| Summary: | Accurate estimation of fine particulate matter with diameter ≤2.5 μm (PM<sub>2.5</sub>) at a high spatiotemporal resolution is crucial for the evaluation of its health effects. Previous studies face multiple challenges including limited ground measurements and availability of spatiotemporal covariates. Although the multiangle implementation of atmospheric correction (MAIAC) retrieves satellite aerosol optical depth (AOD) at a high spatiotemporal resolution, massive non-random missingness considerably limits its application in PM<sub>2.5</sub> estimation. Here, a deep learning approach, i.e., bootstrap aggregating (bagging) of autoencoder-based residual deep networks, was developed to make robust imputation of MAIAC AOD and further estimate PM<sub>2.5</sub> at a high spatial (1 km) and temporal (daily) resolution. The base model consisted of autoencoder-based residual networks where residual connections were introduced to improve learning performance. Bagging of residual networks was used to generate ensemble predictions for better accuracy and uncertainty estimates. As a case study, the proposed approach was applied to impute daily satellite AOD and subsequently estimate daily PM<sub>2.5</sub> in the Jing-Jin-Ji metropolitan region of China in 2015. The presented approach achieved competitive performance in AOD imputation (mean test R<sup>2</sup>: 0.96; mean test RMSE: 0.06) and PM<sub>2.5</sub> estimation (test R<sup>2</sup>: 0.90; test RMSE: 22.3 μg/m<sup>3</sup>). In the additional independent tests using ground AERONET AOD and PM<sub>2.5</sub> measurements at the monitoring station of the U.S. Embassy in Beijing, this approach achieved high R<sup>2</sup> (0.82−0.97). Compared with the state-of-the-art machine learning method, XGBoost, the proposed approach generated more reasonable spatial variation for predicted PM<sub>2.5</sub> surfaces. Publically available covariates used included meteorology, MERRA2 PBLH and AOD, coordinates, and elevation. Other covariates such as cloud fractions or land-use were not used due to unavailability. The results of validation and independent testing demonstrate the usefulness of the proposed approach in exposure assessment of PM<sub>2.5</sub> using satellite AOD having massive missing values.
|
|---|
