Deep Learning for Air Quality Prediction after COVID-19 Pandemic Based on Pollutant and Meteorological Data
14 Pages Posted: 5 Dec 2022
Date Written: July 30, 2022
Abstract
The second wave of COIVD19 has jolted the environment and economy worldwide. Air pollution is one of the primary causes of these pandemics. Therefore, it is important to analyze the air quality index during the COVID19 pandemic. In literature, machine learning (ML) and deep learning (DL) methods have been deployed to predict PM2.5 to forecast air pollution. The central pollution control board (CPCB) of India has gathered information on pollutants such as Particulate Matter(PM) with a diameter of 2.5 microns, called PM2.5, Particulate Matter with a diameter of 10 microns or less, called PM10, Nitrogen dioxide, Sulfur dioxide, Ozone (O3), Carbon Monoxide, Temperature, Relative Humidity (R.H.), Wind Speed (W.S.), Wind Direction(W.D.), and Solar Radiation (S.R.) CPCB has recorded that Bhiwadi, Rajasthan is the world's most polluted city in 2021 PM2.5 Ranking IQAir. In this research, the air quality of Bhiwadi has been analyzed during the COVID19 pandemic based on the above features. The features are considered from three perspectives 1) pollutant features, 2) meteorological features, and 3) Overall features. The analysis has been performed in two-phase, i.e., 1) meteorological + Pollutant and 2) meteorological+PM2.5. ExtraTreesRegressor of ML and LSTM of DL algorithms have achieved the best among other algorithms over the overall dataset (meteorological+pollutant) based on root mean squared error (RMSE) performance measures.
Keywords: Deep learning, Air quality index, Machine learning, PM2.5, COVID19
JEL Classification: C02
Suggested Citation: Suggested Citation