Incision dynamics along river profiles: the influence of uplift and lithology

Abstract:

We investigate how river long profiles and their knickzones and knickpoints interact with uplift (i.e. base level fall) rate and lithology. These results highlight lithogenic controls on knickpoint evolution. First, based on dated and regionally correlated incision markers from West Africa, we reconstructed the entire paleo-long profiles of big rivers such as the Niger at ca. 24, 11 and 6 Ma, as well as the Eocene topography those rivers have dissected. Though spatially and temporally variable, incision remained mostly below 10 m/my with a mean around 5 m/my. The spatial stability of both the river outlets and divides imposed maintenance or increasing concavity of the river long profiles through time, resulting from spatially contrasted adjustment of river segments bounded by recurrent lithogenic knickzones that persisted since 24 Ma. Drainages evolved preferentially by very slow slope decrease or uniform incision in between the stationary knickzones of evolving amplitude, with apparently no relation to base level change. Therefore, knickzone height or position may not simply reflect the transient response of big rivers to base level fall as predicted by stream-power incision river models. Second, we study river profiles experimentally on an 80 cm-long and 2 cm-wide flume filled with a ‘bedrock’ cohesive paste made of silica flour mixed with water and kaolinite. Water is introduced from one extremity at a constant discharge to erode the bedrock. Uplift rate is controlled by a sliding gate at the outlet of the flume while the properties of the bedrock evolve by changing the percentage of kaolinite. During the experiment, knickpoints cyclically migrate upstream. At the foot of the knickpoint face, bedrock is over-eroded to form a plunging pool. The eroded kaolinite is washed away while the silica is rapidly deposited downstream to form a sediment layer whose slope adjusts to the flow regime. A new knickpoint will not be created until the stream removes this layer and carves into the bedrock, providing a nice example of self-organization along the profile. The rate of knickpoint retreat is rather constant with increasing uplift rate and diminishes with increasing bedrock strength. Rather than increasing knickpoint velocity, the profile adapts to increasing uplift rate by increasing the rate of knickpoint creation.