A Hydraulic Model Wrapper to Optimize Floodplain Reconnection Scenarios

Floodplain reconnection projects are nature-based solutions that have great potential to restore channel stability, attenuate stormwater, improve flood resilience, and improve water quality. Floodplain reconnection practices can include berm lowering or removal, floodplain benching, strategic placement of culverts and bridges, and planting and maintaining buffers. However, the suite of possible restoration options for a given study area must consider the multiple uses and functions of river corridors, along with the potential impacts and benefits to adjacent infrastructure. Stakeholders are in need of decision-support frameworks to help identify a combination of restoration projects that will best meet these multiple and sometimes competing objectives and minimize flooding impacts under an overall constraint of limited financial resources. We have created a tool using Python that can be wrapped around a two-dimensional hydraulic model (HEC-RAS) to guide selection of optimal suites of floodplain reconnection scenarios. We apply the wrapper to assess impacts of restoration alternatives on flood velocity, depth and duration for test rivers in Vermont. An exhaustive search was performed on seven floodplain reconnection scenarios for the Black Creek to first validate the optimization algorithm. We are then applying this tool to perform a multi-objective optimization on the Mad River to analyze a larger search space of possible restoration tools, and increase the combinatorial problem. Our optimization wrapper is transferable to other sites with existing HEC-RAS models to help stakeholders and managers design and prioritize an optimal suite of river restoration practices.