التنبؤء بكميه الماء المستهلكة في مدينة أربيل باستخدام النموذج السلاسل الزمنية (SARIMA)‎

Abstract EN

The purpose of this research is to estimate the best model for forecasting the amount of water consumed in Erbil city using ARIMA time series model.

For this purpose, the series have been tested for stationary and for randomness.

Through analyzing the series, it showed that the series is non-stationary and not random and it also contained the seasonality element.

The suitable transformations have been taken to achieve stationary and randomness which represents taking the natural logarithm, first non-seasonal differencing, and first seasonal differencing.

The RMSE criterion has been used for selecting the optimal model among suggested models.

It showed that the best model of the transformed series is a multiplier seasonal model SARIMA(1,1,1)(3,1,0)12.

The significant test of the model has been taken using the goodness of fit test (Box & Pierce) and then forecasting the series values of the year months (2015 – 2017).