ABSTRACT
Models constitute the heart of early warning systems, providing a description of the hazard and its evolution in time. Hydrologic modelling is affected by four main sources of uncertainty: input uncertainty, output uncertainty, structural uncertainty and parametric uncertainty. The suitability of a rainfall-runoff model structure for a certain catchment has been studied through the use of flexible hydrological model structures, which focus on the diagnosis of their differences. The initial parameters for the model were obtained from existing soil, land cover and topographical data of the basin. Model diagnosis is a process by which inferences are made about the representation of hydrological processes through targeted evaluation of the input-state-output response of the model. The precipitation and evapotranspiration data are considered as the main source of uncertainty in the models. The analysis of the discharge ensemble spread in the models shows a higher sensitivity of TOPMODEL to variation in the rainfall.
