عنوان مقاله [English]
نویسندگان [English]چکیده [English]
River flow behavior is one of the major problems considered in the design, operation and studies related to water resources. Recently, applying new techniques such as chaos theory in hydrology and water resources due to its innovation and ability has been considered. An application of chaos theory is characterization and quantitative analysis of river flow time series. The aim of this study is determining properties and prediction of Ahar-chai river daily flow during 19 years using the concepts of chaos theory. Chaos time series state space reconstruction is based on selecting two suitable parameters namely delay time and embedding dimension. In this study, average mutual information and false nearest neighbors method are implemented for estimating these two parameters. Correlation dimension method is used for determining the chaos ability and estimating the fractal dimension of daily river flow. The results of computations show that fractal dimension of 4 (chaotic low), with delay time of 65 days and embedding dimension of 13 for reconstruction the river flow dynamic state space, can be used. Local prediction algorithm for prediction of time series and correlation coefficient, and also Nash-Sutcliffe index and root mean square error are used for assessing the performance. Results illustrate good and acceptable accuracy of chaos theory in prediction of river flow.