Natural Stream Channel Research Articles

KNICKPOINT BEHAVIOR IN NONCOHESIVE MATERIAL: A LABORATORY STUDY

Short oversteepened reaches were molded in the beds of model channels formed in well-sorted, noncohesive sands 0.67 mm and 2 mm in diameter. In each run of the experiment a fall of 0.1 foot in a length of 1.0 foot was provided. The over-all slopes of the channels upstream and downstream from the oversteepened reach were made equivalent and ranged from 0.0012 to 0.0088. The abrupt break in the profile at the head of the over-steepened segment constituted a knickpoint. Progressive changes in the position of the knickpoint and in the slope of the oversteepened reach were measured during runs in which discharge, over-all slope, and particle size were varied independently. I n every run, the slope of the water surface and the slope of the bed below the knickpoint decreased with time. As the knickpoint moved upstream, the channel directly above the knickpoint first steepened and narrowed. Following the initial steepening, the slope became progressively less. At the lower end of the oversteepened reach, sediment eroded from above was deposited as a dune, which advanced downstream and caused the channel to widen and locally to steepen. Following the passage of the dune, the slope again flattened. For runs with identical initial (over-all) slopes, discharge, and widths, the slope below the knickpoint decreased faster in the channel of finer sand. The rate of change of slope in the oversteepened reach below the knickpoint depends upon the magnitude and the rate of change of erosion along this reach which, in turn, depends upon the magnitude and the rate of change of the sediment transport along the reach. If the rate of transport is great, the oversteepened slope below the knickpoint is reduced rapidly. Analysis of the data indicates that the higher the ratio of the oversteepened slope to the average slope the more rapid the rate of decrease of the oversteepened slope. These results are comparable to changes observed in natural stream channels following meander cutoffs. The laboratory experiment confirms the observation that upstream migration of knickpoints accompanied by undiminished slopes does not occur in noncohesive, homogeneous bed material. Several hypothetical cases are discussed in which it is assumed that resistant material is present in the channel profile.