Spectral analysis of time series: an application for air pollution in the city of Vitória, Espírito Santo, Brazil

• Autor(a) principal: Santo, Geisiely do Espírito
• Data de Publicação: 2022
• Outros Autores: Pinto, Wanderson de Paula
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Revista de Engenharia Química e Química
• Texto Completo: https://periodicos.ufv.br/jcec/article/view/13892
• topic: Spectral analysis. Time series. Fisher's G test.; Análise espectral. Séries temporais. Teste G de Fisher.

Air pollution is a factor causing an increase in the mortality rate, an increase in the rate of chronic diseases, an increase in the cost of healthcare expenses, a decrease in productivity and quality of life in general. It is worth remembering that the effects caused by air pollution may not only be local, it means that these pollutants can travel thousands of kilometers through the atmosphere, thus reaching communities far from the point of emission. Based on the risks that these pollutants can cause, this study aimed to analyze the time series of concentrations of atmospheric pollutants of PM10, PM2,5 and SO2, monitored at RGV, Espírito Santo, Brazil, using spectral analysis techniques. The data used in this research were measured at the stations of the Automatic Air Quality Monitoring Network (RAMQAr) and collected through the IEMA database. Through the spectral analysis of the time series, it was possible to observe that the concentrations of the pollutants PM10 and PM2,5 presented seasonality with a period of seven days, and for the pollutant SO2, there was no presence of seasonality. Thus, the seasonal variation presented can be explained mainly by the pace of transport movement in the RGV, that is, the highest measured concentrations of PM2,5 and PM10, occur during weekdays, when the flow of vehicles is greater, and the opposite occurs on weekends. As for SO2, as it does not present seasonal variation, it does not have a fixed repetition period.