In a study to understand the historical effects of the construction and decommissioning of dams on the Kalamazoo River, Plainwell to Otesgo, Michigan, and to simulate channel changes that may result if the dams were removed, early to mid-1800s General Land Office surveys and aerial photographs from 1938, 1981, and 1999 were compared in order to identify historical changes in the river’s planform. This analysis of the 80-mile reach from Morrow Dam to the river mouth at Saugatuck provided insight into how susceptible the river has been to channel migration. The comparison showed that changes in channel width and location were caused mainly by construction of dams and subsequent water-level adjustments in the impounded reaches upstream from the dams. Braiding also occurred downstream from one of the dams. Minor changes in channel form that were not caused by the dams, such as the development and cutoff of meander bends, were observed.

A more detailed study in a 5-mile reach passing through the Plainwell and Otsego City Dams included compiling existing valley cross section and longitudinal profile data into a database, assessing bank stability, and using a hydrologic model to simulate the channel as if the dams were removed. Fifty-four valley cross sections compiled from United States Geological Survey and consultant data sets were used as a base for a bank-stability assessment and to design a hypothetical stable channel without the two dams. The channel design involved adjusting the slope, hydraulic geometry, and floodplain width to ensure that water could be transferred through the reach without increasing flooding or erosion problems.

The bank-stability assessment focused on conditions that are critical to failure. This was accomplished through the use of a two step process. The first involved evaluating the sediment removed from the bank toe when the stage is high. The second involved calculating the factor of safety for the bank based on the water table being elevated higher than the stage, mimicing a bank storage effect. Using these paired proccesses, two scenarios of critical conditions were evaluated: dams present and dams removed.

Results of the bank assessments showed that, under both critical-condition scenarios, the streambanks were more susceptible to toe erosion than to block failure. As toe erosion progresses, the banks will eventually collapse as supporting material underneath is removed. Toe erosion for the damsremoved scenario resulted in higher amounts of erosion than for the dams-present scenario, leading to an overall decrease in bank stability. Effects of vegetation on the bank stability were variable; stability for some banks increase if vegetation was present but remain the same for other banks.