Estimating hourly surface PM2.5 concentrations across China from high-density meteorological observations by machine learning.

The spatial-temporal variations of the ground-based and satellite-derived PM 2.5 are crucial for studying air quality, human health, and climate change. However, the existing ground-based PM 2.5 monitoring network has sparsely-distributed sites and satellite cannot give 24-h PM 2.5 , which make it difficult to grasp the spatial and sub-daily variation characteristics of PM 2.5. This study aims to fill that gap by establishing a virtual network of hourly PM 2.5 concentration using the LightGBM model, based on the high-density ground meteorological observations at ~2400 sites across China. The virtual network shows a desirable performance of hourly PM 2.5 estimation across China, with R 2 of 0.86, root-mean-square error values of 14.99 μg/m3, and mean absolute error of 9.48 μg/m3 (the results of Cross-Validation). It also exhibits high spatial-temporal consistencies with the observed PM 2.5. Spatially, the heaviest PM 2.5 pollution is mainly distributed in eastern China (especially the Beijing-Tianjin-Hebei, the Yangtze and Pearl river deltas, and the Sichuan-Chongqing areas). Temporarily, PM 2.5 exhibits remarkable seasonal and diurnal changes characterized by higher concentration in winter and nighttime and lower in summer and daytime. Meanwhile, we found that visibility can be used as the primary predictor in the machine learning model to enhance the accuracy of estimated PM 2.5. The established virtual hourly PM 2.5 network (~2400 stations) provides a more intuitive and detailed PM 2.5 data for us to understand the diurnal variation of PM 2.5 and monitor inter-regional transport of haze over China. It thus is of benefit to the study of air pollution control and related diseases. • Hourly PM 2.5 in China is estimated using high-density weather records by LightGBM. • The model shows a superior estimation accuracy at national, regional, and site scales. • A virtual network of high-accuracy and -density hourly PM 2.5 in China is established. • This network fills the gaps of sparse air quality sites and satellite-derived PM 2.5. [ABSTRACT FROM AUTHOR]