Influence of Threshold for Bedrock Erosion on River Long Profile Development and Knickzone Retreat in Response to Tectonic Perturbation

Prominent convexities in channel long profiles, or knickzones, are an expected feature of bedrock rivers responding to a change in the rate of base level fall driven by tectonic processes. In response to a change in relative uplift rate, the simple stream power model which is characterized by a slope exponent equal to unity predicts that knickzone retreat velocity is independent of uplift rate and that channel slope and uplift rate are linearly related along the reaches which have re-equilibrated with respect to the new uplift condition (i.e., downstream of the profile convexity). However, a threshold for erosion has been shown to introduce non- linearity between slope and uplift rate when associated with stochastic rainfall variability. We present field data regarding the height and retreat rates of knickzones in rivers upstream of active normal faults in the central Apennines, Italy, where excellent constraints exist on the temporal and spatial history of fault movement. The knickzones developed in response to an independently-constrained increase in fault throw rate 0.75 Ma. Channel characteristics and Shield stress values suggest that these rivers lie close to the detachment-limited end-member but the knickzone retreat velocity (calculated from the time since fault acceleration) has been found to scale systematically with the known fault throw rates, even after accounting for differences in drainage area. In addition, the relationship between measured channel slope and relative uplift rate is non-linear, suggesting that a threshold for erosion might be effective in this setting. We use the Channel-Hillslope Integrated Landscape Development (CHILD) model to quantify the effect of such a threshold on river long profile development and knickzone retreat in response to tectonic perturbation. In particular, we investigate the evolutions of 3 Italian catchments of different size characterized by contrasted degree of tectonic perturbation, using physically realistic threshold values based on sediment grain-size measurements along the studied rivers. We show that the threshold alone cannot account for field observations of the size, position and retreat rate of profile convexities and that other factors neglected by the simple stream power law (e.g. role of sediments) have to be invoked to explain the discrepancy between field observations and modeled topographies.