COVID-19 outbreak and air quality

Abstract

This work compares variations in the concentrations and air quality indexes of the pollutants PM10, PM2.5, CO, and NO2, during the COVID-19 outbreak in São Paulo Municipality. Such comparisons considered the period of physical distancing (autumn) and the three first months of economic activities’ resumption (winter). The concentrations and indexes observed in 2020 were compared with their correspondent values measured in the three preceding years. Also, from a supervised machine learning algorithm, the correspondent 2020 expected values were predicted and used in these comparisons. In order to analyze the influence of meteorological conditions, the seasons studied were characterized using remote sensing and surface data. The pollutants predominantly emitted by the vehicle fleet (CO and NO2) had reductions in their concentrations, with values always below the predictions and good air quality indexes. However, the pollutants whose concentration is less dependent on automotive emissions (PM10 and PM2.5) had high proximity to the predictions during the autumn and lower values during some periods in winter. This reduction was not enough to avoid days with a moderate air quality index. The approximation of the average concentrations of PM10, PM2.5, and CO to the prediction, from the second-half August 2020, indicates the influence of activities’ resumption in air quality.