Intelligent Predictive Systems for Modeling PM10, Ozone, and Temperature Dynamics: A Forecasting Approach for Urban Air Quality in Delhi

Abstract

Study investigates the relationship between PM₁₀, ozone and temperature in Delhi's metropolitan areas, a densely populated and industrialized city with significant air pollution problems. Understanding the complex interactions between PM₁₀ and ozone concentrations and climatic variables is the study's main objective. The data included PM₁₀ and ozone concentrations as well as daily readings of the temperature, relative humidity, and other relevant meteorological factors. Regression modelling and correlation analysis is used to examine the relationship between PM₁₀, ozone concentration and temperature. Machine learning approach is used to forecast the Ozone, PM₁₀ and temperature by using the algorithm of auto regressive integrated moving average (ARIMA). Preliminary data indicates a clear relationship between ozone, PM₁₀ concentration and temperature in Delhi's metropolitan areas. The results demonstrate a substantial correlation between these variables, where higher temperatures are linked to higher ozone concentrations and lower PM₁₀ concentrations. Precursor pollutants are implicated in heat-induced photochemical reactions that culminate in the synthesis and deposition of ozone. Study examines the influence of several meteorological factors, such as solar radiation and relative humidity, on the relationship between ozone, PM₁₀, and temperature by observing their past and predicated trends through the years 2018–2026. ARIMA model is used to forecast ozone, temperature, and PM₁₀ levels in Delhi’s urban environment that shows good accuracy. The predictions indicate a significant rise in ozone and PM₁₀ concentrations in the coming years, highlighting the urgent need for effective mitigation measures to maintain air quality within national and international standards.