Factors influencing spatiotemporal variability of NO2 concentration in urban area: a GIS and remote sensing–based approach

Abstract

The growing global attention on urban air quality underscores the need to understand the spatiotemporal dynamics of nitrogen dioxide (NO₂) and its environmental and anthropogenic factors, particularly in cities like Dhaka (Gazipur), Bangladesh, which suffers from some of the world's worst air quality. This study analysed NO₂ concentrations in Gazipur from 2019 to 2022 using Sentinel-5P TROPOMI data on the Google Earth Engine (GEE) platform. Correlations and regression analysis were done between NO₂ levels and various environmental factors, including land surface temperature (LST), normalized difference vegetation index (NDVI), land use and land cover (LULC), population density, road density, settlement density, and industry density. The results reveal significant seasonal variations. The highest annual mean NO₂ concentration (3.1 × 10^–4 mol/ m²)was recorded for winter 2021, and the lowest (1.1 × 10^–4 mol/m²) was for monsoon 2022. The study demonstrates a significant positive correlation between NO₂ concentrations and LST (0.47), road density (0.55), settlement density (0.44), and industrial density (0.35) and a negative correlation with NDVI (− 0.4). Regression analysis revealed that NO₂ concentrations were positively associated with land surface temperature (LST; β = 0.02, R² = 0.22), road density (β = 0.002, R² = 0.30), settlement density (β = 0.002, R² = 0.19), and industrial density (β = 0.007, R² = 0.12), while a negative association was observed with NDVI (β = − 0.28, R² = 0.16). This research offers critical insights for policymakers and urban planners, advocating for enhanced green infrastructure, stringent emission controls, and sustainable urban development strategies to mitigate air pollution in Gazipur. Our methodological approach and findings contribute to the broader discourse on urban air quality management in developing countries.