Coupling Risk-Informed Design and Stochastic Erosion Modeling To Reduce Habitat Impacts

Flood Risk Management projects have historically sacrificed riparian and upland habitat to ensure robust protection from highly erosive, discrete flow events. In most circumstances, designers are forced to place revetment (usually in the form of riprap) high in the flow cross section to provide protection against erosive velocities and shear stresses associated with the design hydraulic loading of the system. The development of the Risk-Informed Design concept and the use of Semi-Quantitative Risk Assessments (SQRA) has allowed design teams within the USACE to shift the focus of the design away from discrete flow events and towards a quantification of the risk. This approach provides design teams with a means for further optimization of the revetment extents and subsequent habitat impacts. However, the hydraulic tools available to quantify this risk have largely remained unchanged, resulting in the continued use of conservative designs.

The design team for the Lower American River Contract 3 project in Sacramento, California was forced to reconcile this issue during the design of the project’s erosion control. The Lower American River is a designated Wild and Scenic River with valuable riparian/upland habitat nestled within a high-risk levee system. The design team was tasked with meeting the project’s risk objectives through a SQRA while limiting the removal of existing vegetation. The quantification of lateral retreat during the design event is generally not used as a design tool for flood risk management projects due to the dynamic and uncertain nature of erosion. To address this issue, the Bank Stability and Toe Erosion Model (BSTEM) program developed by the USDA was utilized to produce stochastic estimates of bank erosion and create an opportunity to reduce the extents of the required riprap. The design team was able to iterate the horizontal and vertical revetment extents using BSTEM by allowing some bank deformation for the design flow events but keeping the bank erosion risk at an acceptable level. Using fitted distributions for several erosion input parameters, the Monte Carlo analysis produced a range of potential lateral retreat estimates that were compared to the proximity of the levee prism during the SQRA process. This allowed the SQRA team to assess the system risk with the optimized design and incorporate natural bank deformation in the flood risk management project. This approach helped to reduce the total acres of habitat impacts and required mitigation. The approach identified above aligns closely with the goals of the USACE Engineering with Nature initiative to incorporate nature-based processes to achieve design goals while minimizing environmental footprints.

In this paper, the authors will present the iterative design process developed for this project with the novel coupling of BSTEM and the SQRA process in a high-risk environment to reduce habitat impacts. Challenges and lessons learned will be presented along with suggestions for active communication of project risk quantification with resource agencies, flood control districts, and citizen groups.