BED SEDIMENT GRAIN SIZE DISTRIBUTION AND FLOW DYNAMICS OF INDIANHEAD RESERVOIR, ST. CROIX RIVER, MN/WI

Karen Jackson, Macalester College
Kelly R. MacGregor, Macalester College
Daniel J. Hornbach, Macalester College
Mark C. Hove, Macalester College

Dams alter sediment transport and flow dynamics in rivers by acting as a physical barrier to the downstream movement of bed sediment, by slowing water velocity, and allowing suspended material to settle into the reservoir. Interstate Park, located several kilometers below the St. Croix Falls Dam, is host to a large population of native mussels, including threatened and endangered species. Over the past 20 years Hornbach and others (2009) have documented a 90% decrease in the juvenile mussel population at this same location, as well as a gradual decrease in the grain size from mixed sand/gravel to sand at the riverbed. One hypothesis is that the Indianhead Reservoir above the dam may be a significant source of the fine sediment found at Interstate Park. The build up of sediment behind the dam could allow sand size particles to be transported over the dam. Changes in dam operation to run-of-river in the last decade may also affect sand transport across the dam. We aim to quantify grain size distribution at the bed, measure suspended and bedload sediment in transport across a range of water discharges, and characterize flow dynamics of Indianhead Reservoir in an effort to understand spatial and temporal variability in hydrology and sediment transport. Our objective is to determine if the sediment at Interstate Park could originate from the river upstream of the dam.

Reservoir samples were collected from Lions Park boat ramp to the safety line just above the dam. This 2 km portion of the reservoir was divided into 13 evenly spaced transects. Grab samples and gravity cores of bed sediment, suspended sediment concentration (SSC) samples, vertical water velocity profiles using an Acoustic Doppler Current Profiler (ADCP), and high resolution bathymetry readings were collected at three points along each transect. Bathymetric data demonstrate a deeper thalweg exists in the reservoir, with depths ranging from 2-15 m. Vertical velocity profiles reveal flow velocities between 0 and 32 cm/s during the summer low flows. Bed sediments get finer closer to the dam, however, evidence from long cores reveals stochastic deposition of coarse sand and woody debris. The amount of sediment in SSC samples increases with depth in the water column. As water discharge increases, calculated shear stresses increase, thus increasing the likelihood of sediment transport. One site under moderate water discharge (8,050 cfs) exhibits basal shear stress measurements similar to fluvial systems (14.84 dynes/cm²). However, at low discharges (3,100 cfs) ADCP profiles show little vertical variation in water velocity. During spring or flood stages (~25,000 cfs), the larger shear stresses would allow the sand grain size fraction to likely reach and breach the dam. Results demonstrate sand is present in the reservoir close to the dam even during summer low flow discharges. Calculations of settling velocities for available bed sediment, and estimates of residence time for water in the reservoir will further illuminate our understanding of sand transport behind the dam. Additional electron microscopy to compare grain size, composition, and grain morphology of reservoir and Interstate Park bed sediment will improve our understanding of possible sediment transport from the reservoir to mussel beds downstream.