Estimating daily ground-level PM2.5 in China with random-forest-based spatiotemporal kriging

Abstract

Ambient fine particulate matter (PM2.5) plays an important role in cardiovascular- and respiratory-related death. Empirical statistical models have been widely applied to estimate ambient PM2.5 concentrations with correlated variables. However, empirical statistical models ignore the nonlinear relationship between PM2.5 and covariates and assume that residuals are independent and identically distributed random variables. Here, a hybrid approach, which integrates random forest (RF) model and spatiotemporal kriging, is proposed to estimate the daily PM2.5 concentration. The proposed RF-based spatiotemporal kriging (RFSTK) model effectively captures nonlinear interactions among different predictors and accounts for the detailed spatiotemporal dependence of the PM2.5 concentration. The RFSTK model performs well in predicting the daily PM2.5 concentration. The 10-fold overall cross-validation R2 value is 0.881, the mean absolute error (MAE) is 6.89 μg/m3 and the root-mean-square error (RMSE) is 11.48 μg/m3, indicating better performance than the original RF model (R2 = 0.848, MAE = 7.88 μg/m3 and RMSE = 13.26 μg/m3). The spatiotemporal prediction of the PM2.5 concentration shows that approximately 90.04% of China had a daily exposure to PM2.5 in 2018 that was below the nation's air quality standard of 75 μg/m3. The proposed hybrid method is entirely general and can be applied to map the ambient PM2.5 concentration over a large spatiotemporal domain.