Comprehensive Modeling of Seasonal Variation of Surface Ozone Over Southern Tropical City, Bengaluru, India

• Published in: Nature environment and pollution technology Vol. 21; no. 3; pp. 1269 - 1277
• Main Authors: Dhanya, G., Pranesha, T. S., Nagaraja, Kamsali, Chate, D. M., Beig, G.
• Format: Journal Article
• Language: English
• Published: Karad Technoscience Publications 01.09.2022
• Subjects: Artificial neural networks; Business metrics; Carbon monoxide; Humidity; Mathematical models; Meteorological parameters; Neural networks; Neurons; Nitric oxide; Nitrogen dioxide; Nitrogen oxides; Numerical analysis; Ozone; ozone, tropospheric pollutant, seasonal variation, meteorological parameters, statistical variations; Performance measurement; Pollutants; Principal components analysis; Radiation; Regression analysis; Regression coefficients; Regression models; Relative humidity; Seasonal variations; Solar radiation; Statistical analysis; Statistical methods; Variables; Weather forecasting; Wind; Wind direction; Wind speed; India
• ISSN: 2395-3454; 0972-6268; 2395-3454
• DOI: 10.46488/NEPT.2022.v21i03.033

Surface ozone (O3) is an important pollutant. In this study we investigated the effects of precursor gases on the difference in ozone concentration utilizing various statistical methods like Multiple Linear Regression (MLR), Principal Component Regression (PCR), Artificial Neural Network (ANN), and Principal Component and Artificial Neural Network (PC-ANN) in conjunction with meteorological parameters for forecasting. The pollutants ozone (O3), carbon monoxide (CO), nitric oxide (NO), nitrogen dioxide (NO2), oxides of nitrogen (NOx), and the meteorological parameters temperature (temp), relative humidity (RH), solar radiation (SR), wind speed (WS) and wind direction (WD) observed during 2019 are taken as inputs for MLR, PCR, ANN, and PCANN. The mathematical models obtained from the numerical analysis showed that O3 concentration was significantly affected by the CO, NO, NO2, NOX, temp, RH, SR, WS, and WD factors. PCR model’s regression coefficient was lower than the MLR model, but the same for ANN and PCANN models was much better in all the seasons than the linear models such as MLR and PCR. The efficiency of all methods is inspected using several performance metrics.