Abstract—Air quality prediction plays a crucial role in addressing urban environmental challenges and supporting sustainable policy interventions. This study presents a comparative modeling analysis of five major air pollutants (PM₁₀, SO₂, NO₂, O₃, and CO) across three Malaysian cities: Nilai, Larkin, and Pasir Gudang. Observed pollutant concentrations were modeled using three approaches: Pearson correlation, Granger causality, and Artificial Neural Networks (ANN), with performance evaluated using Root Mean Square Error (RMSE), Mean Error (ME), Normalized Absolute Error (NAE), and the coefficient of determination (R²). The results revealed that Multiple Linear Regression (MLR), guided by Pearson correlation-based variable selection, consistently provided the highest accuracy (R² > 0.95) across all cities, highlighting the predominance of linear pollutant-meteorological relationships. ANN exhibited moderate-to-strong performance (R² = 0.53–0.74), particularly in Pasir Gudang, where industrial complexity introduced nonlinear emission behaviors. Granger causality contributed supplementary insights into temporal dependencies but yielded lower predictive accuracy. Comparative outcomes demonstrated clear spatial variability, with Nilai and Larkin favoring linear modeling due to stable emission patterns, while Pasir Gudang benefited from ANN’s nonlinear adaptability. This study underscores the importance of combining correlation-driven linear modeling with data-intelligent nonlinear approaches to improve air-quality forecasting reliability in Malaysia. The findings provide valuable guidance for data-driven environmental policy and the design of hybrid predictive frameworks for sustainable urban air management in Southeast Asia.
Keywords—air pollution modeling, air quality prediction, granger causality, Pearson correlation, artificial neural network
Cite: Norazrin Ramli, Zulkifli A. Rais, Norazian M. Noor, Ahmad Z. Ul-Saufie, Hazrul A. Hamid, and Mohd Khairul N. Mahmad, "Comparative Modeling of Air Pollutants in Malaysian Cities Using Correlation, Causality, and Neural Network Approaches," International Journal of Environmental Science and Development vol. 17, no. 4, pp. 348-356, 2026.
Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).