CROSS-SECTIONAL GEOMETRY OF RIVER MEANDER BENDS WITH NONCOHESIVE BANKS: IMPLICATIONS ON MIGRATION RATE AND FORMATIVE PROCESSES

Laboratory experiments have long demonstrated the difficulty in obtaining a single-threaded meandering channel when formed in a floodplain of noncohesive soils. In the presence of excess boundary shear stress, an initially sinuous noncohesive channel will tend to rapidly widen and shallow, ultimately devolving into a braided channel with large width-depth ratio. Despite this phenomenon, single-threaded meandering rivers with noncohesive banks are observed in nature, most commonly as rivers with composite banks, in which the lower bank is comprised of lateral accretion deposits of sand and gravel from a former river planform configuration. In such materials, previous researchers have hypothesized that the bank resistance must be increased to prevent the widening and shallowing process through mechanisms such as slump block armoring or bank vegetation. How much extra bank resistance is necessary to achieve the resulting geometry is difficult to ascertain, because a typical migrating channel cross-sectional geometry formed in noncohesive banks has never been determined to provide a basis of comparison. While physics-based equilibrium cross-sectional geometry has been determined numerically, analytically, and experimentally in straight channels and in channel bends with fixed outer banks, it has yet to be determined under conditions with migrating banks. In this analysis, a highly conceptualized steady-state migrating channel cross-section in an infinite bend of constant centerline radius is determined numerically. The numerical analysis reproduces the phenomena of an excessively wide, shallow channel when the channel migrates due to noncohesive sediment transport processes subjected to helical bend flow. When the critical shear stress is increased beyond that suggested by the grain-size distribution, natural channel cross-sections result. Field data from a highly dynamic reach of the Mackinaw River in Illinois are evaluated in this framework. Deductions are made regarding the increase in the critical shear stress that is required to reproduce the observed cross-sectional geometry and the factors that may be responsible. The migration rates are compared between the numerical model and aerial photograph records of bend migration.