Enhanced Bank-Stability Modeling With Coupled Geotechnical, Hydraulic and Near-Bank Groundwater Sub-Models: Development and Validation

Physically-based, deterministic bank-stability models have recently been developed to effectively simulate the driving and resisting forces governing streambank erosion. Significant advances have been made in the manner in which groundwater flow through variably saturated porous media, planar and circular geotechnical failures and fluvial sediment transport are simulated. However, to date, coupling these models has required tedious exporting and conversion of geometries and results and manual remeshing. In this presentation, we introduce the first fully integrated suite of models that deterministically simulate the controlling hydrologic, hydraulic and geotechnical processes that govern streambank erosion and channel-width adjustment. The model suite incorporates routines that: 1. Permit the user to enter between 5 and 23 points to describe the bank cross-sectional geometry; 2. Automatically generate a mesh by which to implicitly discretize the 2-D Richards equation utilizing finite volumes. The resulting pentadiagonal matrix is solved iteratively with Stone's Strongly Implicit Procedure (SIP). Timesteps are automatically adjusted to minimize mass balance and truncation errors; 3. Evaluate the force-equilibrium factor of safety (Fs), permitting the simulation of planar and cantilever shear failures with a horizontal slice method and planar shear failures with tension cracks with a rigorous vertical slice method. A random walk approach is adopted to search for the minimum Fs; 4. Estimate the increase in cohesion due to vegetation with a global load-sharing Fibre Bundle Model; and 5. Simulate the erosion of the bank face and bank toe with an excess shear stress approach. Management options to increase slope stability (through the addition of vegetation) and reduce channel- boundary erodibility (through the addition of natural and artificial structures) are also incorporated. We illustrate the efficacy of the modeling approach with a series of case studies in which the components are validated both separately and in unison. Case studies are presented including those from previously published lysimeter experiments and an actively eroding meander bend on Goodwin Creek, Mississippi. The examples illustrate the use of the model suite to: 1. Understand the rates and importance of geotechnical and hydraulic processes in streambank erosion; 2. Estimate the absolute and relative contribution of geotechnical failures and hydraulic scour to volumetric fluvial sediment loads; and 3. Estimate the impact of potential streambank-erosion management and river restoration strategies.