ABSTRACT
Streamflow time series are of utmost importance for a catchment-hydrological analysis and in general to water resources management. Nonetheless, continuous or reliable data are not always available everywhere. Pattern characterization as well as the estimation of incomplete data in time series are often a challenge for many scientists because they are not easy tasks in engineering. Specially in the early stages of water management, a sound characterization of river discharges series and their seasonal variations is extremely important (Black, 1996; Simonovic, 2009).
