Behaviour of Daily River Flow: Chaotic?

This study was conducted to provide evidence of the chaotic behavior of the daily river flow data at Lubuk Paku station on the Pahang River in the Pahang River Basin in Malaysia. Four nonlinear dynamic methods are employed: (1) phase space reconstruction; (2) average mutual information algorithm; (3) false nearest neighbors algorithm; and (4) correlation dimension method. First, average mutual information method is used to determine a first minimum of delay time. Second, the sufficient embedding dimension is estimated using the false nearest neighbour algorithm. The time delay and sufficient dimension are used in phase space reconstruction. The presence of chaos in river flow has been analyzed through the correlation dimension method. The correlation dimension is less than 3. Hence, we can conclude that low correlation dimension presence by examined river flow time series data Lubuk Paku station on the Pahang River, Malaysia.