News Release

MADISON, WI, NOVEMBER 5, 2007 -- Researchers with the United States Geological Surveyís Great Lakes Science Center in Porter, Indiana and the United States Environmental Protection Agency in Athens, Georgia conducted a study to improve the timeliness of bacteria monitoring in a study funded by the National Park Service. The scientists have developed an empirical predictive model for determining E. coli concentrations at two Indiana beaches jointly impacted by two creeks. These models could explain 48-70% of the variation in E. coli, which is a great improvement over the traditional monitoring method, which explained 1-4% of the E. coli variation. It was also determined that shifts in the current direction were a noticeable influence on E. coli counts, with a large current shift toward a receiving beach resulting in higher E. coli counts. This impact was limited in range, where the closer the creek and beach were to each other, the greater the impact.

Recent efforts to model indicator bacteria have typically focused on single beaches, but this is among the first to attempt to define the individual and combined influences of two outfalls on E. coli concentrations at two intervening beaches. Results of this study were published in the September-October 2007 issue of the Journal of Environmental Quality.

Over the summer swimming season of 2004, the researchers collected daily water samples at both beaches and in both creeks and analyzed them for E. coli concentrations. At the same time, water chemistry, hydrology, and weather parameters were measured continuously. Using a mathematical equation, the researchers developed the best-fit model for describing E. coli fluctuations. Models were selected that included the parameters wave height, specific conductivity and turbidity in the creek, barometric pressure, wind direction, and wave period.

Efforts are being made to improve the accuracy and reliability of beach monitoring methods for bacterial water quality. In recent years, traditional monitoring approaches have been increasingly criticized due to the length of time required to conduct the test and the rapidly changing nature of indicator bacteria in swimming water. The need for real-time results has led researchers to explore the applicability of predictive modeling for determining bacterial concentrations in beach water. Differences in morphology, hydrology, and sources among beaches complicate model application, but predictive modeling effectiveness is improving as the state of the science advances.

Results of this study will help beach managers separate the impacts of different point sources on microbiological water quality at their beaches. By doing so, remediation efforts may be directed in a tiered fashion, with attempts first directed at the primary contributing factors. Further, with the increased interest in using empirical models for monitoring beaches, multiple beaches can be assessed simultaneously, making predictions more timely and efficient.