Introducing and Application of Combined BL-ARCH Model for Daily River flow Forecasting (Case study: Shahar-Chai River)

One of the applicable ways for simulation and forecasting hydrological processes is time series modeling. An important problem in forecasting hydrological data using time series is generating stochastic data. Any changes in stochastic series will change generating data. In this study nonlinear ARCH model presented in order to modeling and generating stochastic component of time series. After combing ARCH model with nonlinear bilinear model، BL-ARCH model suggested to forecasting river flow discharge. Daily river flow of Shahar-Chai River located in the west of Urmia Lake and West Azarbaijan province have been used for data analysis and 11 years forecasting. As results shown suggested model with 4. 52 error has better than bilinear model with 6. 77 error. So this model can be used for short-time river flow forecasting specially daily series.