BEDLOAD TRANSPORT FOLLOWING THE REMOVAL OF MILLTOWN DAM, CLARK FORK RIVER, MONTANA

Current understanding of sediment transport processes, sediment budgets and modeling associated with removal of large dams is limited because data are deficient. Future dam removals will benefit from quantitative monitoring studies and model validation and development. The March 2008 removal of Milltown Dam on the Clark Fork River, MT, presented a unique opportunity to better understand the rates and processes by which sediment evacuates a reservoir. In 2008 and 2009, 57 total cross-section integrated bedload samples were collected from bridges upstream of, within and downstream of Milltown Reservoir using either a Toutle River (30 cm inlet nozzle), BL-84 (8 cm nozzle), or modified Helley-Smith (15 cm nozzle) sampler at discharges between 30 and 500 m³/s. Bedload samples collected 4 km below the dam in 2008 contained about 95% sand or smaller sized particles, and bedload transport rates were much higher at a moderate discharge on the rising limb than at the peak. In contrast, at the same site in 2009, bedload was dominated by gravel-sized particles, with the highest rate of transport occurring at peak discharges. These results suggest that in 2008, the rising limb of the hydrograph produced substantial bedload transport of mobile fine sediments out of Milltown Reservoir, with these sediments moving into suspension as transport stage increased. Further, as transport stage increased, coarse sediment deposits mobilized in the upper reaches of the reservoir may have armored remaining fine sediment deposits, reducing subsequent fine grained volumetric transport rates. Observed sediment transport downstream of the Milltown Dam site, including our bedload measurements and suspended sediment measurements by USGS, greatly exceeded rates predicted by one-dimensional (HEC-6) sediment transport modeling performed prior to the removal, highlighting the uncertainties in applying such models to unconfined reservoir sediment deposits.