What have we learned by applying the data driven modelling techniques in the field of hydrological modelling

Well known, praised and widely used are the so-called conceptual models that are based on the prior theoretical knowledge of all the hydrological processes in the form of theoretically developed or empirically derived equations. A modeller needs a lot of detailed data like river network, land cover, soil characteristics and other geographical data or topographical maps to sucessfully calibrate a conceptual model. The availability of all of the aforementioned data can present quite a significant problem in the process of modelling. On the basis of different approaches to the inclusion of geographical, topographical and other data in conceptual models, we distinguish between distributed, semi-distributed and lumped conceptual models. Empirical, mostly black box models simply connecting input and output hydrolgical data have also been widely used in the field of hydrology, especially in the hydrological praxis.