Environmental impact evaluation and forecast of dust in the cities of Khuzestan province using time series models (during the statistical period of 1990-2010)

Saeed Maleki, Rasol Sarvestan, Mohammad Mansourzadeh


Dust is an atmospheric phenomenon that causes the adverse environmental effects. Problems associated with dust phenomenon more striking are that the frequency of occurrence of this phenomenon in our country has been an upward trend. The aim of this study is to predict and identify the best models for time series in the province during the years 2016-2020. In order to identify and predict the best time-series models over the years 2016-2020, eight selected cities of the province including Masjed Soleiman, Ahvaz, Rāmhormoz, PA, Dezful, Aghajari, Abadan and Omidiyeh are selected. Statistical methods and annual data are performed by using the software 17 Minitab, spss19 and Excel 2013. For this purpose the annual dust data from eight meteorological stations in Khuzestan province during the period 1990-2010 are obtained, and homogeneity test data on dust, are analyzed. Using time-series models dusts are analyzed and then the best model is proposed for prediction. In addition, the accuracy of the model are verified based on statistics, the average absolute error (MAE), mean absolute percentage error (MAPE), root mean square error (RMSE) and Standard (BIAS) approaches. Results indicates that the most appropriate model for the city of Masjed Soleiman and Behban is Holt Winters model, and for the cities of Ahwaz, Rāmhormoz, Aghajari, Abadan and Omidiyeh is a model simple flat development are selected. Furthermore ARIMA model (0,0,0) is proper only for the Dezful city.


dust, time series, Holt Winters, Arima, Anderson-Darling, Khuzestan.

Full Text:



  • There are currently no refbacks.