Evaluation of Atmospheric Pollutants Using Satellite Derived and Low-Cost Sensor Data Over Accra, Ghana.

Abstract

It is well established that high pollution concentrations negatively impact the climate and public health. According to the WHO, poor air quality is responsible for approximately 7 million deaths worldwide each year. With rising urban emissions and population growth, the situation is similar in sub-Saharan Africa, where poor air quality is increasing alarmingly. Therefore, regular monitoring is needed to assess the levels of pollutants at both local and regional scales. However, there needs to be more monitoring in sub Saharan Africa since purchasing, installing, and maintaining many high-grade air quality monitoring sensors is expensive. Hence, the gap in studies focusing on associations between particulates with aerodynamic diameter less than 2.5 microns (PM2.5) and gas pollutants like nitrogen dioxide (NO2) and ozone (O3) in sub-Saharan African cities. Furthermore, this study seeks to bridge this gap by utilizing data from five Clarity Node-S PM2.5 sensors, total column particulates or Aerosol Optical Depth (AOD), NO2 and O3 data from satellites, and data from five different Ghana Environmental Protection Agency (GEPA) air quality traffic sites in the Greater Accra Metropolitan Area (GAMA). The AOD, NO2 and O3 data were retrieved from Moderate Resolution Imaging Spectro-Radiometer (MODIS) Terra and Ozone Monitoring Instrument (OMI). The long-term trends over the 5 stations on (25 × 25) 𝑘𝑚2 grid for OMI and (50 × 50) 𝑘𝑚2 grid for MODIS Terra AOD from 2012 to 2021 were assessed using Mann-Kendall sequential test while the Pearson correlation coefficient was used to find correlations between the pollutants. In addition, characterization of PM2.5, AOD, NO2, and O3 levels in the GAMA was evaluated. Overall, there was an increasing trend in NO2 (with p < 0.05) over 4 stations, no trend in O3 (with p > 0.05) and a decreasing trend in AOD (with p < 0.01). Coefficients of determination between PM2.5 data and MODIS Terra AOD on (50 × 50) 𝑘𝑚2 grid across the stations were (R2 = 0.72, 0.72, 0.67, 0.58 and 0.57) respectively. Coefficients of determination between total column NO2 and O3 were (R2 = 0.83 ±0.030, p < 0.01), AOD and O3 (R2 = 0.43 ± 0.003, p < 0.01) and, NO2 and AOD (R2 = 0.21 ± 0.010, p > 0.01). PM2.5, AOD and NO2 levels were generally high during the dry season while high concentrations of O3 were observed in the wet season across the stations. Moreover, the PM2.5 daily mean level of 32.8 μgm-3 for 25 months between 2018 and 2021 was more than twice World Health Organization (WHO) recommended daily mean level of 15 μgm-3. A high correlation coefficient was observed between total column NO2 and O3. Low correlation coefficients between AOD, NO2 and O3 may reveal different emission sources from open burning, street cooking, traffic, and industrial activities in the GAMA. High population growth coupled with increasing traffic, biomass burning, and climate change in growing sub Saharan African cities requires urgent policy measures and regulations as ground air quality monitoring sensors are limited.