Water quality data collected at 17 sites in urbanized and rural parts of the 438 sq km Irondequoit Creek basin from July 1980 through August 1981 were used to compute annual loads of eight selected constituents. Of the total annual loads of these constituents, 50 to 70% was transported to Irondequoit Bay during a 3.5-month period from late January to early May. Of six mixed-land-use subbasins, the two most highly urbanized had the highest loads of all constituents. Of the four sites representing single land uses, the high density residential site and the housing construction site had the highest loads of all constituents except cadmium. A rainfall-runoff model was used to predict quantity and quality of storm runoff leaving one commercial site, two residential sites, and a large mixed-land-use subbasin. Predicted volume and peak discharges were within 10 to 30% of the measured values; predicted runoff loads of most constituents were within 40 to 60% of measured values. Stormflow modification in the Irondequoit Creek wetlands by two hypothetical control structures was simulated. Results indicate that outflow from the upper wetland could be decreased and stormflows dispersed to a greater part of the upper wetland to increase retention of suspended solids and associated chemical constituents.
