Determining the Role of Different Climatic Factors in the Occurrence of Internal and External Dust in Ahvaz Metropolis using Neural-artificial Network

In addition to domestic dust sources, Ahvaz city is affected by most of the foreign sources of dust which is mainly due to locating in the passage of southeast-northwest winds and vice versa. In this study, internal and external dusts were separated to determine the role of climatic factors in the occurrence of dust in Ahvaz. Therefore, the data related to parameters and climatic phenomena affecting the occurrence of dust have been used. Besides, the analysis and comparison of different neural networks have been applied to figure out the importance of each climatic factor in the occurrence of dust. In this study, 70% of the data are entered the network as educational data and 30% as test data. The square sum of educational data error and test data, the relative error of educational data test data and also the correlation coefficient between the measured and estimated values have been compared to evaluate the accuracy of different functions. Finally, the model showing the lowest error rate and the highest correlation coefficient has been selected as the optimal model to investigate the contribution of climatic factors affecting the occurrence of internal and external dust. The results show that wind factor with a speed of more than 6 meters per second and then the wind less than 6 meters per second have played the most important role in the occurrence of internal dust. Moreover, rainfall-changes factor is the third effective one. In the case of foreign dust, since Ahvaz is influenced by the conditions of more remote areas that are almost outside the political borders of Iran, drought factors, SPEI and DTR indices, are considered as the important factors in the occurrence of regional widespread dusts that reach Ahvaz.