Evaluating model accuracy: Predicted and observed sediment deposition following the Chiloquin Dam removal

Hydrodynamic and sediment transport models are often used to reduce uncertainty regarding the outcomes of dam removal, though the accuracy of these models is not regularly evaluated post-removal. With the goal of improving our understanding on the accuracy and limitations of making predictions about sediment dynamics following dam removal, we compare predicted and observed sediment deposition at sites downstream of the Chiloquin Dam removal. Results from 1D hydraulic and sediment transport modeling are compared to cross-sectional and longitudinal profile and sediment surveys collected one year following the dam removal. Model predictions provide estimates of mass and thickness for different grain sizes of sediment remaining in downstream channel features following one year of winter flows. Pre-removal and post-removal bathymetry and grain size distributions document the changes in depth, size, and mass of sediment deposited downstream. Field observations are coupled with continuous TSS monitoring, resulting in a sediment budget for the reservoir and downstream reach. The relationships between the model predictions and observations are presented, model uncertainty and observation error is characterized, and approaches for improving model predictions are discussed.