A comparison of metric and conceptual approaches in rainfall-runoff modeling and its implications

The aim of the present paper is to compare metric and conceptual approaches to rainfall-runoff modeling in terms of calibration and simulation performances and parameter invariance. This is investigated by applying two models of equal complexity (i.e., possessing the same number of parameters), but with different levels of "conceptualization," to two catchments with different climatology. Level of conceptualization is understood as the degree to which the model structure and its parameters can be related to catchment-scale hydrological processes. The results suggest that the model with less conceptualization provides, in general, a more accurate reproduction of streamflow, even on independent data sets, but this difference only becomes clear when models are applied to the drier catchment. The paper corroborates that the more process complexity one wants to include in the model structure, the more types of data and higher information content are required to estimate the process parameters and to test the model performance. When only rainfall-runoff data are available, it is difficult to justify substantial conceptualization of complex processes.