Assessing Accuracy of Nitrogen Dioxide Concentration Data from Sentinel-5P Satellite over Bangkok during COVID-19 Lockdown

Abstract

Nitrogen dioxide is a gaseous pollutant which damage the human respiratory system. Generally, pollution data was obtained from the ground-based measurements with high accuracy. However, it is a point measurement which couldn't represent the wide area, especially upcountry area. The satellite measurement become popular choice in recent, but it's accuracy still low. Therefore, this study aims to access and improve accuracy of NO2 concentration data retrieving from Sentinel-5P satellite. Moreover, to compare the performance between the NO2 concentration data which obtained from Sentinel-5P and the ground-based data by analyzing the NO2 concentration changes relating to the COVID-19 lockdown. Results showed that the Sentinel-5P concentration data has R = 0.3614 and RMSE = 0.000898. Then after the accuracy improvement has done, the Sentinel-5P data has higher R (0.6860) and lower RMSE (0.000502), however, there are performed underestimation at high values. Moreover, the analysis of NO2 concentration changes exhibited deference trends of NO2 concentration data which measured by the Sentinel-5P satellite and the ground-based observatory. The results indicated that the data from Sentinel-5P can be used for air pollution monitoring due to it covers for wide area, however, it requires accuracy improvement process.