Compositional Principal Component Analysis on Air Concentrations data from Kodungaiyur region of Chennai city

This study applies Compositional Principal Component Analysis (CoDA PCA) to analyze air pollutant concentrations in the Kodungaiyur region of Chennai, focusing on seven key pollutants: PM10, PM2.5, NO2, SO2, NH3, O3, and CO. Compositional transformations—Centered Log-Ratio (CLR), Isometric Log-Ratio (ILR), and Additive Log-Ratio (ALR)—are employed to address the inherent constraints of compositional data, ensuring the sum of the proportions remains constant. The PCA results reveal that the first two principal components explain approximately 76% of the variance, providing valuable insights into the relationships between pollutants and their potential sources. The CLR, ILR, and ALR biplot highlight different pollutant groupings, suggesting distinct emission sources and environmental impacts. This study underscores the importance of using compositional data techniques in air quality research and offers a detailed understanding of pollutant dynamics in Kodungaiyur, aiding in more targeted pollution control strategies.