prediction of precipitation applying wavelet network model(case study: zarringol station, golestan province, iran)

abstract obligatory modelling of precipitations in various periods, have a lot of problems and weakness because of their casual nature in time and space. moreover, their uncertainty in predictions, reduce credibility of estimations which have done via these models. wavelet is one of the novel and very effective methods in analyzing of time series and signals considered in the hydrology in recent years. in this research, precipitation signal has been decomposed via selected mother wavelet, and then the resulted data are used by fitting direct equations to anticipate the precipitation. these mentioned methods are applied in zarringol station in golestan province (iran) for 33 years predict monthly precipitation with 808 mm annually during 1975-76 until 2007-2008. as a result, decomposed signal via wavelet, correlation among observed and calculated data is 84% and the precipitation prediction can be done with more precise. meaningless of f test in 90% and above verifies this phenomenon.