This study examines changes in air quality during anthropogenic restrictions in Dhaka, Bangladesh, during the COVID-19 lockdown, with an emphasis on variations associated with reduced human activity. The pollutant levels of PM 2.5 , PM 10 , NO 2 , SO 2 , O 3 , and CO were collected and compared before, during, and after lockdown. Significant reductions were observed in NO 2 (−20.77%), PM 10 (−11.47%), SO 2 (−12.85%), and CO (−9.03%), but O 3 increased by 2.35%. PM 2.5 levels exhibited a statistically insignificant decline (−16.44%). To evaluate whether the observed changes deviated from historical patterns, Air Quality Index (AQI) levels for 2020 were forecasted using pre-pandemic data. The forecasts indicated AQI values in the range of 122.24–124.32, compared to the observed value of 107.74 during the same period, suggesting that the improvement in air quality exceeded typical variations. The potential role of meteorological factors was further examined through statistical testing and causality analysis. Although some weather variables showed temporal variation, no significant causal relationships with AQI were identified within the study period, indicating a limited explanatory role relative to other factors. Finally, to determine which policy measures were most strongly associated with changes in air quality, the Random Forest model and ridge regression analysis were employed. The findings consistently identify stay-at-home requirements and the closure of schools and workplaces as the measures most strongly associated with reductions in AQI, with additional contributions from transport and gathering restrictions.
Zaman et al. (2026) studied this question.