STATISTICAL ANALYSIS OF OZONE POLLUTION IN DELHI: BEFORE AND AFTER LOCKDOWN

Abstract

Ozone (O3) is technically known as, a Greenhouse Gas, has its own importance or harmful depending on where it is found in the earth's atmosphere. The present work mainly focuses on the study of concentration variation levels in O3 over a 5–year period (2015–2020), using the Multiple Imputation by Chained Equations (MICE) and Time Series analysis on multivariate data. Machine learning models based on SARIMA is built to for analysis of Ozone pollution especially in Delhi. Data were obtained from National Pollution Board of India which was decomposed into seasonal, reside and trend components. Through the time–series analysis of O3 in Delhi, the results showed values above average during the cold seasons. The study analysis the effects of O3 on the health during the 5-years and prediction of future Ozone pollution level. The performance evaluation of the prediction model is done by calculating mean square error (MSE), root mean square error (RMSE), and mean absolute error, may help control the degraded ozone quality. The SARIMA model for O3 gave 20.41 RMSE values for Delhi for the year 2021 and tested through one-way ANOVA hypothesis and got a positive result.