Modeling of multilayer cohesive bank erosion with a coupled bank stability and mobile-bed model

Yong G. Lai, Robert E. Thomas, Yavuz Ozeren, Andrew Simon, Blair P. Greimann, Kuowei Wu
2015 Geomorphology  
16 Streambank erosion can be an important form of channel change in unstable alluvial 17 environments. It should be accounted for in geomorphic studies, river restoration, dam removal, 18 and channel maintenance projects. Recently, one-dimensional and two-dimensional flow and 19 mobile-bed numerical models have become useful tools for predicting morphological responses 20 to stream modifications. Most, however, either ignore bank failure mechanisms or implement 21 only simple ad hoc methods. In
more » ... this study, a coupled model is developed that incorporates a 22 complex streams. The developed model is shown to be robust and easy to apply; it may be used 32 as a practical tool to predict bank erosion caused by fluvial and geotechnical processes. 33 Keywords: bank erosion; 2D mobile-bed model; coupled bank model; cohesive bank 34 1977) or capacity (White et al., 1982) , minimum variance (Williams, 1978) , and the principle of 60 least action (Huang and Nanson, 2000) . Although empirical/analytical models are relatively 61 simple to use they are inappropriate for short-and medium-term predictions of unsteady 62 geomorphic response of streams to disturbances (Simon et al., 2007) . 63 Process-based models attempt to explicitly simulate the physical processes that are most 64 important for bank erosion and thus aim to provide reliable short-to medium-term predictions of 65 bank retreat in both stable and unstable channels. The ASCE (1998) provided a review of the 66 models that existed in 1996, Rinaldi and Darby (2008) updated and expanded this review to 67 include finite element seepage modeling, and Rinaldi and Nardi (2013) provided a review on 68 modeling interactions between riverbank hydrology and mass failures. Langendoen and Simon 69 (2008) provided a review focusing primarily on the geotechnical modeling elements and Motta et 70 al. (2012) provided a review of models that linearized and nondimensionalized the two-71 dimensional (2D) mass and momentum equations. 72 Early process-based models assumed that the rate of bank retreat was proportional to the 73 difference (or perturbation) between the depth-averaged near-bank velocity and cross-sectional 74 mean velocity (Hasegawa, 1977; Ikeda et al., 1981) . Osman and Thorne (1988) introduced 75 probably the first process-based model to explicitly consider both lateral basal erosion and mass 76 failure of cohesive sediments. Their method simulated both circular and planar failures for 77
doi:10.1016/j.geomorph.2014.07.017 fatcat:5akcxnmicbffddzbd3ir2ja2sm