Search

Health and environmental hazards related to high pollution concentrations have become a serious issue from public policy perspectives and human health. Using Machine Learning (ML) approach, this research aims to improve the estimation of grid‐wise PM2.5, a criteria pollutant, by reducing systematic bias from speciation provided by MERRA‐from the Modern‐Era Retrospective analysis for Research and Applications version 2 (MERRA‐2). The ML model was trained using various meteorological parameters and aerosol species simulated by MERRA‐2 and ground measurements from Environmental Protection Agency (EPA) air quality system stations. The ML approach significantly improved performance and reduced mean bias in the 0–10 μg m−3 range. We also used the Random Forest ML model for each EPA region using 1 year of collocated data sets. The resulting ML models for each EPA region were validated, and the aggregate data set has a Spearman Rank correlation (SR) of 0.73 (RMSE = 4.8 μg m−3) and 0.69 (RMSE = 5.8 μg m−3) for training and testing, respectively. The SR (and RMSE in μg m−3) increased to 0.81 (3.9), 0.89 (1.6), and 0.90 (1.1) for daily, monthly, and yearly averages, respectively. The results from the initial implementation of the ML model for the global region are encouraging. Still, they require more research and development to overcome challenges associated with data gaps in many parts of the world. Plain Language Summary: PM2.5 is one of the most critical pollutants monitored globally since it is a serious health and environmental hazard. Monitoring of PM2.5 helps understand the air quality and provides advisory to the public. This advisory becomes more critical in the scenarios of natural or anthropogenic events where there is a sudden increase in its concentration. Although a ground‐based monitor can detect a high pollution episode, its spatial extent is limited. Numerical model outputs like Modern‐Era Retrospective analysis for Research and Applications version 2 (MERRA‐2) reasonably estimate various meteorology and aerosol fields to understand the spatial distribution. However, coarse resolution and averaging large grid boxes result in certain uncertainties, leading to biases. In this study, we implemented a machine learning model to better estimate PM2.5 at the grid level using MERRA‐2 as a base model. The developed machine learning model estimated the PM2.5 with good accuracy and low biases. Key Points: Machine learning (ML) model to estimate grid‐wise PM2.5 from Modern‐Era Retrospective analysis for Research and Applications version 2 meteorology and aerosols was developedML model was successful in estimating PM2.5 when compared with surface measurementsRegional ML models performed better than the overall ML model [ABSTRACT FROM AUTHOR]

To quantify the impact of transported air pollutants on local air quality, this study applies an inverse modeling technique with tropospheric nitrogen dioxide column retrievals from TROPOspheric Monitoring Instrument, and the CMAQ‐ISAM model, which computes source attribution over six regions of interest in spring 2019 over East Asia. Satellite‐constrained nitrogen oxides (NOx) emissions successfully improve model performance by significantly reducing the normalized mean bias of concentrations of NOx, Ozone (O3), and PM2.5 from most regions in China and South Korea. Anticyclones over the Yellow Sea transport air pollutants and increase the contribution of emissions from adjacent regions to local air quality. Fluxes of air pollutants with posterior emissions showed notable enhancements in their magnitude by 3%–24% for NOy, O3, and PM2.5. The contributions of emissions from neighboring regions on local NOx and O3 budgets in the planetary boundary layer were 22.96%–35.24% and 24.23%–42.26%. The contribution of emissions from North Korea was notable despite its comparably lower emission rates. Regarding the impact of external influences on local air quality, the enhanced chemical loss of NOx compensated for increases by transport. Also, the increased vertical gradient of maximum daily 8‐hr average O3 amplified vertical transport and dry deposition processes in the opposite direction along with slight changes in the chemical net O3 production rate. This study shows the nonlinear response of local air quality to changes in the transport of air pollutants, which could be further investigated concerning changes in emissions and climatological factors that would modify the inflow of air pollutants. Key Points: Top‐down nitrogen oxides emissions by inverse modeling with TROPOspheric Monitoring Instrument increase the modeling reliability and magnitude of air pollutants fluxesSpringtime anticyclones over Yellow Sea increase adjacent regions' emission impact on local air quality, even with North Korean emissionsTransported air pollutants modulate physical and chemical environments, leading to a nonlinear response to local air quality [ABSTRACT FROM AUTHOR]

This paper proposes a deep learning model that integrates a convolutional neural network and a gated recurrent unit with groups of neighboring stations to accurately predict PM2.5 concentrations at 25 stations in Seoul, South Korea. The deep learning model uses observations obtained from one Korea Meteorological Administration (KMA) station, 25 National Institute of Environmental Research (NIER) stations, and 28 automatic weather stations (AWSs) throughout Seoul. To train the deep learning model, we use all available meteorological and air quality data observed between 2015 and 2017. With the trained model, we predict PM2.5 concentrations at all 25 NIER stations in Seoul for 2018. This study also proposes a geographical polygon group model that determines the optimal number of neighboring NIER stations required to increase the accuracy of PM2.5 concentration predictions at the target station. Comparing the model measures for each of the 25 monitoring sites in 2018, we find that the geographical polygon group model achieves an index of agreement of 0.82–0.89 and a Pearson correlation coefficient of 0.70–0.83. Compared to the method using only meteorological and air quality data from one target station (average IOA = 0.77) to predict PM2.5 concentrations at the 25 stations in Seoul, the proposed method using geographical correlation-based neighboring NIER stations as polygonal groups (average IOA = 0.85) improves the PM2.5 prediction accuracy by an average of about 10%. This approach, based on deep learning, can be updated to predict air pollution or air quality indices up to several days in advance. [ABSTRACT FROM AUTHOR]