A six-year (2015-2020) dataset integrating spatial-temporal features with six daily ground-level pollutant concentrations (O3, CO, NO2, SO2, PM2.5, PM10) across China

This comprehensive air pollutant monitoring dataset encompasses 4,453,372 daily records collected from 2,040 monitoring stations over a six-year period (2015–2020) across China. Each record integrates 128 carefully curated features, including meteorological, geographic, anthropogenic indicators, and concentration values for six key air pollutants: O3, CO, NO2, SO2, PM2.5, and PM10. This dataset is in tabular format and is designed with Machine Learning (ML) and Deep Learning (DL) readiness. It provides unified variables, consistent monitoring station identifiers, and a complete feature space. The Python code for benchmarking the ML/DL models is also provided for reference.