We propose a data exploration framework for the visualisation and evaluation of hydrological data for the enhancement of numerical simulations. A user-controlled 3D space allows an assessment of large input data sets. Typical examples for inconsistencies between heterogeneous data sets are given and it is demonstrated how these can be detected via a number of modifications done to the scene. Also, the framework allows for the discretisation of the model domain by the generation of finite element meshes based on geo-metric input data. A number of measurements are provided to verify the quality of these meshes and evaluate their adequacy for a subsequent numerical simulation of hydrological processes such as groundwater flow. We demonstrate the application of the framework in two case studies of different scale in central Germany.