ABSTRACT

ABSTRACT: Physically based distributed models, when properly applied, have the potential of being operational tools that can be useful to water resources planning and management. They can account for the distributed variation of physical properties of the domain being modelled as well as produce spatially distributed predictions. In theory, these models do not need to be calibrated because their parameters may be assessed directly from field measurements. However, in practice, due to the lack of accurate and well distributed data, sub-grid variability of parameters, the uncertainty attached to the definition of initial and boundary conditions and the discrepancies among the scale to which the mathematical structure of the models were derived, the scale to which field data are collected and the scale to which these models are finally applied, this type of models have to be calibrated. In this context, with respect to the application of distributed hydrologic models on a catchment scale, this study presents 5 study cases to briefly address some modelling aspects such as (i) how best to represent hydrological processes; (ii) how best to identify the parameters of distributed models at different scales; (iii) how best to incorporate new sources of data into the modelling process; and (iv) how to assess general uncertainty attached to the application of distributed hydrologic models on a catchment scale. In doing so, different study sites located within Belgium were modelled. Finally, the paper presents some brief thoughts on the future of distributed hydrologic modelling.