Atmospheric Pollutant Levels in Southeast Brazil During COVID-19 Lockdown: Combined Satellite and Ground-based Data Analysis

• Main Author: Cruvinel Brandao Fonseca Marinho, Rayssa
• Other Authors: Environmental Science and Engineering
• Format: Others
• Published: Virginia Tech 2021
• Subjects: air pollution; satellite; surface air quality; COVID-19; lockdown; Brazil
• Online Access: http://hdl.handle.net/10919/102026

With the ongoing COVID-19 pandemic being spread all over the world, lockdown measures are being implemented making air pollution levels go down in several countries. In this context, the air quality changes in the highly populated and trafficked Brazilian states of Sao Paulo (SP) and Rio de Janeiro (RJ) are hereby going to be addressed using a combination of satellite and ground-based data analysis. We explored nitrogen dioxide (NO2) and particulate matter (PM2.5) daily levels for the month of May during different years within 2015-2020. Daily measurements of NO2 column concentrations from the Ozone Monitoring Instrument (OMI) aboard NASA's Aura satellite were also gathered and averaged decreases of 42% and 49.6% were found for the year of 2020 compared to previous averaged 2015-2019 years. In parallel to the NO2 column retrieval, the ground-based data, measured by the Brazilian States Environmental Institutions, is analyzed, and correlated with satellite retrievals. Correlation coefficients between column and ground-based concentrations were 77% and 53% in SP and RJ, respectively. It was found a 13.3% (p-value = 0.099) and 18.8% (p-value = 0.077) decrease in NO2 levels for SP and RJ, respectively, in 2020 compared to 2019. For PM2.5, no significant change was observed for the same time period in the SP region, although the high number of fire burnings in the Southeast region seemed to be affecting PM2.5 levels. In addition to natural emissions (fire burnings), the combined data was also evaluated taking meteorological parameters, such as temperature and wind speed, into account. No interference of weather or fire was found in 2020 NO2 ground levels compared to previous years, This integrated analysis is innovative and has yet to be more explored in Brazilian studies. This is true specifically because the ground-based stations are spatially and temporally sparse in Brazil. === Master of Science === This study aims to explore satellite data applied to the lockdown context resultant from the COVID-19 pandemic in Brazil. Satellite data usage in air quality management is yet to be explored to its full potential. Two highly populated states were chosen: Sao Paulo (SP) and Rio de Janeiro (RJ). Local governments have been imposing limitations on private and public vehicle circulation, inducing a decrease in atmospheric pollutant levels, specifically nitrogen dioxide (NO2), which is directly emitted to the air by fuel combustion. NO2 is also short-lived in the atmosphere, so its variation within days can be easily captured. PM2.5, a category of fine inhalable particles, can be produced by wildfires, in addition to fuel burning and mechanical processes such as resuspension by cars. Here we retrieved daily NO2 vertical column densities for the month of May within the 2015-2020 years from the OMI instrument onboard of NASA's Aura satellite. Ground daily NO2 and PM2.5 measurements were also collected from local environmental agencies. Results showed an average 42% decrease of the NO2 column values in SP in 2020 compared to 2015-2019. The decrease was 49.6% in RJ for the same timeframe. Correspondent surface data showed a decrease of 13.3% (p-value = 0.099) and 18.8% (p-value = 0.077) during 2020 compared to 2019 in SP and RJ stations, respectively. No significant divergence in PM2.5 values was found between 2019 and 2020. Finally, weather data was added to the pollutant analysis. PM2.5 concentrations were associated with wildfires, while the NO2 levels found in 2020 for SP and RJ were attributed to local lockdown decrees. Satellite retrievals showed significant potential in filling out ground datasets, correlating with the SP and RJ surface data in 77% and 53%, respectively.