Submitted to: Journal of the American Water Resources Association
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/15/2006
Publication Date: 6/1/2007
Citation: Wells, R.R., Langendoen, E.J., Simon, A. 2007. Modeling Pre- and Post-Dam Removal Sediment Dynamics: The Kalamazoo River, Michigan. Journal of the American Water Resources Association, 43(3): 773-785. DOI:10.1111/j.1752-1688.2007.00062.x
Interpretive Summary: Concerns over the fate of PCB-laden channel sediments in the Kalamazoo River between Otsego and Plainwell, Michigan resulted in the U.S. Geological Survey (USGS) supporting a study to simulate sediment loads and channel changes in the reach. The 8.8 km reach of the Kalamazoo River contains two low-head dams. The state of Michigan is interested in removing these dams while minimizing impacts to the study reach and downstream reaches, and to provide for improved fisheries. This study was designed to evaluate the erosion, transport, and deposition of sediments in the Kalamazoo River between Plainwell and Otsego City, Michigan. Three numerical simulations were carried out over a 17.7-year period of record to evaluate the response of the Kalamazoo River between Otsego and Plainwell, Michigan to current channel conditions instantaneous removal of two low-head dams, and a design channel without the low-head dams. Flows for all three scenarios are based on a modified discharge record from the USGS gage on the Kalamazoo River at Comstock, Michigan. Bank and bed erosion increased significantly for the dams out case. For this scenario, the average-annual sediment load increased 34 times and the erosion of material finer than 0.01 mm increased 3 times. For the scenario including the design channel average-annual sediment load was half of the dams out case but still 15 times greater than the existing dams in case. The dams in (baseline) case clearly provides the smallest loads for both total and fine-grained sediment transport.
Technical Abstract: The study was designed to evaluate the erosion, transport, and deposition of fine-grained sediments in the Kalamazoo River between Otsego and Plainwell, Michigan. The state of Michigan is interested in removing two low-head dams within the study reach while minimizing impacts and providing for improved fisheries. Numerical modeling of channel-erosion processes over a 17.7-year period was conducted using CONCEPTS for three specific scenarios, Dams In or baseline, Dams Out, and Design. The USGS conducted channel surveys, collected and analyzed bed-sediment cores, and performed particle-size analysis for all channel material samples collected by the USDA-ARS, and the USDA-ARS conducted in-situ measurements on channel materials. Flows for all three scenarios are based on a modified discharge record from the USGS gage on the Kalamazoo River at Comstock, Michigan (04106000). Total loads at the downstream boundary are 3040 T/y (deposition) for the Dams In case, 72,000 T/y (erosion) for the Dams Out case, and 21,900 T/y (erosion) for the Design case. Passing the downstream boundary there is net erosion of 1560 T/y fro the Dams In case, net erosion of 53,300 T/y for the Dams Out case, and net erosion of 23,200 T/y for the Design case. Fine-grained loads (<65'm and <10'm) show a similar pattern as total loads with the Dams In case contributing 1360 T/y and 227 T/y, the Dams Out case contributing 5250 T/y and 725 T/y, and the Design case contributing 4640 T/y and 282 T/y. For the Dams In case, the banks contributed 12% of the total in the 65'm class and 46% of the total in the 10'm class. For the Dams Out case, the banks contributed 17% of the total in the 65'm class and 87% of the total in the 10'm class. For the Design case, the banks contributed 5% of the total in the 65'm class and 56% of the total in the 10'm class. The most significant findings of this research are that (1) Bed erosion is an important contributor of suspended-sediment, (2) The Plainwell reach is the greatest contributor of suspended-sediment and fine-grained sediment, (3) Sediment delivery could be significantly reduced by controlling streambank erosion between cross-sections P12 and P15, POC11 and POC16, and G7 and G4, and (4) Removal of the low-head dams will cause loads to increase significantly.