A Meteorology Based Particulate Matter Prediction Model for Megacity Dhaka

Dhaka, the capital of Bangladesh, is one megacities in world with worst air quality. In this study, we develop statistical models for predicting particulate matter (PM) concentration ambient Dhaka using meteorological and quality data from 2002 to 2004 a continuous monitoring station (CAMS). Model finer fraction PM (PM2.5) explains up 57% variability daily PM2.5 concentration, whereas model coarser (PM2.5-10) 35% its variability, indicating that influenced more by meteorology than PM2.5-10. Temperature, wind speed, direction account 94% total explained model, while relative humidity contributes 75% PM2.5-10 variability. Inclusion lag effect increases models’ predictive power 4–16%. general, our developed show promising performance capturing seasonal Dhaka’s although overestimate low concentrations during wet season (April–September). We validate these recent dataset (2013–2017) same site, which modeled strong positive correlations observed (r = 0.81 0.76 respectively). Models also exhibit forecasting levels two other CAMSs Dhaka. Thus, have potentials explain temporal spatial within These can be helpful policymakers as they predict at any location reasonable accuracy if previous day’s are available. The climate change scenarios on pollution dynamics assessed models.