shobe_tucker_sensitivity.pdf (14.74 MB)
When hillslope-derived blocks alter river evolution: A sensitivity analysis
This is a poster presented on December 15th, 2016 at the American Geophysical Union annual meeting in San Francisco, California, USA. It presents initial results for sensitivity analysis of a numerical model of river incision in the presence of hillslope-derived blocks of rock (Shobe et al., 2016; Geophysical Research Letters).
Abstract:
Many studies rely on simple models of river channel evolution for predicting channel response to perturbations and landscape evolution outcomes, as well as extracting climatic and tectonic signals from river longitudinal profiles. However, recent work has shown that large, hillslope-derived blocks delivered to rivers in response to rapid incision may noticeably alter the form and evolution of river profiles from the behavior predicted by the most common models. We use a 1-D model of fluvial erosion and block dynamics to investigate the conditions under which this channel-hillslope feedback influences river erosion and landscape evolution. In the model, block delivery by hillslopes occurs in response to channel bed lowering, and blocks inhibit further erosion by covering the bed and reducing effective bed shear stress. Using formal sensitivity analysis techniques including Morris one-at-a-time screening and variance-based decomposition, we seek to explore the full model parameter space and determine which landscape parameters set the influence of blocks on river incision. We use eight response functions to evaluate model behavior, including the mean, standard deviation, and range of the distributions of elevation and slope in the model domain. Preliminary results indicate that baselevel lowering rate, bed erodibility, initial block size, and a parameter governing the connection between river erosion and volume flux of blocks to the channel control whether or not channel evolution will be influenced by hillslope-derived blocks. Surprisingly, block erodibility and the delay timescale separating channel incision from hillslope response have relatively little influence on model outcomes. Results also show that parameter sensitivity is strongly dependent on the chosen response function, highlighting the need for multiple model metrics. Identification of the most important model parameters allows confident identification of landscapes strongly influenced by block delivery to rivers, where special care may be needed when inverting river profiles for tectonic and climatic signals. Our results support the need for comprehensive model sensitivity analysis in geomorphology, which has the potential to reveal model dynamics that may be missed by simple “brute force” parameter space exploration.
