Modeling PM2.5 Concentrations in the Export Processing Area of Dhaka Over a 12-Month Period Using Time Series

Air pollution, particularly PM2.5, poses significant health risks, with Dhaka, Bangladesh, experiencing some of the highest concentrations of fine particulate matter. This study employs 11 time-series models to analyze PM2.5 concentrations in the export processing area of Dhaka. Using a recently published dataset from January 2019 to December 2023, the models were trained to forecast PM2.5 concentrations for a given day of each month over a 12-month period. The results indicated that the Holt-Winters model, followed by a Neural Network and SARIMA, achieved the best performance. An ensemble model incorporating these top three models was developed, demonstrating high forecasting accuracy compared to the other models. This study provides valuable insights for policymakers and businesses, offering an advanced framework for forecasting PM2.5 concentrations to address the ongoing air pollution issue in Dhaka.