Meteorological Adjustment and Regression Modeling of Air Pollutants in Dhaka, Bangladesh: Insights from Pre-, During-, and Post-Lockdown Periods (2017–2023)

This study provides a comprehensive assessment of air quality variability in Dhaka, Bangladesh, from 2017 to 2023, with emphasis on the pre-lockdown, lockdown, and post-lockdown phases of the COVID-19 pandemic. Long-term records of major pollutants (PM\((_{2.5})\), PM\((_{10})\), NO\((_2)\), CO, and O\((_3)\)) from the Department of Environment (DoE) were integrated with meteorological data from the Bangladesh Meteorological Department (BMD) to evaluate the combined impacts of anthropogenic activity and atmospheric conditions. Descriptive statistics, trend analysis, and multiple linear regression (MLR) models were applied to characterize pollutant dynamics. The lockdown period exhibited substantial reductions in NO\((_2)\) (50.26%), PM\((_{2.5})\) (35.04%), PM\((_{10})\) (27.99%), and CO (50.23%), attributed primarily to decreased vehicular, industrial, and construction emissions. Conversely, O\((_3)\) increased by 14.55% due to weakened NO titration and enhanced photochemistry. Post-lockdown increases across pollutants indicated that air-quality improvements were temporary. Seasonal analysis identified winter as the most polluted period, with particulate concentrations nearly three times higher than during the monsoon. Regression models showed strong meteorological influence, with temperature exerting pronounced negative effects on PM\((_{2.5})\) (\((\beta = -9.181)\)) and PM\((_{10})\) (\((\beta = -10.415)\)). The strengthened NO\((_2)\)--traffic correlation post-lockdown (\((r = 0.664)\)) underscores vehicular emissions as a dominant source. These findings highlight the need for meteorologically informed, data-driven strategies to manage air quality in rapidly urbanizing megacities.