Escherichia coli bacteria density in relation to turbidity, streamflow characteristics, and season in the Chattahoochee River near Atlanta, Georgia, October 2000 through September 2008—Description, statistical analysis, and predictive modeling

Abstract

Water-based recreation—such as rafting, canoeing, and fishing—is popular among visitors to the Chattahoochee River National Recreation Area (CRNRA) in north Georgia. The CRNRA is a 48-mile reach of the Chattahoochee River upstream from Atlanta, Georgia, managed by the National Park Service (NPS). Historically, high densities of fecal-indicator bacteria have been documented in the Chattahoochee River and its tributaries at levels that commonly exceeded Georgia water-quality standards. In October 2000, the NPS partnered with the U.S. Geological Survey (USGS), State and local agencies, and non-governmental organizations to monitor Escherichia coli bacteria ( E. coli ) density and develop a system to alert river users when E. coli densities exceeded the U.S. Environmental Protection Agency (USEPA) single-sample beach criterion of 235 colonies (most probable number) per 100 milliliters (MPN/100 mL) of water. This program, called BacteriALERT, monitors E. coli density, turbidity, and water temperature at two sites on the Chattahoochee River upstream from Atlanta, Georgia. This report summarizes E. coli bacteria density and turbidity values in water samples collected between 2000 and 2008 as part of the BacteriALERT program; describes the relations between E. coli density and turbidity, streamflow characteristics, and season; and describes the regression analyses used to develop predictive models that estimate E. coli density in real time at both sampling sites. Between October 23, 2000, and September 30, 2008, about 1,400 water samples were collected and turbidity was measured at each of the two USGS streamgaging stations in the CRNRA near the cities of Norcross and Atlanta, Georgia. At both sites, water samples were collected at frequencies ranging from daily to twice per week and analyzed in the laboratory for E. coli bacteria, using the Colilert-18® and Quanti-tray-2000® defined substrate method, and turbidity. Beginning in mid-2002, turbidity and water temperature were measured in real time at both sites. Streamflow at both sites is affected by the operation of two hydroelectric facilities upstream that release water in response to daily peak power demands in the area. During dry weather, offpeak water released from both dams ranges from about 600 to 1,500 cubic feet per second. During dry weather, 98 and 93 percent of water samples from Norcross and Atlanta sites, respectively, contained E. coli densities below the USEPA single-sample beach criterion (235 MPN/100 mL). Conversely during stormflow, only 26 percent of the samples from Norcross and 10 percent of the samples from Atlanta contained E. coli densities below the USEPA beach criterion. At both sites, median E. coli density and turbidity were statistically greater in stormflow samples than dry-weather samples. Furthermore, median E. coli density and turbidity were statistically lower at Norcross than at Atlanta during dry weather. During stormflow, median turbidity values were statistically similar at the two sites (36 and 35 formazin nephelometric units at Norcross and Atlanta, respectively); whereas the median E. coli density was statistically higher at Atlanta (810 MPN/100 mL) than at Norcross (530 MPN/100 mL). During dry weather, the maximum E. coli density was 1,200 MPN/100 mL at Norcross and 9,800 MPN/100 mL at Atlanta. During stormflow, the maximum E. coli density was 18,000 MPN/100 mL at Norcross and 28,000 MPN/100 mL at Atlanta. Regression analyses show that E. coli density in samples was strongly related to turbidity, streamflow characteristics, and season at both sites. The regression equation chosen for the Norcross data showed that 78 percent of the variability in E. coli density (in log base 10 units) was explained by the variability in turbidity values (in log base 10 units), streamflow event (dry-weather flow or stormflow), season (cool or warm), and an interaction term that is the cross product of streamflow event and turbidity. The regression equation chosen for the Atlanta data showed that 76 percent of the variability in E. coli density (in log base 10 units) was explained by the variability in turbidity values (in log base 10 units), water temperature, streamflow event, and an interaction term that is the cross product of streamflow event and turbidity. Residual analysis and model confirmation using new data indicated the regression equations selected at both sites predicted E. coli density within the 90 percent prediction intervals of the equations and could be used to predict E. coli density in real time at both sites.