Abstract :
|
Time series analysis is an established method of generating data, this was used to analyze 31 years of average monthly stream flows of River Kaduna Three models were fitted to the data, viz: MA (Moving average), AR (Autoregressive), and ARIMA (Autoregressive integrated moving average) models .The parameters for the three models were estimated accompanied by diagnostic checks, which showed the model fitness to the data adequately, also the histogram of the residuals revealed that the models assumption was correct for the three models(i.e. residuals are normally distributed), this was further tested by forecasting of 10 years of stream flows at 95% confidence. The results demonstrated the usefulness of the method in generating additional data in watersheds where there is paucity of time series data
|