Evaluation and Selection of Streamflow Network Stations Using Entropy Methods


Abstract: The entropy methods provide reliable results in evaluating the performance of existing stream-gauging networks. Two entropy methods used to design network problems for selecting the priority stations in the multivariate case are presented in this study. One of these entropy methods is based on the combination of stations with the least transinformation, which was developed with normal and log-normal distributions. The other method, based on ranking stations, is applied for normal, log-normal, and gamma distributions. The aim of this study is to investigate the effect of distribution types on these entropy methods. For this reason, these entropy methods were applied under different distributions for the annual observations of 5 runoff stations in the Kızılırmak Basin. It was found that the stations selected to continue observations on the existing stream gauging network were not the same for different distribution types. This indicates that the distribution type for streamflow data is very important for these methods.

Keywords: Entropy method, Normal distribution, Log-normal distribution, Gamma distribution

Full Text: PDF