Ensemble Modeling of Suspended Sediment in Steep Mountain Catchments

Climatic and land cover changes present important uncertainties into the rates of soil erosion and sedimentation in watersheds. Soil erosion adds constituents to streams, altering water chemistry and streambed morphology, which can adversely affect aquatic life and poses a critical challenge for water treatment and reservoir management. The goal of this research is to establish estimates of sediment transport within large-scale mountainous catchments (>1000 km2). As sedimentation rates are impacted by numerous physical processes including soil, land cover, slope and climate; the results from seven models will be presented to quantify uncertainty and improve predictability. A broader inquiry made here is into the efficacy of model structure under different conditions. We present the results from empirical, stochastic, conceptual and physical models. These include empirical models: monovariate rating curve, multivariate regression and the Modified Universal Soil Loss Equation (MUSLE), to models with conceptual components: Soil Water Assessment Tool (SWAT) to more physically based models: Water Erosion Prediction Project (WEPP), Precipitation Runoff Modeling System (PRMS) and Distributed Hydrology Soil Vegetation Model (DHSVM). Key uncertainties will be characterized resulting from forcing inputs, parameter selection, scale discretization, and model structure. Calibration results from a multi-objective optimization routine will be presented that optimize parameters and identify performance trade-offs that will be used to develop uncertainty estimates in both streamflow and sediment projections. The outcomes of this research will highlight critical issues relevant to large-scale hydrologic and suspended sediment prediction initiatives.