Paula L. Sturdevant-Rees
Disinfection byproducts may be carcinogenic compounds and are present within most drinking water supply systems. These hannful compounds result from an mteraction between natural organic matter and conventional disinfectants such as chlorine or chlorine-based solutions. The level of disinfection byproducts within a distribution system is controlled by federal and state regulations; pending federal regulations propose increasingly stringent allowable levels of disinfection byproducts. The goals of this investigation were to 1) identify long-term variability associated with natural organic matter across the country, 2) develop empirical relationships between anthropogenic activity, land use/land cover and natural organic matter and 3) construct a mechanistic model that describes the fate and transport of natural organic matter within a water supply catchment. To accomplish the first goal, statistical analysis, specifically a seasonal Kendall test, was implemented to analyze natural organic matter trends across the country. This technique provided accurate descriptions of the long-term variability across the nation. The second goal was accomplished via regression analysis. Multiple linear regression techniques developed empirical models that related land use/land cover to natural organic matter. The third goal was satisfied through mass balance equations and a Fibonacci optimization technique. An inputoutput model was constructed and an optimization algorithm detennined reaction constants. The results of the seasonal Kendall analysis yielded trends of regional similarity. The East and Gulf Coasts exhibited increasing trends in natural organic matter, while the Pacific Northwest exhibited decreasing trends. The interior of the country expressed significant variability in trend and no regional conclusions were drawn. Multiple linear regression identified several similarities between watersheds across the country. Wetlands and urban development were shown to most significantly influence natural organic matter. The regression analyses also presented the effects of seasonality and the impact of varied hydrologic conditions upon natural organic matter within water supply catchments. Mechanistic modeling, via optimization, produced kinetic rate constants related to a riverine system. These rate constants accurately described the fate and transport of natural organic matter within the system. The optimization techniques and simplified model adequately described the particular watershed. The statistical and mechanistic techniques were quite accurate for the available data, but the results and findings of this investigation are by no means complete. There exists voluminous data regarding natural organic matter and disinfection byproducts. These available data, from local collection agencies, civic organizations and academic research, should be incorporated into the current database to gain a more thorough analysis of long-term variability across the country and the affects of anthropogenic activity.