Forecasting of Monthly Sum-raining by Stochastic Models in Time Series

Recently, limiting water resources for agricultural and non agricultural usages pose some difficulties and rainfall is the most important water resource. One of rainfall input component can be considered as of hydrological systems. In most cases for studies of groundwater runoff, floods, droughts, sediment, it is necessary and essential to study and measure. Therefore, for optimal allocation of water resources, forecasting rainfall for a region of special importance. One of the methods of evaluate and forecast of precipitation, is the use of time series. For this purpose, there are a variety of methods and models, such as including models of auto regressive (AR), moving average (MA), Auto regressive integrated moving average (ARIMA) and seasonal Auto regressive integrated moving average (SARIMA). In this article, the performance of any of the models listed on the monthly total precipitation amounts and estimates in the hashimabad district of Gorgan during 1983-2012 were studied. Following the elimination of seasonal effects, trend and irregular variations, an SARIMA model was presented. To ascertain the properties of the proposed model, residuals and errors were examined and the model coefficients were estimated. Finally it shows that the model of SARIMA (3,1,2; 0,1,2) time series models have a better performance and less error time series changes to the simulation. To ascertain the properties of the proposed model, residuals and errors were examined and the model coefficients were estimated. Finally it shows that the model of SARIMA (3,1,2; 0,1,2) time series models have a better performance and less error time series changes to the simulation.