Estimation of hourly PM1 concentration in China and its application in population exposure analysis

Particulate pollution is closely related to public health. PM1 (particles with an aerodynamic size not larger than 1 μm) much more harmful particles sizes because it goes deeper into the body and hence arouses social concern. However, sparse unevenly distributed ground-based observations limit understanding of spatio-temporal distributions in China. In this study, hourly concentrations central eastern China were retrieved based on a random forest model using aerosol optical depth (AOD) from Himawari-8, meteorological geographic information as inputs. Here spatiotemporal autocorrelation was also considered model. Experimental results indicate that although performance proposed shows diurnal, seasonal spatial variations, relatively better others, determination coefficient (R2) 0.83 calculated 10-fold cross validation method. Geographical map implies level Beijing-Tianjin-Hebei higher other regions, mean value ∼55 μg/m3. Based exposure analysis, we found about 75% population lives environment 35 μg/m3 whole study area. The retrieval dataset great significance for further exploring impact