A probabilistic framework to evaluate the uncertainty of design hydrograph: case study of Swannanoa River watershed

This study presents a probabilistic framework to evaluate the impact of uncertainty of design rainfall depth and temporal pattern as well as antecedent moisture condition (AMC) on design hydrograph attributes - peak, time to peak, duration and volume, as well as falling and rising limb slopes - using an event-based hydrological model in the Swannanoa River watershed in North Carolina, USA. Of the six hydrograph attributes, falling limb slope is the most sensitive to the aforementioned uncertainties, while duration is the least sensitive. In general, the uncertainty of hydrograph attributes decreases in higher recurrence intervals. Our multivariate analysis revealed that in most of the return periods, AMC is the most important driver for peak, duration and volume, while time to peak and falling limb slope are most influenced by rainfall pattern. In higher return periods, the importance of rainfall depth and pattern increases, while the importance of AMC decreases.