出版社：Chinese Association for Aerosol Research in Taiwan
摘要：Aerosol optical depth (AOD) and top-of-atmosphere (TOA) reflectance are two useful sourcesof satellite data for estimating surface PM2.5 concentrations. Comparison of PM2.5 estimatesbetween these two approaches remains to be explored. In this study, satellite observations ofTOA reflectance and AOD from the Advanced Himawari Imager (AHI) onboard the Himawari-8geostationary satellite in 2016 over Yangtze River Delta (YRD) and meteorological data are usedto estimate hourly PM2.5 based on four different machine learning algorithms (i.e., random forest,extreme gradient boosting, gradient boosting regression, and support vector regression). Forboth reflectance-based and AOD-based approaches, our cross validated results show thatrandom forest algorithm achieves the best performance, with a coefficient of determination (R2)of 0.75 and root-mean-square error (RMSE) of 18.71 µg m–3for the former and R2 = 0.65 andRMSE = 15.69 µg m–3for the later. Additionally, we find a large discrepancy in PM2.5 estimatesbetween reflectance-based and AOD-based approaches in terms of annual mean and their spatialdistribution, which is mainly due to the sampling difference, especially over northern YRD inwinter. Overall, reflectance-based approach can provide robust PM2.5 estimates for both annualmean values and probability density function of hourly PM2.5. Our results further show thatalmost all population lives in non-attainment areas in YRD using annual mean PM2.5 fromreflectance-based approach. This study suggests that reflectance-based approach is a valuableway for providing robust PM2.5 estimates and further for constraining health impact assessments.
