Retrieving spatial variation of aerosol level over urban mixed land surfaces using Landsat imageries: Degree of air pollution in Dhaka Metropolitan Area

Aerosols, or suspended particles in the atmosphere, are critical in the production of cloud nuclei, climate change, poor air quality and reduced ambient visibility, as well as public health and disease. Since there are limited ground monitoring stations in the Dhaka Metropolitan Area (DMA), calculating the degree of Aerosol Optical Depth (AOD) and PM2.5 is thus a challenge. The research aimed to retrieve the AOD at 30 m spatial resolution using the Simplified Aerosol Retrieval Algorithm (SARA) based on Landsat imageries. Landsat 4–5 and 8 cloud-free images at 30 m spatial resolution were obtained from 1999 to 2019 at 5 years interval to retrieve AOD and estimate Particulate Matter 2.5 (PM2.5) concentrations. The results showed significant increase in PM2.5 concentrations and the average PM2.5 were estimated for the year of 1999, 2004, 2009, 2014, 2019 were 48.41, 55.26, 66.42, 72.9 and 90.13 μg/m3, respectively. The AOD, mostly generated from the built-up area was 2.6 in 2019, which was 3 times higher than in 1999. Besides, to find out the relationship among AOD, LST (Land Surface Temperature) and NDVI (Normalized Difference Vegetation Index), results revealed a strong positive correlation of AOD with LST but negative correlation with NDVI. Based on AERONET and NASA SEDAC data, the AOD and PM2.5 were validated using statistically significant (p value 0.05) correlation analysis. By presenting a growing pattern and coverage of air pollution over time, the study will assist urban planners, policymakers, and health sectors.