Uncertainty propagation in derived flood frequency analysis driving a rainfall-runoff model with stochastically generated climate input data

For derived flood frequency analysis (DFFA), weather generators are usually linked to rainfall-runoff models. This approach is accompanied by uncertainties from different sources that need to be quantified. In order to find the main sources of uncertainty, possible relations among them and their contribution to the total uncertainty, this study examines in more detail the uncertainties resulting from rainfall-runoff-modelling using stochastically generated input climate data. The conceptual rainfall-runoff model HBV-IWW is driven by stochastically generated climate input data, namely rainfall, temperature and potential evaporation, on an hourly time step for 140 meso- and macroscale (30km² - 1500km²) catchments in Germany. Monte Carlo simulations are carried out varying (a) climate data, (b) model parameters, and (c) climate data and model parameters so as to analyse the uncertainty propagation which might also be linked with catchment conditions.