-quality samples indicates that turbidity and concentrations of total ammonia plus organic nitrogen, total nitrogen, total phosphorus, total organic carbon, total lead, total zinc, total suspended solids, and suspended-sediment concentrations increased with increasing discharge at all watersheds. Specific conductance, however, decreased during stormflow at all watersheds, and total dissolved solids concentrations decreased during stormflow at a few of the watersheds. Total suspended solids and suspended-sediment concentrations typically were two orders of magnitude higher in stormflow samples, turbidities were about 1.5 orders of magnitude higher, total phosphorus and total zinc were about one order of magnitude higher, and total ammonia plus organic nitrogen, total nitrogen, total organic carbon, and total lead were about twofold higher than in base-flow samples. Seasonal patterns and long-term trends in flow-adjusted water-quality concentrations were identified for five representative constituents—total nitrogen, total phosphorus, total zinc, total dissolved solids, and total suspended solids. Seasonal patterns for all five constituents were fairly similar, with higher concentrations in the summer and lower concentrations in the winter. Significant linear long-term trends in stormflow composite concentrations were identified for 36 of the 60 constituent-watershed combinations (5 constituents multiplied by 12 watersheds) for the period of record through water year 2011. Significant trends typically were decreasing for total nitrogen, total phosphorus, total suspended solids, and total zinc and increasing for total dissolved solids. Total dissolved solids and total suspended solids trends had the largest magnitude changes per year. Stream water loads were estimated for 10 water-quality constituents. These estimates represent the cumulative effects of watershed characteristics, hydrologic processes, biogeochemical processes, climatic variability, and human influences on watershed water quality. Yields, in load per unit area, were used to compare loads from watersheds with different sizes. A load estimation approach developed for the Gwinnett County LTTM program that incorporates storm-event composited samples was used with some minor modifications. This approach employs the commonly used regression-model method. Concentrations were modeled as a function of discharge, time, season, and turbidity to improve model predictions and reduce errors in load estimates. Total suspended solids annual loads have been identified in Gwinnett County’s Watershed Protection Plan for target performance criterion. The amount of annual runoff is the primary factor in determining the amount of annual constituent loads. Below average runoff during water years 2004–09, especially during water years 2006–08, resulted in corresponding below average loads. Variations in constituent yields between watersheds appeared to be related to various watershed characteristics. Suspended sediment (total suspended solids and suspended-sediment concentrations) along with constituents transported predominately in solid phase (total phosphorus, total organic carbon, total lead, and total zinc) and total dissolved solids typically had higher yields from watersheds that had high percentages of impervious areas or high basin slope. High total nitrogen yields were also associated with watersheds with high percentages of impervious areas. Low total nitrogen, total suspended solids, total lead, and total zinc yields appear to be associated with watersheds that have a low percentage of high-density development. Total suspended solids yields were lower in drought years, water years 2007–08, from the combined effects of less runoff and the result of fewer, lower magnitude storms, which likely resulted in less surface erosion and lower stream sediment transport.