Incorporation of rating curve uncertainty in dynamic identifiability analysis and model structure evaluation

Abstract. When applying hydrological models, different sources of uncertainty are present and the incorporation of these uncertainties in evaluations of model performance are needed to assess model outcomes correctly. Nevertheless, uncertainty in the discharge observations complicate the model identification, both in terms of model structure and parameterization. In this paper, two different lumped model structures (PDM and NAM) are compared taking into account the uncertainty coming from the rating curve. The derived uncertainty bounds of the observations are used to derive limits of acceptance for the model simulations. The DYNamic Identifiability Approach (DYNIA) is applied to identify structural failure of both models and to evaluate the configuration of their structures. The analysis focuses on different parts of the hydrograph and evaluates the seasonal performance. In general, similar model performance is observed. However, the model structures tend to behave differently in function of the time. Based on the analyses we did, the probability based soil storage representation of the PDM model outperformed the NAM structure. The incorporation of the observation error did not prevent the DYNIA analysis to identify potential model structural deficiencies that are limiting the representation of the seasonal variation.