Long-term spatiotemporal trends in surface and satellite based PM2.5 i n Bangladesh

Abstract

Currently, the air quality in Bangladesh is one of the worst in the world. Pollution due to particulate matter especially PM2.5 is the major driver of air pollution in Bangladesh. It is particularly important to understand the temporal trend of air pollution in the country to assess the change in the extent of pollution in the country. Furthermore, evaluating spatial variability is another key issue while assessing exposure risk. This study focuses on analyzing the temporal and spatial trends of PM2.5 to demonstrate the full spectrum of particulate pollution in the country. In addition to that, where the existing ground monitoring stations provide limited spatial coverage satellite-based Aerosol optical depth (AOD) has the potential to be used as a proxy of PM2.5. Therefore, this study also analyzes the temporal and spatial variability of AOD to reveal the extent of particulate pollution throughout Bangladesh. In this study, the long-term temporal trend is analyzed by adopting Mann-Kendall and Sen’s slope estimation method along with seasonal decomposition with locally estimated scatterplot smoothing. Principal component analysis and hierarchical clustering analysis are conducted to demonstrate spatial variability. The overall trend of the 9 years (2013-2021) data shows an increasing trend for both PM2.5 and AOD in Bangladesh. Dhaka and adjacent regions emerge as pollution epicenters. Darussalam, Dhaka, records the highest rate of PM2.5 increase (2.28 µg/m3/year), while Sylhet displays a significant decline in PM2.5 concentration. Both PM2.5 and AOD data reveal that the north and southeastern parts are less polluted compared to the western part and mid-region. A significant increasing trend in AOD is observed in Rajshahi, Dhaka, and its nearby cities. According to the cross-correlation analysis for AOD, temperature has the strongest association with AOD in all locations. While the association between temperature and AOD is negative in all locations, the highest negative correlation with temperature is observed in Gazipur during winter (0.73). AOD shows significant negative relationships with windspeed, relative humidity, and total rainfall during winter and monsoon. By comprehending the intricate dynamics of PM2.5 and AOD variations and their connections with meteorological factors, the findings of this research may aid policymakers in developing targeted interventions to combat particulate pollution.