Paper Title
PREDICTION AND ANALYSIS OF PM LEVELS IN DELHI AND THEIR RELATIONSHIP WITH METEOROLOGICAL FACTORS AND GASEOUS POLLUTANTS

Abstract
Abstract - In this study, we used concentration and meteorological data from the year 2018 to 2023 for 5 different sites. This study investigates the impact of air pollution, especially particulate matter (PM), on the environment and human health in Delhi, one of the most polluted cities in the world. The study analyzes the sources and levels of various gaseous pollutants and meteorological factors that affect PM levels at different sites in Delhi by analyzingthe varying characteristics of PM concentration and the relationships between the different characteristics and meteorological variables. In see of time, with the assistance of diverse machine learning calculations the outcomes appeared that the PM concentration showed maximum increase with increase in Nitrogen Dioxide levels and Carbon Monoxide levels in terms of Gaseous pollutants while Atmospheric Temperature, Solar Radiance and wind speed had maximum negative correlation with PM levels among the meteorological factors chosen. The study also compares three machine learning models for predicting PM 2.5 and PM 10 levels: Multiple Linear Regression (MLR), Random Forest (RF) and Xgboost (XGB). The results show that MLR performs better than RF and XGB in terms of accuracy and error metrics. The correlation with meteorological factors from high to low was the atmospheric temperature>solar radiance>wind speed>wind direction>relative humidity>rainfall. Amidst the above meteorological factors rainfall, temperature, wind speed and humidity are negatively correlated with PM concentration, but the relationship between relative humidity and PM2.5 concentration is a positive correlation whereas with PM10, it shows negative correlation.