The Effectiveness of the Chemical Perturbation Index for Monitoring Water Quality in Three Mixed-use Urban Watersheds, Knoxville, Tennessee

Landscape change within the United States has resulted in significant physical and chemical alteration of our nation's surface waters. Many research projects have demonstrated that landscape features moderate or cause observed water quality conditions. Urbanization is a rapidly growing form of landscape change in the United States and ranks second to agriculture as a major cause of stream degradation. Understanding the effect of urbanization on surface waters is only one component of the larger issue of restoring and maintaining the integrity of urban stream water quality. Effective watershed management is a social process that requires the inclusion of local citizens and community alliances. To this end, communities need the tools to gather useful and interpretable data about water quality. The Chemical Perturbation Index (CPI) may potentially be an inexpensive and easily interpretable index of water quality parameters that may be able to characterize both spatial and temporal changes in stream chemistry due to urbanization. The primary purpose of this study is to test the usefulness of the Chemical Perturbation Index as a tool for urban water quality assessment. To gain insight into this question, I explored the ability of the CPI to describe differences in the water quality of three mixed-use urban watersheds: Third Creek, Second Creek, and Goose Creek in Knox County, Tennessee.

To explore the usefulness of the CPI for urban water quality monitoring, I compared the CPI to other methods of describing and determining water quality found in the literature. My research into the CPI's effectiveness as a water quality monitoring tool yielded mixed results. Statistical measures of the individual components of the CPI correlated with changes in landscape and land use characteristics throughout the research subbasins. The CPI itself did not show a relationship with the landscape or land use characteristics within the subbasins, but did show a general relationship with geologic characteristics. Pollution indicators modeled using the Hydrologic Simulation Program Fortran (HSPF) also failed to correlate significantly with the CPI or to demonstrate clear relationships between simulated pollution indicators· and the individual components of the CPI. The lack of expected correlations between landscape and land use factors and many of my chemical measures raises several questions about water quality in urban watersheds.

While previous research has demonstrated a connection between water chemistry and land use, previous studies have generally been done in larger watersheds than those in my research. In smaller watersheds that are urbanized to a high degree, the relationship between the degree of urbanization and water quality may not be as strong. This suggests that the relationship between water chemistry and landscape alteration does not necessarily extend across threshold levels for watershed scale and amount of urbanization. The CPI reflected to geologic characteristics, suggesting that, in extensively urbanized watersheds, factors other than land use account for the observed differences in water quality.