An Entropy-Based Approach to Station Discontinuance

Abstract

In the design and operation of hydrologic data collection networks, the question of how long data gathering should be continued is a problem that is not often addressed. The current situation is that no definite criteria have been established to decide upon when and where to terminate data collection. Entropy-based measures of information, as used in this study, present convenient and objective means of assessing the status of an existing station with respect to information gathering. Such an assessment can be realized by evaluating the redundancy of information in both the time and the space domains Accordingly, a particular site that either repeats the information provided by neighboring stations or produces the same information by successive measurements can be discontinued. The presented study shows that the entropy concept can be effectively employed to evaluate the spatial and temporal redundancy of information produced by a hydrologic network. The application of the method is demonstrated on case studies comprising water quality and quantity monitoring networks in selected Turkish river basins.