Spatiotemporal dynamics of NO₂ concentration with linear mixed models: a Bangladesh case study

There is currently a limited understanding of how climatic and anthropogenic factors affect atmospheric NO₂ concentration, and how these factors are associated with air pollution over space and time. Using high-resolution TROPOMI satellite data, this study estimates both the degree of association between climatic and anthropogenic factors, and the spatiotemporal variability of NO₂ concentration over Bangladesh. Several linear mixed models were developed to isolate possible factors affecting the NO₂ concentration values recorded between July 2018 and June 2019). This included monthly mean maximum temperature (MMAXT), rainfall, wind speed (WS), relative humidity (RH), enhanced vegetation index (EVI), population density, and distance from industrial activities. The study revealed that the very urbanized central region of Bangladesh experienced high NO₂ concentrations, particularly from September through to March. Dynamic variables such as RH, MMAXT, RAIN, and WS can positively or negatively influence NO₂ depending on the time of year. Areas with a high vegetation cover, a low population density, and located some distance from industrial areas tended to have low NO₂ concentrations. This study concluded that policy measures such as transboundary air quality agreements, the introduction of a month-specific green tax, decentralization, industrial relocation, and increased urban tree plantation activities could all prove valuable in reducing NO₂ pollution in Bangladesh.