Backing a model – how to establish predictions from limited observations in water sciences

This lecture will provide an introduction into simulation model development in water sciences. Backing a Model will focus on the (limited) availability of hydrological observations that strongly depends on the spatial scale to which a model is meant to be applied. By presenting two case studies, one at the scale of several m² and the other at the scale of entire Europe, I will elaborate the decisions and simplifications that were necessary to develop a prediction model from the available data at the respective scale. Strength and weaknesses of the selected approaches at each scale will be discussed and some ideas for improving future model development in water sciences will be provided.