Modeling sediment mobilization using a distributed hydrological model coupled with a bank stability model

In addition to surface erosion, stream bank erosion and failure contributes significant sediment and sediment-bound nutrients to receiving waters during high flow events. However, distributed and mechanistic simulation of stream bank sediment contribution to sediment loads in a watershed has not been achieved. Here we present a full coupling of existing distributed watershed and bank stability models and apply the resulting model to the Mad River in central Vermont. We fully coupled the Bank Stability and Toe Erosion Model (BSTEM) with the Distributed Hydrology Soil Vegetation Model (DHSVM) to allow the simulation of stream bank erosion and potential failure in a spatially explicit environment. We demonstrate the model's ability to simulate the impacts of unstable streams on sediment mobilization and transport within a watershed and discuss the model's capability to simulate watershed sediment loading under climate change. The calibrated model simulates total suspended sediment loads and reproduces variability in suspended sediment concentrations at watershed and subbasin outlets. In addition, characteristics such as land use and road-to-stream ratio of subbasins are shown to impact the relative proportions of sediment mobilized by overland erosion, erosion of roads, and stream bank erosion and failure in the subbasins and watershed. This coupled model will advance mechanistic simulation of suspended sediment mobilization and transport from watersheds, which will be particularly valuable for investigating the potential impacts of climate and land use changes, as well as extreme events.