Air Pollutants and Meteorological Parameters Influence on PM2.5 Forecasting and Performance Assessment of the Developed Artificial Intelligence-Based Forecasting Model

Particulate matter with an aerodynamic diameter lower than 2.5 mu m (PM 2.5) is one of the most important air pollutants. Current regulations impose measuring and limiting its concentrations. Thus, it is necessary to develop forecasting models programs that can inform the population about possible pollution episodes. This paper emphasizes the correlations between PM2.5 and other pollutants, and meteorological parameters. From these, nitrogen dioxide and temperature showed have the best correlations with PM2.5 and have been selected as inputs for the proposed forecasting model besides four PM2.5 concentrations (the values from current hour to three hours ago), the output of the model being the prediction of the next hour PM2.5 concentration. Two methods from artificial intelligence were used to build the forecasting model, namely adaptive neuro-fuzzy inference system (ANFIS) and artificial neural networks (ANN). The comparative study between these methods showed that the model which uses ANN have better results in terms of statistical indicators and computational effort.