Bank Erosion By Wind-Generated Waves Ii: Application of A Wind-Wave Sub-Model In Bstem-Dynamic

The USDA-ARS Bank-Stability and Toe-Erosion model has become a widely used tool since its initial development over 20 years ago. A somewhat simplified version of the Static model (v. 5.4) has been incorporated into HEC-RAS. The Dynamic version, which handles a varying-flow series has been continually developed and enhanced to include new features such as variable roughness (and effective stress) by layer, choice of a variable energy slope by time-step and accounting for shear stress from boat-generated waves. Some or all of these features have been successfully deployed by the authors in recent studies in the US, Australia and New Zealand to evaluate the relative impacts on bank erosion from dam operations and releases, high flows, roughness from established vegetation and boat-generated waves. In response to the need to provide reliable analysis of the causes of bank-retreat and effective mitigation strategies in wide, impounded rivers and embayments, both a wind-prediction algorithm (WPA) and a wind-wave erosion module were developed and added to the BSTEM-Dynamic code (v. 2.4). The enhanced model was applied to two studies where at least one of the objectives was to evaluate the relative roles of various bank-erosion processes on rates of erosion and bank retreat: A 29-km reach of the Tennessee River between Pickwick Dam and Savannah, TN and a 1-km reach of the Fox River at De Pere, WI, in backwater from Lake Michigan. BSTEM-Dynamic was calibrated at 21 sites along the Tennessee River over the 1985 to 2016 period using the retreat measured from the aerial imagery. One cross-section was used for calibration on the Fox River for the period of 1990 to 2020.

At least two runs were conducted initially at each site, with and without waves, to determine the role of wind-generated waves on erosion. Results from the two studies provided very different outcomes. In the reach of the Fox River which can be characterized as a “drowned channel” with a beach-like bank toe and low bank heights, wind-generated waves accounted for 77% to ~100% of the bank erosion that ranged from 0.75 to 7.5 ft3/ft/yr. Water-surface elevations (wse) vary only within a ~7-ft range, thereby concentrating wave action at the base of the bank face above the beach. In contrast, on the study reach of the Tennessee River where wse varied over a much larger 43-ft range, wind-generated waves accounted for <1% of the bank erosion at the vast majority of sites, notwithstanding longitudinal zones of turbid water that were observed along bank toes on windy days. It is hypothesized that because of fluctuating water levels and the relatively resistant nature of the surficial bank materials, the cumulative impacts of waves were minimized. Average-annual rates of bank erosion were 56.4 ft3/ft/yr and were largely attributed due to high flows in this incised channel.

The BSTEM-Dynamic model that includes the wind-wave sub-model has proven effective in evaluating the roles of a range of processes that impact streambanks. Application to real-world issues has also shown its utility in determining effective of mitigation alternatives.