Calibration of a Water-Quality Model for Low-Flow Conditions on the Red River of the North at Fargo, North Dakota, and Moorhead, Minnesota, 2003: Usgs

A time-of-travel and reaeration-rate study was conducted by the U.S. Geological Survey, in cooperation with the North Dakota Department of Health, the Minnesota Pollution Control Agency, and the cities of Fargo, North Dakota, and Moorhead, Minnesota, to provide information to calibrate a water-quality model for streamflows of less than 150 cubic feet per second. Data collected from September 24 through 27, 2003, were used to develop and calibrate the U.S. Environmental Protection Agency Water Quality Analysis Simulation Program model (hereinafter referred to as the Fargo WASP water-quality model) for a 19.2-mile reach of the Red River of the North. The Fargo WASP water-quality model was calibrated for the transport of dye by fitting simulated time-concentration dye curves to measured time-concentration dye curves. Simulated peak concentrations were within 10 percent of measured concentrations. Simulated traveltimes of the dye cloud centroid were within 7 percent of measured traveltimes. The variances of the simulated dye concentrations were similar to the variances of the measured dye concentrations, indicating dispersion was reproduced reasonably well. Average simulated dissolved-oxygen concentrations were within 6 percent of average measured concentrations. Average simulated ammonia concentrations were within the range of measured concentrations. Simulated dissolved-oxygen and ammonia concentrations were affected by the specification of a single nitrification rate in the Fargo WASP water-quality model. Data sets from August 1989 and August 1990 were used to test traveltime and simulation of dissolved oxygen and ammonia. For streamflows that ranged from 60 to 407 cubic feet per second, simulated traveltimes were within 7 percent of measured traveltimes. Measured dissolved-oxygen concentrations were underpredicted by less than 15 percent for both data sets. Results for ammonia were poor; measured ammonia concentrations were underpredicted by as much as 70 percent for both data sets. Overall, application of the Fargo WASP water-quality model to the 1989 and 1990 data sets resulted in poor agreement between measured and simulated concentrations. This likely is a result of changes in the waste-load composition for the Fargo and Moorhead wastewater-treatment plants as a result of improvements to the wastewater-treatment plants since 1990. The change in waste-load composition probably resulted in a change in decay rates and in dissolved oxygen no longer being substantially depressed downstream from the Moorhead and Fargo wastewater-treatment plants. The Fargo WASP water-quality model is valid for the current (2008) treatment processes at the wastewater-treatment plants.