Forecasting the amount of water consumed in Erbil City using time : series model (SARIMA)‎

Abstract EN

Forecasting in the time series is one of the essential subjects in statistical science because its indicators helps the government to make a good plan to get an accurate decision in the future.

This research was conducted to analyse the time series data for the amount of water consumed in Erbil City for the years proceeding the period between (1/1/2014) until of (30/6/2019).

In the analysis, the study found non-stationary in the time series, so the necessary transformations were taken on the series to remove the non-stationary.

Thus, the best model was found among the obtained models, SARIMA (0, 1, 2) (1, 1, 1)12 produces the best results based on the minimum statistical criteria (RMSE, MAE, MAPE) used for comparison.

Finally, using the best model to forecast the water monthly average consumed in Erbil City for 12 months (6 months in 2019 and 6 months in 2020) and the results appeared to be consistent with those actual values in the time series.