Assessing urban-rural NO2 trend of the Tibetan Plateau via satellite observations

• High-resolution NO 2 column from 2015 to 2022 is obtained by fusion algorithm. • Improved ground-level NO 2 shows fine spatial details with long term series. • Decline rate of surface NO 2 in non-urban area is 14% higher than urban area. • Significant overestimation of NO 2 exposure by 49% pollution days is observed. In the background of global climate change, the trend of atmospheric pollution of the Tibetan Plateau region (TP), where ecosystem is susceptible to their synergistic effects, requires prompt attention. Nitrogen dioxide (NO 2 ), as an important air pollutant of the TP with a profound social and ecological effect, lacks sufficient investigations. Current assessments on NO 2 greatly rely on machine-learning models based on NO 2 column derived by Ozone Monitoring Instrument (OMI), which provides observations over 20 years. Meanwhile, an unexpected overestimation in non-urban areas of the TP due to the coarse spatial resolution of OMI strongly impacts the long-term assessment. Consequently, this study proposes a fusion strategy to improve the long-term OMI NO 2 column based on the recent available fine-scale NO 2 column from TROPOspheric Monitoring Instrument (TROPOMI). Besides the fine resolution and extended applicability to periods long before available TROPOMI observations, the improved NO 2 column obtained by fusion strategy retrieves ground-level NO 2 concentrations with enhanced accuracy (with R 2 0.831–0.880) and approximately 20% mitigated overestimation. Based on the improved dataset, the urban–rural contrast is revealed, as ground-level NO 2 in non-urban regions declines by 0.88 μg/m 3 per year, while 14% lower rate is observed in urban regions. It further indicates that previous exposure to NO 2 exceeding WHO health standards is significantly overestimated (by 49 % pollution days). The results contribute to understanding the different impacts of human activities and climate change on air quality of urban–rural areas.