KNICKPOINT MIGRATION IN WESTERN IOWA STREAMS

Knowing knickpoint migration rates would help prevention of further channel bed incision in western Iowa streams. Past field studies in western Iowa have reported knickpoint migration rates in alluvial loess deposits at average rates as high as 1,780 m/yr and maximum rates as high as 68.6 m/day.

For over four years, the upstream migration of a knickpoint on Mud Creek in Mills County, IA was monitored. The knickpoint migrated a total of 45.8 m, at a rate of 10.8 m/yr; the most rapid retreat of 31.9 m occurred over a six-month period in February-August 2010. The knickpoint maintained a fairly constant height of 1.5-2.2 m, approximately at the same elevation of a stratigraphic discontinuity between silty, stronger soils above and sandy, weaker soils below. Approximately 0.2 m was visible above the downstream normal water surface.

The discrepancy between available hydraulic shear stress and shear strength and erodibility of the knickpoint indicate fluvial erosion is not responsible for most or all of the observed knickpoint migration. A cyclic process, noted in previous studies, is probably responsible where: 1) a pressure differential exists on the vertical knickpoint face due to increased flow velocity as water plunges over the knickpoint; 2) the upper part of the knickpoint is undercut by erosion of less resistant strata below; 3) the undercut sediment fails; 4) the failed material is deposited in the scour hole; 5) the debris is removed, and the vertical knickpoint face is reestablished.

Formation of ice over the knickpoint surface may cause pipe-like flow under the ice or, if the ice attaches to the knickpoint, the plucking and removal of large blocks of sediment; but, ice formation does not directly cause knickpoint retreat.

Past field studies in western Iowa reporting higher knickpoint migration rates were probably eroding through the weaker Roberts Creek Member, whereas this knickpoint is eroding through the upper Gunder Member. Jet erosion tests show the knickpoint bed material (Gunder) is an-order-of-magnitude more resistant to fluvial erosion than the bank soil (Roberts Creek).

The presence and propagation of multiple knickpoints in series caused the erosional topography observed at the current knickpoint; in other words, in highly unstable streams, the current knickpoint is likely the bed scar of past knickpoints.