Analysis of the Effect of the Truck Strike and COVID-19 on the concentration of NOx and O3 in the Metropolitan Region of the Vale do Paraiba, São Paulo, Brazil

• Published in: Aerosol and air quality research Vol. 22; no. 8; p. 210364
• Main Authors: Conti, Ludmilla Manera, Herdies, Dirceu Luís, Alvim, Débora Souza, Corrêa, Sergio Machado
• Format: Journal Article
• Language: English
• Published: Taoyuan City Taiwan Association of Aerosol Research 01.08.2022
• Subjects: Algorithms; Coronaviruses; COVID-19; Humidity; Machine learning; Metropolitan areas; Multivariate statistical analysis; Nitrogen dioxide; Nitrogen oxides; Pandemics; Peak hour traffic; Photochemicals; Pollutants; Principal components analysis; Radiation; Relative humidity; Solar radiation; Statistical analysis; Towns; Wind; Wind direction; Wind speed; Brazil
• ISSN: 1680-8584; 2071-1409
• DOI: 10.4209/aaqr.210364

The daily diurnal data pattern of nitrogen oxides (NO, NO2) and ozone (O3), temperature, relative humidity, pressure, wind direction and speed and solar radiation were studied from 2017 to 2020 within a period of 21 days in two towns in Paraiba Valley: São José dos Campos (SJC) and Guaratinguetá (GRT). In 2018, there was a truckers' strike in Brazil and in 2020 a partial lockdown was imposed in response to the coronavirus pandemic; in this study, Machine Learning techniques and a multivariate statistical analysis were conducted to compare these different periods. During both 2018 and 2020, there was a reduction in the NO and NO2 concentrations, (particularly NO), which is a primary pollutant during peak hours of vehicular traffic; this was notably the case in 2018 owing to the truckers´ strike. Through an application of the Tukey test, a comparison was made between the NO, NO2 and O3 data which showed that there was a similarity in each element of the dataset on a decreasing scale, however they continue to be statistically significant. Regarding the Principal Component Analysis (PCA), this procedure identified the first major component for both towns in the entire study period and explained around 42% of the data and the proper interconnections between the data, with a strong positive influence of O3 concentrations, temperature (T), wind speed (WS) and solar radiation (SR). In addition, when analyzing data by means of the Boruta algorithm, there was a considerable difference in the variables that influence O3 concentrations, with GRT showing NO2 and relative humidity, while SJC, NO2 and global solar radiation were the most important variables for feature selection.