PM2.5 Estimated Directly from Satellite Data and from Fused Data Produced by an Interpretable Multi-Model Stacking Ensemble Method

Rapid urbanization and industrialization in China have resulted in an increase of PM2.5 concentrations. In this study, an interpretable multi-model stacking ensemble method (IMSEM) with top-of-the-atmosphere reflectance (TOAR) from the Himawari-8 satellite were used to acquire high-resolution PM2.5 data in China. In contrast to the traditional approach whereby PM2.5 is estimated with single models, using TOAR data, IMSEM outperformed single models in terms of several skill scores. The hourly average R2(RMSE) of 10-fold-cross validation reached 0.84 (9.52 μg/m3) in 2021 by IMSEM. The feature importance results of IMSEM showed the significant contributions of TOAR and meteorological variables. The PM2.5 estimates of IMSEM were also fused with surface observations using interpolation for correction and optimization. When this was done for PM2.5 concentrations in 2022, it was found that, among the four seasons, the fusion-based estimate of PM2.5 concentration was highest in winter (49.94 μg/m3), followed by autumn (31.59 μg/m3) and spring (29.07 μg/m3), and lowest in summer (19.25 μg/m3).