Thursday, November 8, 2007 - 11:35 AM
349-12

Assessing Trends in River Water Quality.

H.J.O. Johnson, Minnesota Department of Agriculture, 625 Robert Street North, St. Paul, MN 55155, Satish Gupta, 1991 Upper Buford Circle, University of Minnesota, Department of Soil, Water, & Climate, St. Paul, MN 55108, and A. Vecchia, USGS, 821 East Interstate Ave., Bismarck, ND 58501.

Excessive loading of sediment and nutrients to rivers is a major problem in many parts of the United States. One such river in Minnesota is the Minnesota River. The goal of this study was to test statistical procedures that can quantify trends in water quality of the Minnesota River. This quantification is important for assessing the impact of management practices before their large-scale implementation in the watershed. The two procedures tested in this research were the non parametric Seasonal Kendall (SK) trend model and the parametric United States Geological Survey Quality of Water trend (QW) analysis program. Except for nitrate-N concentration, there were decreasing trends in water quality variables at all three locations. However, magnitudes of trends between the two methods were quite different. Further analysis indicated that decreasing trends in water quality variables were mainly due to decreasing trends in earlier periods (1976 to the mid to late 1980's). Nitrate-N concentrations generally increased with time at all three locations. Trends in total suspended solids concentrations on a sliding time frame showed that trends were strongly sensitive to addition or deletion of one or more years of water quality data. Also, significance of the trends was somewhat opposite of each other for the SK and the QW models. These discrepancies between methods stress the need for refinement in these models or development of additional models that are less sensitive to the length of the data and its manipulations.