Noise level estimation of chaotic hydrological time series

A new method of estimating the noise level present in a chaotic hydrological time series is presented. The effectiveness of the method is first demonstrated using two artificial chaotic time series, i.e. the Henon map and the Lorenz equation, whose dynamic characteristics are known a priori, and then tested on two real hydrological time series: daily streamflow series observed in the Chao Phraya River basin in Thailand (raw data), and the same data "cleaned" by the method of Schreiber (1993a). Different levels of noise are added to the artificial chaotic time series in order to demonstrate the effectiveness of the method. A comparison of the results obtained using the proposed method and the method by Schreiber (1993a) clearly indicate a much better performance of the proposed method.