199 - Prediction of E. coli Concentrations in Highly Urbanized Watersheds: An Analysis Using Diverse Environmental Stressors

Fecal indicator bacteria (FIB) such as Escherichia coli (E. coli) serve as indicators of water quality. Found primarily in the intestines of humans and animals, elevated concentrations of E. coli can pose risks to human health, aquatic ecosystems, and recreational water activities. The level of E. coli concentration in a watershed can be influenced by several environmental factors such as hydrological characteristics, watershed topography, landscape, and meteorological conditions. These factors notably impact stream runoff, subsequently affecting contamination levels. Understanding the intricate interplay between E. coli and other environmental variables can help predict future E. coli concentration in waterbodies. In this study, we leverage remote sensing and geographic information systems to extract diverse parameters related to land use, topography, soil, and meteorological parameters that includes rainfall, humidity, temperature. We also sourced water quality and hydrological data from three United States Geological Survey (USGS) stations. Drawing from data spanning 2000-2023, we employ four decision-treed based machine learning algorithms such as Extreme Gradient Boosting and Random Forest to predict future E. coli concentrations in a watershed. Moreover, our research aims to identify the most significant environmental factors influencing these concentrations. Our predictive model will be tested across three distinct watersheds within the Chattahoochee River National Recreation Area (CRNRA), a region with historically high records of E. coli concentrations. The prediction of future concentrations of E. coli can help in formulating effective water quality management strategies and safeguarding public health policies. Our study on CRNRA watersheds offers a framework potentially applicable to other watersheds.