Rainfall and hence resulting river runoff are the random outcomes of natural processes. In the present research, possibility of predicting average runoff of Kulfo River of Ethiopia has been analyzed through stochastic time series modeling using historical data. In formulating the river runoff predictive model a univariate time series autoregressive integrated moving average (ARIMA) process of time series analysis as suggested by Box and Jenkin is used. In order to make the river runoff data stationary first order differencing is done. ARIMA (1,1,1) model has produced best fit simulation and forecasts of runoff. Statistical measures and error criteria were used to test the validity and performance of the developed ARIMA model. The forecast produced by the model are good enough to be used for the decision making in water supply projects and flood disaster management.
Loading....