Abstract:
Amid the COVID-19 crisis, governments all over the world, and not
excluding India, took to lockdown measures to deaccelerate the
spread of the COVID-19 virus. This led to reduction of atmospheric
pollution by declining the harmful Nitrogen and Sulphur Oxide
(NOX and SOX) concentrations. However, one hand while the
stratospheric Ozone (O3) showed repair, the lower atmospheric O3
concentrations demonstrated a remarkable increase during lockdown phase over India. This study aims to estimate the O3 concentration during the Covid-19 lockdown over Pune city in India
using freely available Sentinel-5P satellite datasets. The study
makes use of the Ordinary Least Squares (OLS) and Random
Forest (RF) regressions and compares the findings of the two algorithms based on estimation results. This study utilizes lower atmospheric O3 concentration data from Sentinel-5P satellite of the
European Space Agency (ESA) over the Indian mainland for a
month of lockdown scenario (March 22nd, 2020, to April 25th,
2020) and shows the remarkable increase in concentration of O3
gas as a pollutant. Despite the complete lockdown over India during this given time frame, there has been enough emission of O3
precursors from other sources such as stubble burning. The estimates of tropospheric O3 concentration for May 2020 for Pune
city, using OLS and RF Regressions, have been validated with May
2020 data. The results have provided a RMSE of 1.05 and 1.23
with R2
-statistics of 0.90 and 0.857 in training and testing phases
for OLS and RMSE of 0.98 and MAE of 1.07 with R2
-statistics of
0.968 and 0.895 in training and testing phases of the RF. The outcome of this study has proven that O3 gas concentrations in the
atmosphere depends upon various other causative factors apart
from the precursor gases. The study also shows that the remotely
sensed Sentinel-5P datasets, supplemented with ground-based
sensor data can help in time and cost saving estimation of O3
concentrations in the troposphere with considerable accuracy.
