The use of maximum entropy to increase the informational content of hydrological networks by additional gauges

ABSTRACT

The optimization and extension of existing gauging networks are a challenging task, which can be done under consideration of many different aspects. One possibility is to maximize the obtained information on regional hydrological characteristics by new gauges compared to existing ones. For this, information theory approaches are most suitable. Here, the principle of maximum entropy is applied to calculate the probability of non-similarity of catchments to determine locations of new gauges according to the catchment characteristics that are most relevant for the hydrological conditions. The realization in an interactive application, provided online, makes use easy for practitioners and scientists. Goodness-of-fit measures are applied to investigate the explanatory power of the model and the contribution of each characteristic to the model, which gives information on the most influential properties of the catchment. The relevance of the proposed approach is proven by comparing hydrological signatures between similar and non-similar catchment.