Over a 1.5-year period, water-quality data were collected for seven small drainage basins in urban aeas of Portland, Oreg. Analysis of the data followed three approaches. First, the constituent concentrations were analyzed. Average concentrations of suspended sediment, settleable solids, and fecal coliform bacteria generally exceeded levels expected for secondary waste-treatment plant effluent, whereas biochemical oxygen demand concentrations were lower than expected. The second analytical approach established correlations and bivariate regression relationships between constituents for individual storms in each basin, for all storms in each basin, and for all storms in all basins. Generally, correlation coefficients decreased when progressing from data for individual storms in each basin, to data for all storms in each basin, to data for all storms in all basins. In the third approach, storm yields for 10 constituents were related to basin and precipitation characteristics by use of multiple-linear-regression techniques. Storm yields for suspended sediment varied by about four orders of magnitude. Generally, results of the multiple-regression analysis indicated that variations in storm yields were highly dependent on precipitation characteristics, with total rainfall of the storm frequently explaining most of the variation of the dependent variable.
