Spatiotemporal patterns of air quality and its dominant factors in middle–lower reaches of Yellow River

Monitoring air quality is a core scientific issue in atmospheric environmental research. Based on the hourly air quality index (AQI) data from 298 monitoring stations, a multi-scale analysis of air quality was conducted in the middle–lower reaches of the Yellow River from June 2016 to May 2020. Methodologically, unified quality control and missing-data imputation were applied to the hourly AQI records, after which annual, seasonal, and monthly mean AQI values were calculated using arithmetic averaging. The Kriging interpolation was applied to map multi-temporal spatial patterns, while the coefficient of variation and global/local spatial autocorrelation were used to quantify spatial disparity and clustering of AQI. Then, the effects of natural and socioeconomic factors were identified and cross-validated using the geographical detector and multiple linear regression. The results show that AQI exhibits a pronounced downward trend over time, with clear seasonal characteristics and cyclical variations at the monthly scale; spatially, AQI has long displayed a pattern of higher values in downstream and lower values in midstream, with high-concentration areas mainly clustered in the adjacent regions of Henan, Shandong and Shanxi provinces; significant positive spatial autocorrelation is observed at all temporal scales, with stable high–high and low–low clustering patterns; population density, civilian vehicle ownership, and industrial dust emissions show the strongest explanatory power, while air temperature and afforestation area exert moderating effects on AQI variations. This study is of potential significance to cross-provincial coordinated control and refined air pollution management.