Dust Level Forecasting and Its Interaction with Gaseous Pollutants Using Artificial Neural Network: A Case Study for Kermanshah, Iran

Abstract

An artificial neural network (ANN) was used to forecast natural airborne dust as well as five gaseous air pollutants concentration by using a combination of daily mean meteorological measurements and dust storm occurrence at a regulatory monitoring site in Kermanshah, Iran for the period of 2007-2011. We used local meteorological measurementsand air quality data collected from three previous days as independent variables and the daily pollutants records as the dependent variables (response). Neural networks could be used to develop rapid air quality warning systems based on a network of automated monitoring stations. Robustness of constructed ANN acknowledged and the effects of variation of input parameters were investigated. As a result, dust had a decreasing impact on the gaseous pollutants level. The prediction tests showed that the ANN models used in this study have the high potential of forecasting dust storm occurrence in the region studied by using conventional meteorological variables.