Estimating hourly PM2.5 concentrations in Beijing with satellite aerosol optical depth and a random forest approach

Jin Sun; Jianhua Gong; Jieping Zhou

doi:10.1016/j.scitotenv.2020.144502

Estimating hourly PM_2.5 concentrations in Beijing with satellite aerosol optical depth and a random forest approach

Sci Total Environ. 2021 Mar 25:762:144502. doi: 10.1016/j.scitotenv.2020.144502. Epub 2020 Dec 14.

Authors

Jin Sun¹, Jianhua Gong², Jieping Zhou³

Affiliations

¹ National Engineering Research Center for Geoinformatics, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100101, China; University of Chinese Academy of Sciences, Beijing 100049, China.
² National Engineering Research Center for Geoinformatics, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100101, China; University of Chinese Academy of Sciences, Beijing 100049, China; Zhejiang-CAS Application Center for Geoinformatics, Jiashan 314100, China.
³ National Engineering Research Center for Geoinformatics, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100101, China. Electronic address: zhoujp@aircas.ac.cn.

PMID: 33360341
DOI: 10.1016/j.scitotenv.2020.144502

Abstract

Assessing short-term exposure to PM_2.5 requires the concentration distribution at a high spatiotemporal resolution. Abundant researches have derived the daily predictions of fine particles, but estimating hourly PM_2.5 is still a challenge restrained by the input data. The recent aerosol optical depth (AOD) product from Himawari-8 provides hourly satellite observations informative to modelling. In this study, we developed separate random forest models with and without AOD and combined the estimates to obtain a full-coverage hourly PM_2.5 distribution. 10-fold cross validation R² ranged from 0.92 to 0.95 and root mean square errors from 14.1 to 16.9 μg/m³, indicating the good model performance. Spatial convolutional layers of PM_2.5 measurements and temporal accumulation effects of meteorological features were added into the model. They turned out to be of the most important predictors and improved the performance significantly. Finally, we mapped hourly PM_2.5 at a 1-km resolution in Beijing during a pollution episode in 2019 and studied the pollution pattern. The study proposed a method to obtain 24-h full-coverage hourly PM_2.5 estimates which are useful for acute exposure assessment in epidemiological researches.

Keywords: AOD; Full-coverage; High resolution; Himawari-8; MAIAC; Machine learning.