This historical review addresses the issues of the evaluation and testing of hydrological models, with a focus on rainfall–runoff models. After a discussion of the general philosophies of hydrological modeling, nine different philosophies of model evaluation are considered, focusing on the period of modeling on digital computers since the 1960s. In addition, some discursions to discuss the definitions of calibration and validation, how much data is needed for model calibration, equifinality and uncertainty, probabilities and possibilities, the evaluation of model ensembles, and model benchmarking. The paper finishes with a final discursion on the philosophical problem of induction.This article is categorized under:
Science of Water > Methods
Science of Water > Hydrological Processes