Journal
ENVIRONMENTAL POLLUTION
Volume 268, Issue -, Pages -Publisher
ELSEVIER SCI LTD
DOI: 10.1016/j.envpol.2020.115736
Keywords
Downscaling; Surface ozone; Ensemble learning; Simulation interpretability
Categories
Funding
- Bavarian State Ministry of the Environment and Consumer Protection [K3-8503-PN 18-17]
- Bavarian State Ministry of Health and Care
Ask authors/readers for more resources
This study developed a hybrid modeling system to improve the accuracy of hourly surface O-3 observations at European air quality stations. By using machine learning algorithms and analyzing meteorological factors, the study found variations in the effect of temperature on O-3 concentrations.
Ozone (O-3) is a harmful pollutant when present in the lowermost layer of the atmosphere. Therefore, the European Commission formulated directives to regulate O-3 concentrations in near-surface air. However, almost 50% of the 5068 air quality stations in Europe do not monitor O-3 concentrations. This study aims to provide a hybrid modeling system that fills these gaps in the hourly surface O-3 observations on a site scale with much higher accuracy than existing O-3 models. This hybrid model was developed using estimations from multiple linear regression-based eXtreme Gradient Boosting Machines (MLR-XGBM) and O-3 reanalysis from European regional air quality models (CAMS-EU). The binary classification of extremely high O-3 events and the 1- and 24-h forecasts of hourly O-3 were investigated as secondary aims. In this study thirteen stations in Northern Bavaria, out of which six do not monitor O-3, were chosen as test sites. Considering the computational complexity of machine learning algorithms (MLAs), we also applied two recent MLA interpretation methods, namely SHapley Additive explanations (SHAP) and Local interpretable model-agnostic explanations (LIME). With SHAP, we showed an increasing effect of temperature on O-3 concentrations which intensifies for temperatures exceeding 17 degrees C. According to LIME, O-3 concentration peaks are mainly governed by meteorological factors under dry and warm conditions on a regional scale, whereas local nitrogen oxide concentrations control base O-3 concentrations during cold and wet periods. While recently developed MLAs for the spatial estimation of hourly O-3 concentrations had a station-based root-mean-square error (RMSE) above 27 g/m(3), our proposed model significantly reduced the estimation errors by about 66% with an RMSE of 9.49 g/m(3). We also found that logistic regression (LR) and MLR-XGBM performed best in the site -scale classification and 24-h forecast of O-3 concentrations (with a station-averaged accuracy and RMSE of 0.95 and 19.34 g/m(3), respectively). (C) 2020 Elsevier Ltd. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available