Superior PM2.5 Estimation by Integrating Aerosol Fine Mode Data from the Himawari-8 Satellite in Deep and Classical Machine Learning Models

Artificial intelligence is widely applied to estimate ground-level fine particulate matter (PM2.5) from satellite data by constructing the relationship between the aerosol optical thickness (AOT) and the surface PM2.5 concentration. However, aerosol size properties, such as the fine mode fraction (FMF), are rarely considered in satellite-based PM2.5 modeling, especially in machine learning models. This study investigated the linear and non-linear relationships between fine mode AOT (fAOT) and PM2.5 over five AERONET stations in China (Beijing, Baotou, Taihu, Xianghe, and Xuzhou) using AERONET fAOT and 5-year (2015–2019) ground-level PM2.5 data. Results showed that the fAOT separated by the FMF (fAOT = AOT × FMF) had significant linear and non-linear relationships with surface PM2.5. Then, the Himawari-8 V3.0 and V2.1 FMF and AOT (FMF&AOT-PM2.5) data were tested as input to a deep learning model and four classical machine learning models. The results showed that FMF&AOT-PM2.5 performed better than AOT (AOT-PM2.5) in modelling PM2.5 estimations. The FMF was then applied in satellite-based PM2.5 retrieval over China during 2020, and FMF&AOT-PM2.5 was found to have a better agreement with ground-level PM2.5 than AOT-PM2.5 on dust and haze days. The better linear correlation between PM2.5 and fAOT on both haze and dust days (dust days: R = 0.82; haze days: R = 0.56) compared to AOT (dust days: R = 0.72; haze days: R = 0.52) partly contributed to the superior accuracy of FMF&AOT-PM2.5. This study demonstrates the importance of including the FMF to improve PM2.5 estimations and emphasizes the need for a more accurate FMF product that enables superior PM2.5 retrieval.

[1]  J. Lelieveld,et al.  Global and regional trends in aerosol optical depth based on remote sensing products and pollutant emission estimates between 2000 and 2009 , 2010 .

[2]  A. Okuyama,et al.  An Introduction to Himawari-8/9— Japan’s New-Generation Geostationary Meteorological Satellites , 2016 .

[3]  Yuesi Wang,et al.  Size-resolved aerosol chemical analysis of extreme haze pollution events during early 2013 in urban Beijing, China. , 2014, Journal of hazardous materials.

[4]  Arata Okuyama,et al.  Preliminary validation of Himawari-8/AHI navigation and calibration , 2015, SPIE Optical Engineering + Applications.

[5]  Robert E. Holz,et al.  Towards a long-term global aerosol optical depth record: applying a consistent aerosol retrieval algorithm to MODIS and VIIRS-observed reflectance , 2015 .

[6]  J. H. Belle,et al.  Satellite-based estimation of hourly PM2.5 levels during heavy winter pollution episodes in the Yangtze River Delta, China. , 2019, Chemosphere.

[7]  Sundar A. Christopher,et al.  Satellite remote sensing of fine particulate matter (PM2.5) air quality over Beijing using MODIS , 2014 .

[8]  Zhengqiang Li,et al.  Merging MODIS and Ground-Based Fine Mode Fraction of Aerosols Based on the Geostatistical Data Fusion Method , 2017 .

[9]  Gerrit de Leeuw,et al.  Joint retrieval of the aerosol fine mode fraction and optical depth using MODIS spectral reflectance over northern and eastern China: Artificial neural network method , 2020 .

[10]  Yang Liu,et al.  Random forest models for PM2.5 speciation concentrations using MISR fractional AODs , 2020, Environmental Research Letters.

[11]  D. Tanré,et al.  Validation of GRASP algorithm product from POLDER/PARASOL data and assessment of multi-angular polarimetry potential for aerosol monitoring , 2020, Earth System Science Data.

[12]  Lianfa Li,et al.  A Robust Deep Learning Approach for Spatiotemporal Estimation of Satellite AOD and PM2.5 , 2020, Remote. Sens..

[13]  Qingyang Xiao,et al.  Predicting monthly high-resolution PM2.5 concentrations with random forest model in the North China Plain. , 2018, Environmental pollution.

[14]  Xiaohua Tong,et al.  Virtual image pair-based spatio-temporal fusion , 2020, Remote Sensing of Environment.

[15]  Wenzhong Shi,et al.  Satellite-based PM2.5 estimation using fine-mode aerosol optical thickness over China , 2017 .

[16]  T. Eck,et al.  Global evaluation of the Collection 5 MODIS dark-target aerosol products over land , 2010 .

[17]  Piet Stammes,et al.  Simulation study of the aerosol information content in OMI spectral reflectance measurements , 2007 .

[18]  G. Leeuw,et al.  Exploring the relation between aerosol optical depth and PM 2.5 at Cabauw, the Netherlands , 2008 .

[19]  Xiaoling Chen,et al.  Spatiotemporal relationship between Himawari-8 hourly columnar aerosol optical depth (AOD) and ground-level PM2.5 mass concentration in mainland China. , 2020, The Science of the total environment.

[20]  Fanghua Wu,et al.  Analysis of influential factors for the relationship between PM 2.5 and AOD in Beijing , 2017 .

[21]  Bingyue Pan,et al.  Application of XGBoost algorithm in hourly PM2.5 concentration prediction , 2018 .

[22]  Zhanqing Li,et al.  New global aerosol fine-mode fraction data over land derived from MODIS satellite retrievals. , 2021, Environmental pollution.

[23]  Zhengqiang Li,et al.  Remote sensing of atmospheric fine particulate matter (PM2.5) mass concentration near the ground from satellite observation , 2015 .

[24]  O. Dubovik,et al.  Validation of POLDER GRASP aerosol optical retrieval over China using SONET observations , 2020 .

[25]  Thomas F. Eck,et al.  GOCI Yonsei Aerosol Retrieval (YAER) algorithm and validation during the DRAGON-NE Asia 2012 campaign , 2015 .

[26]  Xingying Zhang,et al.  Satellite remote sensing of atmospheric particulate matter mass concentration: Advances, challenges, and perspectives , 2021 .

[27]  Roger D Peng,et al.  Community-level spatial heterogeneity of chemical constituent levels of fine particulates and implications for epidemiological research , 2011, Journal of Exposure Science and Environmental Epidemiology.

[28]  Xiaohua Tong,et al.  Spatio-temporal spectral unmixing of time-series images , 2021 .

[29]  Huanfeng Shen,et al.  The relationships between PM2.5 and aerosol optical depth (AOD) in mainland China: About and behind the spatio-temporal variations. , 2019, Environmental pollution.

[30]  Machine learning classification of new asteroid families members , 2020 .

[31]  Pierre Geurts,et al.  Extremely randomized trees , 2006, Machine Learning.

[32]  Zhanqing Li,et al.  New interpretable deep learning model to monitor real-time PM2.5 concentrations from satellite data. , 2020, Environment international.

[33]  Wenzhong Shi,et al.  An improved algorithm for retrieving the fine-mode fraction of aerosol optical thickness, part 1: Algorithm development , 2017 .

[34]  M. Brauer,et al.  Global Estimates of Fine Particulate Matter using a Combined Geophysical-Statistical Method with Information from Satellites, Models, and Monitors. , 2016, Environmental science & technology.

[35]  Melita Keywood,et al.  Impact of biomass burning sources on seasonal aerosol air quality , 2013 .

[36]  Ramesh P. Singh,et al.  Contrasting Aerosol Trends Over South Asia During the Last Decade Based on MODIS Observations , 2011 .

[37]  M. Brauer,et al.  Global Estimates of Ambient Fine Particulate Matter Concentrations from Satellite-Based Aerosol Optical Depth: Development and Application , 2010, Environmental health perspectives.

[38]  W. Shi,et al.  A Spatial-Temporal Interpretable Deep Learning Model for improving interpretability and predictive accuracy of satellite-based PM2.5. , 2021, Environmental pollution.

[39]  LI Chengcai,et al.  Evaluation and possible uncertainty source analysis of JAXA Himawari-8 aerosol optical depth product over China , 2021 .

[40]  A. Berg,et al.  Present and future Köppen-Geiger climate classification maps at 1-km resolution , 2018, Scientific Data.

[41]  Liang-pei Zhang,et al.  Estimate hourly PM2.5 concentrations from Himawari-8 TOA reflectance directly using geo-intelligent long short-term memory network. , 2020, Environmental pollution.

[42]  Junying Sun,et al.  Robust prediction of hourly PM2.5 from meteorological data using LightGBM , 2021, National science review.

[43]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[44]  W. Shi,et al.  Evaluation and Comparison of Himawari-8 L2 V1.0, V2.1 and MODIS C6.1 aerosol products over Asia and the oceania regions , 2020 .

[45]  Hiroshi Murakami,et al.  Common Retrieval of Aerosol Properties for Imaging Satellite Sensors , 2018 .

[46]  Cheng Guo,et al.  Entity Embeddings of Categorical Variables , 2016, ArXiv.

[47]  Zhengqiang Li,et al.  Retrieval of aerosol fine-mode fraction over China from satellite multiangle polarized observations: validation and comparison , 2021 .

[48]  B. R. Gurjar,et al.  The representation of emissions from megacities in global emission inventories , 2008 .

[49]  Jun Wang,et al.  Satellite remote sensing of particulate matter and air quality assessment over global cities , 2006 .

[50]  Jasper R. Lewis,et al.  Advancements in the Aerosol Robotic Network (AERONET) Version 3 database – automated near-real-time quality control algorithm with improved cloud screening for Sun photometer aerosol optical depth (AOD) measurements , 2019, Atmospheric Measurement Techniques.

[51]  Yoshua Bengio,et al.  Understanding the difficulty of training deep feedforward neural networks , 2010, AISTATS.

[52]  Shana Mattoo,et al.  Continuing the MODIS Dark Target Aerosol Time Series with VIIRS , 2020, Remote. Sens..