Data augmentation for bias correction in mapping PM_2.5 based on satellite retrievals and ground observations

Abstract

As most air quality monitoring sites are in urban areas worldwide, machine learning models may produce substantial estimation bias in rural areas when deriving spatiotemporal distributions of air pollutants. The bias stems from the issue of dataset shift, as the density distributions of predictor variables differ greatly between urban and rural areas. We propose a data-augmentation approach based on the multiple imputation by chained equations (MICE-DA) to remedy the dataset shift problem. Compared with the benchmark models, MICE-DA exhibits superior predictive performance in deriving the spatiotemporal distributions of hourly PM_2.5 in the megacity (Chengdu) at the foot of the Tibetan Plateau, especially for correcting the estimation bias, with the mean bias decreasing from -3.4 µg/m³ to -1.6 µg/m³. As a complement to the holdout validation, the semi-variance results show that MICE-DA decently preserves the spatial autocorrelation pattern of PM_2.5 over the study area. The essence of MICE-DA is strengthening the correlation between PM_2.5 and aerosol optical depth (AOD) during the data augmentation. Consequently, the importance of AOD is largely enhanced for predicting PM_2.5, and the summed relative importance value of the two satellite-retrieved AOD variables increases from 5.5% to 18.4%. This study resolved the puzzle that AOD exhibited relatively lower importance in local or regional studies. The results of this study can advance the utilization of satellite remote sensing in modeling air quality while drawing more attention to the common dataset shift problem in data-driven environmental research.

Keywords

Aerosol optical depth / Dataset shift / Spatiotemporal Distribution / Air quality monitoring / Multiple imputation by chained equations

CRediT authorship contribution statement

Taimoor Hassan Farooq: Conceptualization, Investigation, Writing – original draft, Software, Writing – review & editing. Shagufta Jabeen: Methodology, Investigation, Writing – original draft. Awais Shakoor: Writing – review & editing. Muhammad Saleem Arif: Writing – review & editing. Nadia Siddique: Methodology, Software. Khuram Shahzad: Writing – review & editing. Muhammad Umair Riaz: Methodology, Investigation. Yong Li: Writing – review & editing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This research was financially supported by Central South University of Forestry and Technology Research Funding (70702-45200003), and Scientific Research Foundation of Hunan Provincial Education Department (70702-22200007).

References

[1] Aguilera R., Luo N., Basu R., Wu J., Clemesha R., Gershunov A., Benmarhnia T., 2023. A novel ensemble-based statistical approach to estimate daily wildfire-specific PM_2.5 in California (2006-2020). Environ. Int. 171, 107719, 10.1016/j.envint.2022.107719.

[2] Bai H., Zheng Z., Zhang Y., Huang H., Wang L., 2021. Comparison of Satellite-based PM_2.5 Estimation from Aerosol Optical Depth and Top-of-atmosphere Reflectance. Aerosol Ari Qual. Res. 21, 200257. http://dx.doi.org/10.4209/aaqr.2020.05.0257.

[3] Bentéjac C., Csörgő A., Martínez-Muñoz G., 2021. A comparative analysis of gradient boosting algorithms. Artif. Intell. Rev. 54, 1937-1967. 10.1007/s10462-020-09896-5.

[4] Chen T., Guestrin C., 2016. XGBoost: A Scalable Tree Boosting System. 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 785-794. 10.1145/2939672.2939785.

[5] Chen G., Li Y., Zhou Y., Shi C., Guo Y., Liu Y., 2021. The comparison of AOD-based and non-AOD prediction models for daily PM_2.5 estimation in Guangdong province, China with poor AOD coverage. Environ. Res. 195, 110735, 10.1016/j.envres.2021.110735.

[6] Chen J., Yin J., Zang L., Zhang T., Zhao M., 2019. Stacking machine learning model for estimating hourly PM_2.5 in China based on Himawari 8 aerosol optical depth data. Sci. Total Environ. 697, 134021, 10.1016/j.scitotenv.2019.134021.

[7] CIESIN, 2018. Gridded Population of the World, Version 4 (GPWv4): Population Density, Revision 11. 10.7927/H49C6VHW (accessed on 24 November, 2021).

[8] deSouza P., Kahn R.A., Limbacher J.A., Marais E.A., Duarte F., Ratti C., 2020. Combining low-cost, surface-based aerosol monitors with size-resolved satellite data for air quality applications. Atmos. Meas. Tech. 13, 5319-5334, https://doi.org/10.5194/amt-13-5319-2020.

[9] Didan K., 2015. MOD13Q1 MODIS/Terra Vegetation Indices 16-day L3 Global 250m SIN Grid V006. 10.5067/MODIS/MOD13Q1.006 (accessed on 21 November, 2021).

[10] Douzas G., Bacao F., 2017. Self-Organizing Map Oversampling (SOMO) for imbalanced data set learning. Expert Syst. Appl. 82, 40-52. 10.1016/j.eswa.2017.03.073.

[11] Duarte A.L., Schneider I.L., Artaxo P., Oliveira M.L.S., 2022. Spatiotemporal assessment of particulate matter (PM₁₀ and PM_2.5) and ozone in a Caribbean urban coastal city. Geosci. Front. 13, 101168, 10.1016/j.gsf.2021.101168.

[12] Efron B., 2020. Prediction, Estimation, and Attribution. j.Am. Stat. Assoc. 115, 636-655. 10.1080/01621459.2020.1762613.

[13] ESA, 2017. Land Cover CCI Product User Guide Version 2. Tech. Rep 2017. http://www.esa-landcover-cci.org (accessed on 21 November, 2021).

[14] Falah S., Kizel F., Banerjee T., Broday D.M., 2023. Accounting for the aerosol type and additional satellite-borne aerosol products improves the prediction of PM_2.5 concentrations. Environ. Pollut. 320, 121119, 10.1016/j.envpol.2023.121119.

[15] Feng L., Li Y., Wang Y., Du Q., 2020. Estimating hourly and continuous ground-level PM_2.5 concentrations using an ensemble learning algorithm: The ST-stacking model. Atmos. Environ. 223, 117242, 10.1016/j.atmosenv.2019.117242.

[16] Fu J., Tang D., Grieneisen M.L., Yang F., Yang J., Wu G., Wang C., Zhan Y., 2023. A machine learning-based approach for fusing measurements from standard sites, low-cost sensors, and satellite retrievals: Application to NO 2 pollution hotspot identification. Atmos. Environ. 302, 119756, 10.1016/j.atmosenv.2023.119756.

[17] Geng G., Xiao Q., Liu S., Liu X., Cheng J., Zheng Y., Xue T., Tong D., Zheng B., Peng Y., Huang X., He K., Zhang Q., 2021. Tracking Air Pollution in China: Near Real-Time PM_2.5 Retrievals from Multisource Data Fusion. Environ. Sci. Technol. 55, 12106-12115. 10.1021/acs.est.1c01863.

[18] He Q., Wang M., Yim S.H.L., 2021. The spatiotemporal relationship between PM_2.5 and aerosol optical depth in China: influencing factors and implications for satellite PM_2.5 estimations using MAIAC aerosol optical depth. Atmos. Chem. Phys. 21, 18375-18391, https://doi.org/10.5194/acp-21-18375-2021.

[19] Hersbach H., Bell B., Berrisford P., Hirahara S., Horányi A., Muñoz-Sabater J., Nicolas J., Peubey C., Radu R., Schepers D., Simmons A., Soci C., Abdalla S., Abellan X., Balsamo G., Bechtold P., Biavati G., Bidlot J., Bonavita M., De Chiara G., Dahlgren P., Dee D., Diamantakis M., Dragani R., Flemming J., Forbes R., Fuentes M., Geer A., Haimberger L., Healy S., Hogan R.J., Hólm E., Janisková M., Keeley S., Laloyaux P., Lopez P., Lupu C., Radnoti G., de Rosnay P., Rozum I., Vamborg F., Villaume S., Thépaut J.-N., 2020. The ERA 5 global reanalysis. Q. j.Roy. Meteor. Soc. 146, 1999-2049. 10.1002/qj.3803.

[20] Hu X., Belle J.H., Meng X., Wildani A., Waller L.A., Strickland M.J., Liu Y., 2017. Estimating PM_2.5 Concentrations in the Conterminous United States Using the Random Forest Approach. Environ. Sci. Technol. 51, 6936-6944. 10.1021/acs.est.7b01210.

[21] Hua Z., Sun W., Yang G., Du Q., 2019. A Full-Coverage Daily Average PM_2.5 Retrieval Method with Two-Stage IVW Fused MODIS C6 AOD and Two-Stage GAM Model. Remote Sens. 11(13), 1558. 10.3390/rs11131558.

[22] JAXA, 2018. JAXA Himawari Monitor P-Tree System. http://www.eorc.jaxa.jp/ptree/ (accessed on 01 September, 2021).

[23] Jiang T., Chen B., Nie Z., Ren Z., Xu B., Tang S., 2021. Estimation of hourly full-coverage PM_2.5 concentrations at 1-km resolution in China using a two-stage random forest model. Atmos. Res. 248, 105146, 10.1016/j.atmosres.2020.105146.

[24] Kim J., Jeong U., Ahn M.-H., Kim J.H., Park R.J., Lee H., Song C.H., Choi Y.-S., Lee K.-H., Yoo J.-M., Jeong M.-J., Park S.K., Lee K.-M., Song C.-K., Kim S.-W., Kim Y.J., Kim S.-W., Kim M., Go S., Liu X., Chance K., Chan Miller C., Al-Saadi J., Veihelmann B., Bhartia P.K., Torres O., Abad G.G., Haffner D.P., Ko D.H., Lee S.H., Woo J.-H., Chong H., Park S.S., Nicks D., Choi W.J., Moon K.-J., Cho A., Yoon J., Kim S.-k., Hong H., Lee K., Lee H., Lee S., Choi M., Veefkind P., Levelt P.F., Edwards D.P., Kang M., Eo M., Bak J., Baek K., Kwon H.-A., Yang J., Park J., Han K.M., Kim B.-R., Shin H.-W., Choi H., Lee E., Chong J., Cha Y., Koo J.-H., Irie H., Hayashida S., Kasai Y., Kanaya Y., Liu C., Lin J., Crawford J.H., Carmichael G.R., Newchurch M.J., Lefer B.L., Herman J.R., Swap R.J., Lau A.K.H., Kurosu T.P., Jaross G., Ahlers B., Dobber M., McElroy C.T., Choi Y., 2020. New Era of Air Quality Monitoring from Space: Geostationary Environment Monitoring Spectrometer (GEMS). Bull. Am. Meteoral. Soc. 101, E1-E22. 10.1175/BAMS-D-18-0013.1.

[25] Knibbs L.D., van Donkelaar A., Martin R.V., Bechle M.J., Brauer M., Cohen D.D., Cowie C.T., Dirgawati M., Guo Y., Hanigan I.C., Johnston F.H., Marks G.B., Marshall J.D., Pereira G., Jalaludin B., Heyworth J.S., Morgan G.G., Barnett A.G., 2018. Satellite-Based Land-Use Regression for Continental-Scale Long-Term Ambient PM_2.5 Exposure Assessment in Australia. Environ. Sci. Technol. 52, 12445-12455. 10.1021/acs.est.8b02328.

[26] Krawczyk B., 2016. Learning from imbalanced data: open challenges and future directions. Prog. Artif. Intell. 5, 221-232. 10.1007/s13748-016-0094-0.

[27] LAADS, 2018. MCD19A2 - MODIS/Terra+Aqua Land Aerosol Optical Depth Daily L2G Global 1 km SIN Grid. https://ladsweb.modaps.eosdis.nasa.gov/missions-and-measurements/products/MCD19A2#overview (accessed on 01 September, 2021).

[28] Lee H.J., Coull B.A., Bell M.L., Koutrakis P., 2012. Use of satellite-based aerosol optical depth and spatial clustering to predict ambient PM_2.5 concentrations. Environ. Res. 118, 8-15. 10.1016/j.envres.2012.06.011.

[29] Li R., Ma T., Xu Q., Song X., 2018. Using MAIAC AOD to verify the PM_2.5 spatial patterns of a land use regression model. Environ. Pollut. 243, 501-509. 10.1016/j.envpol.2018.09.026.

[30] Li T., Shen H., Yuan Q., Zhang L., 2020. Geographically and temporally weighted neural networks for satellite-based mapping of ground-level PM_2.5. ISPRS J. Photogramm. 167, 178-188. 10.1016/j.isprsjprs.2020.06.019.

[31] Li H., Yang Y., Wang H., Li B., Wang P., Li J., Liao H., 2021. Constructing a spatiotemporally coherent long-term PM_2.5 concentration dataset over China during 1980-2019 using a machine learning approach. Sci. Total Environ. 765, 144263, 10.1016/j.scitotenv.2020.144263.

[32] Li T., Yang Q., Wang Y., Wu J., 2023. Joint estimation of PM_2.5 and O₃ over China using a knowledge-informed neural network. Geosci. Front. 14, 101499, 10.1016/j.gsf.2022.101499.

[33] Liao Q., Zhang W., Fan W., Liu P., Cao P., Du L., 2018. The South Asian subcontinent impact on atmospheric pollutants transport in China and the indicating effect analysis of monitoring in Hailuogou Background Station. Environmental Monitoring in China 34, 41-46 https://www.cnki.net/kcms/doi/10.19316/j.issn.1002-6002.2018.02.06.html. (in Chinese with English abstract).

[34] Liu X., Li C., Liu D., Grieneisen M.L., Yang F., Chen C., Zhan Y., 2022a. Hybrid deep learning models for mapping surface NO₂ across China: One complicated model, many simple models, or many complicated models? Atmos. Res. 278, 106339, 10.1016/j.atmosres.2022.106339.

[35] Liu Y., Li C., Liu D., Tang Y., Seyler B.C., Zhou Z., Hu X., Yang F., Zhan Y., 2022b. Deriving hourly full-coverage PM_2.5 concentrations across China's Sichuan Basin by fusing multisource satellite retrievals: A machine-learning approach. Atmos. Environ. 271, 118930, 10.1016/j.atmosenv.2021.118930.

[36] Liu J., Weng F., Li Z., 2019. Satellite-based PM_2.5 estimation directly from reflectance at the top of the atmosphere using a machine learning algorithm. Atmos. Environ. 208, 113-122. 10.1016/j.atmosenv.2019.04.002.

[37] Lundberg S.M., Erion G., Chen H., DeGrave A., Prutkin J.M., Nair B., Katz R., Himmelfarb J., Bansal N., Lee S.-I., 2020. From local explanations to global understanding with explainable AI for trees. Nat. Mach. Intell. 2, 56-67. 10.1038/s42256-019-0138-9.

[38] Ma Z., Hu X., Sayer A.M., Levy R., Zhang Q., Xue Y., Tong S., Bi J., Huang L., Liu Y., 2016. Satellite-based spatiotemporal trends in PM_2.5 concentrations: China, 2004-2013. Environ. Health Perspect. 124, 184-192. 10.1289/ehp.1409481.

[39] Ma Z., Dey S., Christopher S., Liu R., Bi J., Balyan P., Liu Y., 2022. A review of statistical methods used for developing large-scale and long-term PM_2.5 models from satellite data. Remote Sens. Environ. 269, 112827, 10.1016/j.rse.2021.112827.

[40] Mhawish A., Banerjee T., Sorek-Hamer M., Bilal M., Lyapustin A.I., Chatfield R., Broday D.M., 2020. Estimation of High-Resolution PM_2.5 over the Indo-Gangetic Plain by Fusion of Satellite Data, Meteorology, and Land Use Variables. Environ. Sci. Technol. 54, 7891-7900. 10.1021/acs.est.0c01769.

[41] Moreno-Torres J.G., Raeder T., Alaiz-Rodríguez R., Chawla N.V., Herrera F., 2012. A unifying view on dataset shift in classification. Pattern Recognit. 45, 521-530. 10.1016/j.patcog.2011.06.019.

[42] Nejadkoorki F., Nicholson K., Hadad K., 2011. The design of long-term air quality monitoring networks in urban areas using a spatiotemporal approach. Environ. Monit. Assess. 172, 215-223. 10.1007/s10661-010-1328-4.

[43] OpenStreetMap, 2016. OpenStreetMap Contributors Planet Dump. http://planet.openstreetmap.org (accessed on 10 September, 2021).

[44] Pu Q., Yoo E.-H., 2021. Ground PM_2.5 prediction using imputed MAIAC AOD with uncertainty quantification. Environ. Pollut. 274, 116574, 10.1016/j.envpol.2021.116574.

[45] Quiñonero-Candela J., Sugiyama M., Schwaighofer A., Lawrence N.D., 2008. Dataset shift in machine learning. The Mit Press. 10.7551/mitpress/9780262170055.001.0001.

[46] She Q., Choi M., Belle J.H., Xiao Q., Bi J., Huang K., Meng X., Geng G., Kim J., He K., Liu M., Liu Y., 2020. Satellite-based estimation of hourly PM_2.5 levels during heavy winter pollution episodes in the Yangtze River Delta, China. Chemosphere 239, 124678. 10.1016/j.chemosphere.2019.124678.

[47] Shimodaira H., 2000. Improving predictive inference under covariate shift by weighting the log-likelihood function. j.Stat. Plan. Infer. 90, 227-244, 10.1016/S0378-3758(00)00115-4.

[48] Song Z., Chen B., Huang Y., Dong L., Yang T., 2021. Estimation of PM_2.5 concentration in China using linear hybrid machine learning model. Atmos. Meas. Tech. 14, 5333-5347. 10.5194/amt-14-5333-2021.

[49] SRTM, 2019. The Shuttle Radar Topography Mission. http://dwtkns.com/srtm30m (accessed on 05 November, 2021).

[50] Stafoggia M., Bellander T., Bucci S., Davoli M., de Hoogh K., de' Donato F., Gariazzo C., Lyapustin A., Michelozzi P., Renzi M., Scortichini M., Shtein A., Viegi G., Kloog I., Schwartz J., 2019. Estimation of daily PM₁₀ and PM_2.5 concentrations in Italy, 2013-2015, using a spatiotemporal land-use random-forest model. Environ. Int. 124, 170-179. 10.1016/j.envint.2019.01.016.

[51] Su L., Gao C., Ren X., Zhang F., Cao S., Zhang S., Chen T., Liu M., Ni B., Liu M., 2022. Understanding the spatial representativeness of air quality monitoring network and its application to PM_2.5 in the mainland China. Geosci. Front. 13, 101370, 10.1016/j.gsf.2022.101370.

[52] Sun J., Gong J., Zhou J., 2021. Estimating hourly PM_2.5 concentrations in Beijing with satellite aerosol optical depth and a random forest approach. Sci. Total Environ. 762, 144502, 10.1016/j.scitotenv.2020.144502.

[53] Tang D., Liu D., Tang Y., Seyler B.C., Deng X., Zhan Y., 2019. Comparison of GOCI and Himawari-8 aerosol optical depth for deriving full-coverage hourly PM_2.5 across the Yangtze River Delta. Atmos. Environ. 217, 116973, 10.1016/j.atmosenv.2019.116973.

[54] Tao J., Zhang L., Engling G., Zhang R., Yang Y., Cao J., Zhu C., Wang Q., Luo L., 2013. Chemical composition of PM_2.5 in an urban environment in Chengdu, China: Importance of springtime dust storms and biomass burning. Atmos. Res. 122, 270-283. 10.1016/j.atmosres.2012.11.004.

[55] Thongthammachart T., Shimadera H., Araki S., Matsuo T., Kondo A., 2023. Land use regression model established using Light Gradient Boosting Machine incorporating the WRF/CMAQ model for highly accurate spatiotemporal PM_2.5 estimation in the central region of Thailand. Atmos. Environ. 297, 119595, 10.1016/j.atmosenv.2023.119595.

[56] van Donkelaar A., Martin R V., Brauer M., Boys Brian L., 2015a. Use of Satellite Observations for Long-Term Exposure Assessment of Global Concentrations of Fine Particulate Matter. Environ. Health Perspect. 123, 135-143. 10.1289/ehp.1408646.

[57] van Donkelaar A., Martin R.V., Spurr R.J.D., Burnett R.T., 2015b. High-Resolution Satellite-Derived PM_2.5 from Optimal Estimation and Geographically Weighted Regression over North America. Environ. Sci. Technol. 49, 10482-10491. 10.1021/acs.est.5b02076.

[58] Wang W., He J., Miao Z., Du L., 2021a. Space-Time Linear Mixed-Effects (STLME) model for mapping hourly fine particulate loadings in the Beijing-Tianjin-Hebei region, China. j.Cleaner Prod. 292, 125993, 10.1016/j.jclepro.2021.125993.

[59] Wang J., He L., Lu X., Zhou L., Tang H., Yan Y., Ma W., 2022. A full-coverage estimation of PM_2.5 concentrations using a hybrid XGBoost-WD model and WRF-simulated meteorological fields in the Yangtze River Delta Urban Agglomeration, China. Environ. Res. 203, 111799, 10.1016/j.envres.2021.111799.

[60] Wang X., Sun W., 2019. Meteorological parameters and gaseous pollutant concentrations as predictors of daily continuous PM_2.5 concentrations using deep neural network in Beijing-Tianjin-Hebei, China. Atmos. Environ. 211, 128-137. 10.1016/j.atmosenv.2019.05.004.

[61] Wang Y., Yuan Q., Li T., Tan S., Zhang L., 2021b. Full-coverage spatiotemporal mapping of ambient PM_2.5 and PM₁₀ over China from Sentinel-5P and assimilated datasets: Considering the precursors and chemical compositions. Sci. Total Environ. 793, 148535, 10.1016/j.scitotenv.2021.148535.

[62] Wei J., Li Z., Pinker R.T., Wang J., Sun L., Xue W., Li R., Cribb M., 2021. Himawari-8-derived diurnal variations in ground-level PM_2.5 pollution across China using the fast space-time Light Gradient Boosting Machine (LightGBM). Atmos. Chem. Phys. 21, 7863-7880. 10.5194/acp-21-7863-2021.

[63] Wu Y., Di B., Luo Y., Grieneisen M.L., Zeng W., Zhang S., Deng X., Tang Y., Shi G., Yang F., Zhan Y., 2021. A robust approach to deriving long-term daily surface NO₂ levels across China: Correction to substantial estimation bias in back-extrapolation. Environ. Int. 154, 106576, 10.1016/j.envint.2021.106576.

[64] Xiao Q., Zheng Y., Geng G., Chen C., Huang X., Che H., Zhang X., He K., Zhang Q., 2021. Separating emission and meteorological contributions to long-term PM_2.5 trends over eastern China during 2000-2018. Atmos. Chem. Phys. 21, 9475-9496, https://doi.org/10.5194/acp-21-9475-2021.

[65] Xie Y., Wang Y., Zhang K., Dong W., Lv B., Bai Y., 2015. Daily Estimation of Ground-Level PM_2.5 Concentrations over Beijing Using 3 km Resolution MODIS AOD. Environ. Sci. Technol. 49, 12280-12288. 10.1021/acs.est.5b01413.

[66] Yao F., Wu J., Li W., Peng J., 2019. A spatially structured adaptive two-stage model for retrieving ground-level PM_2.5 concentrations from VIIRS AOD in China. ISPRS J. Photogramm. 151, 263-276. 10.1016/j.isprsjprs.2019.03.011.

[67] Yin S., Li T., Cheng X., Wu J., 2022. Remote sensing estimation of surface PM_2.5 concentrations using a deep learning model improved by data augmentation and a particle size constraint. Atmos. Environ. 287, 119282, 10.1016/j.atmosenv.2022.119282.

[68] You W., Zang Z., Zhang L., Li Y., Pan X., Wang W., 2016. National-Scale Estimates of Ground-Level PM_2.5 Concentration in China Using Geographically Weighted Regression Based on 3 km Resolution MODIS AOD. Remote Sens. 8, 184. 10.3390/rs8030184.

[69] Yu W., Li S., Ye T., Xu R., Song J., Guo Y., 2022. Deep Ensemble Machine Learning Framework for the Estimation of PM_2.5 Concentrations. Environ. Health Perspect. 130, 037004, 10.1289/EHP9752.

[70] Yumimoto K., Nagao T.M., Kikuchi M., Sekiyama T.T., Murakami H., Tanaka T.Y., Ogi A., Irie H., Khatri P., Okumura H., Arai K., Morino I., Uchino O., Maki T., 2016. Aerosol data assimilation using data from Himawari-8, a next-generation geostationary meteorological satellite. Geophys. Res. Lett. 43, 5886-5894. 10.1002/2016GL069298.

[71] Zeng Z., Gui K., Wang Z., Luo M., Geng H., Ge E., An J., Song X., Ning G., Zhai S., Liu H., 2021. Estimating hourly surface PM_2.5 concentrations across China from high-density meteorological observations by machine learning. Atmos. Res. 254, 105516, 10.1016/j.atmosres.2021.105516.

[72] Zhang S., Mi T., Wu Q., Luo Y., Grieneisen M.L., Shi G., Yang F., Zhan Y., 2022. A data-augmentation approach to deriving long-term surface SO₂ across Northern China: Implications for interpretable machine learning. Sci. Total Environ. 827, 154278, 10.1016/j.scitotenv.2022.154278.

[73] Zhang T., Zang L., Wan Y., Wang W., Zhang Y., 2019. Ground-level PM_2.5 estimation over urban agglomerations in China with high spatiotemporal resolution based on Himawari-8. Sci. Total Environ. 676, 535-544. 10.1016/j.scitotenv.2019.04.299.

[74] Zhang W., Zhang R., Wu C., Goh A.T.C., Lacasse S., Liu Z., Liu H., 2020. State-of-the-art review of soft computing applications in underground excavations. Geosci. Front. 11, 1095-1106. 10.1016/j.gsf.2019.12.003.

[75] Zheng Y., Zhang Q., Liu Y., Geng G., He K., 2016. Estimating ground-level PM_2.5 concentrations over three megalopolises in China using satellite-derived aerosol optical depth measurements. Atmos. Environ. 124, 232-242. 10.1016/j.atmosenv.2015.06.046.