Wavelets, functions with zero mean and finite variance, have recently been found to be appropriate tools in investigating geophysical, hydrological, meteorological, and environmental processes. In this study, a wavelet-based modeling technique is presented for suspended sediment discharge time series. The model generates synthetic series statistically similar to the observed data. In the model in which the Haar wavelet is used, the available data are decomposed into detail functions. By choosing randomly from among the detail functions, synthetic suspended sediment discharge series are composed. Results are compared with those obtained from a moving-average process fitted to the data set.
Precise Asymptotics in the Law of Iterated Logarithm for Moving Average Process under Dependence