Comparison of streamflow and water-quality data collection techniques for the Saginaw River, Michigan

Abstract

In 2009, the Michigan Department of Environmental Quality and the U.S. Geological Survey developed a plan to compare the effect of various streamgaging and water-quality collection techniques on streamflow and stream water-quality data for the Saginaw River, Michigan. The Saginaw River is the primary contributor of surface runoff to Saginaw Bay, Lake Huron, draining approximately 70 percent of the Saginaw Bay watershed. The U.S. Environmental Protection Agency has listed the Saginaw Bay system as an "Area of Concern" due to many factors, including excessive sediment and nutrient concentrations in the water. Current efforts to estimate loading of sediment and nutrients to Saginaw Bay utilize water-quality samples collected using a surface-grab technique and flow data that are uncertain during specific conditions. Comparisons of current flow and water-quality sampling techniques to alternative techniques were assessed between April 2009 and September 2009 at two locations in the Saginaw River. Streamflow estimated using acoustic Doppler current profiling technology was compared to a traditional stage-discharge technique. Complex conditions resulting from the influence of Saginaw Bay on the Saginaw River were able to be captured using the acoustic technology, while the traditional stage-discharge technique failed to quantify these effects. Water-quality samples were collected at two locations and on eight different dates, utilizing both surface-grab and depth-integrating multiple-vertical techniques. Sixteen paired samples were collected and analyzed for suspended sediment, turbidity, total phosphorus, total nitrogen, orthophosphate, nitrite, nitrate, and ammonia. Results indicate that concentrations of constituents associated with suspended material, such as suspended sediment, turbidity, and total phosphorus, are underestimated when samples are collected using the surface-grab technique. The median magnitude of the relative percent difference in concentration based on sampling technique was 37 percent for suspended sediment, 26 percent for turbidity, and 9.7 percent for total phosphorus samples collected at both. Acoustic techniques were also used to assist in the determination of the effectiveness of using acoustic-backscatter information for estimating the suspended-sediment concentration of the river water. Backscatter data was collected by use of an acoustic Doppler current profiler, and a Van Dorn manual sampler was simultaneously used to collect discrete water samples at 10 depths (3.5, 7.5, 11, 14, 15.5, 17.5, 19.5, 20.5, 22, and 24.5 ft below the water surface) along two vertical profiles near the center of the Saginaw River near Bay City. The Van Dorn samples were analyzed for suspended-sediment concentrations, and these data were then used to develop a relationship between acoustic-backscatter data. Acoustic-backscatter data was strongly correlated to sediment concentrations and, by using a linear regression, was able to explain 89 percent of the variability. Although this regression technique showed promise for using acoustic backscatter to estimate suspended-sediment concentration, attempts to compare suspended-sediment concentrations to the acoustic signal-to-noise ratio estimates, recorded at the fixed acoustic streamflow-gaging station near Bay City (04157061), resulted in a poor correlation.