A non-linear time series prediction method for missing daily flow rate data of Middle Firat Catchment

After the consideration of Climate Change as a serious threat for Water Resource Management, hydrological studies has become to focus on data observation, management and generation. Water Resources data need correct measurement, analysis, and reliable estimates for future planning and current operations for its purposes such as; drinking water, irrigation and energy production. Water Resource Data mining ensure, monitoring Climate change and its further threats. In this study, the daily flow rate data of four different stations on the Murat River were used to generate the data of other fifth station by using Artificial Neural Networks (ANN). Generated data set was tested with MLR method to control its achievement. As ANN are non-linear statistical data modeling tools their achievement for modeling complex relationships between inputs and outputs or to find patterns in data are more successful than statistical methods. Using, non-linear statistical methods will provide many significant benefits to not only to investors during the planning period of run-off river power stations, but also for further studies in Water Resource engineering.