Quantitative attribution of PM2.5 and NO2 pollution in a coal-based city of Fenwei Plain, China, using a Generalized Additive Model

Abstract

Taiyuan, as a typical resource based city in China, is undergoing an energy transition from heavy industry to clean development. Assessing the air quality benefits resulting from its transformation policies could offer valuable insights for other industrial cities facing similar challenges. To quantitatively elucidate the causes of major air pollution in Taiyuan, we established attribution models for fine particulate matter (PM_2.5) and nitrogen dioxide (NO₂) using generalized additive models for the period January 2020 to December 2024. The non-linear effects of emissions, meteorological variables, and air mass transport were analyzed. The observed concentrations demonstrated that Taiyuan City has made obvious achievements in the treatment of coal smoke pollution with an observed SO₂ decrease of 37.3%. The quantification reveals that emission reductions contributed to a 20% decrease in PM_2.5 and a 15% decrease in NO₂ over the 5 years of the study. The impacts of a 751m increase in boundary layer height, a 24.5% increase in relative humidity, and an 11.3°C increase in temperature on PM_2.5 were about −25%, 65%, and 65%, respectively. NO₂ was mainly affected by local emissions, while air mass transport from the south leads to a 15% increase in PM_2.5 concentrations. By identifying the impacts of emissions and meteorological variability on PM_2.5 and NO₂ concentrations in Taiyuan during the transformation process, we conclude that the results can provide support for optimizing regional environmental management and sustainable development policies in other industrial cities or regions at similar transformation stages.

 

Click here to read the full research: https://doi.org/10.1016/j.apr.2026.103092