A Robust Deep Learning Approach for Spatiotemporal Estimation of Satellite AOD and PM2.5

Abstract

Accurate estimation of fine particulate matter with diameter ≤2.5 μm (PM2.5) 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 PM2.5 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 PM2.5 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 PM2.5 in the Jing-Jin-Ji metropolitan region of China in 2015. The presented approach achieved competitive performance in AOD imputation (mean test R2: 0.96; mean test RMSE: 0.06) and PM2.5 estimation (test R2: 0.90; test RMSE: 22.3 μg/m3). In the additional independent tests using ground AERONET AOD and PM2.5 measurements at the monitoring station of the U.S. Embassy in Beijing, this approach achieved high R2 (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 PM2.5 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 PM2.5 using satellite AOD having massive missing values.

Exceedances and trends of particulate matter (PM2.5) in five Indian megacities

“…This USEC data has been used for the study of PM 2.5 levels in the urban environment for different purposes viz. ; to study the trend and characteristics of PM 2.5 ( Chen et al, 2020 ; Fontes et al, 2017 ; Sreekanth et al, 2018 ; Liang et al, 2016 ; Batterman et al, 2016 ; San Martini et al, 2015 ), to compare with other data and model evaluation ( Jiang et al, 2015 ; Li, 2020 ; Matthias et al, 2017 ; Mukherjee and Toohey, 2016 ; Shimadera et al, 2016 ; Uno et al, 2014 ; Wang et al, 2018 ), and to estimate the health impacts ( Han et al, 2020 ; Lowsen and Conway, 2016 ; Luong et al, 2020 ; Nhung et al, 2020 ; Tian et al, 2020 ; Wang et al, 2020 ; You et al, 2016 ; Zhang et al, 2020 ). While most of the studies are carried out in China, few studies have been carried out for other countries including Vietnam ( Hien et al, 2019 ; Luong et al, 2020 ), Japan ( Shimadera et al, 2016 ) Indonesia ( Kusuma et al, 2019 ), Mongolia ( Hill et al, 2017 ), Bangladesh ( Auvee and Bashar, 2019 ), and Singapore ( Liu and Salvo, 2018 ).…”

SAMIRA-SAtellite Based Monitoring Initiative for Regional Air Quality

“…Due to the complex spatial and temporal relationship between the total column aerosol optical depth and ground-level particulate matter, AODto-PM conversion, the second activity within SAMIRA, is a rather complex challenge. A multitude of methods for the AOD-to-PM conversion were developed throughout the years, for example using empirical and multivariate relations (e.g., [11,12]), scaling of the satellite AOD with the PM 2.5 /AOD ratio from a CTM [13], synergistic satellite and ground-based AOD [14], spectral and synergistic satellite information [15,16], fused satellite and modelcalibrated PM 2.5 [17], and machine learning [18][19][20]. For SAMIRA we chose a physical based AOD-to-PM conversion method, the foundation of which goes back to work of [21].…”

Embedded Generative Air Pollution Model with Variational Autoencoder and Environmental Factor Effect in Ulaanbaatar City

“…Li et al introduced the advantage of using two-stage models in geographically weighted machine learning [34], and deep learning to robustly impute a long time series of multi-angle implementation of atmospheric correction aerosol optical depth [35]. The achievements of embedded methods with deep learning approaches were considered in [36,37], and the high performance of autoencoder-based residual networks consisting of learning models is analyzed in [38].…”