ABSTRACT
A general and important issue in sediment transport investigations is how to deal with limited data. Simulation results of 2D morphodynamic models depend on a variety of parameters as well as on initial conditions and inputs provided. Different parameter settings are possible when calibrating the model output. In this context the sensitivity of a numerical 2D bed load transport model was tested with regard to input uncertainties. Model tests were conducted using different approaches to calculate input sedigraphs based on multi-annual hydrographs and characteristic parameters of reaches located upstream. The common boundary condition of total sediment load was estimated. It was shown that the model is sensitive to the composition of sediment mixtures as well as to the temporal distribution of sediment input. The ability to calibrate the model based on an uncertain input is discussed. Finally, the sensitivity of the model to input variations seems to be less compared to the uncertainty in estimating the total annual loads on catchment scale.
