PREDICTION OF PM2.5 EMISSIONS USING ANN
Keywords:
PM2.5, emissions prediction, air pollution, MLP, ANNAbstract
Rapid economic and industrial development and the emission of various pollutants into the atmosphere result in increasing pressure on the environment. Particulate matter, PM2.5 represent one of the most significant atmospheric pollutants and one of the most significant indicators of air quality. In this sense, the monitoring of this pollutant is extremely important, especially if we take into account the negative consequences on human health caused by long-term exposure to this pollutant. The subject of this paper is the development of a model based on artificial neural networks (ANN) for the prediction of PM2.5 emissions at the national level. For the development of the model, a standard three-layer neural network with one input layer, one hidden layer and one output layer was used in this work. For the development of the ANN model, available data for 23 countries of Europe for the period from 2010 to 2019 were used. One of the most important segments in the development of the ANN model is the selection of the most important input parameters, i.e. those input parameters that have the greatest influence on the output. As industry and traffic have a significant effect on the emissions of PM2.5, in this paper 7 inputs were selected that were considered to contribute to the emissions of this pollutant. Industrial, economic and traffic indicators were used as input data for the development of the model: Energy productivity, Final energy consumption-total, Final energy consumption - transport sector, Final energy consumption in households, Modal split of passenger transport, Primary energy consumption and Gross domestic product. To evaluate the performance of the created model, the statistical performance indicator coefficient of determination R2 was used. The obtained results of the created ANN model show a very good agreement between the measured and the ANN model predicted PM2.5 emissions values, with the value of the coefficient of determination R-0.869. During the development of the ANN model, the significance of individual inputs was also assessed. It turned out that Final energy consumption in households and Gross domestic product have the greatest influence on PM2.5 emissions. Other inputs have a slightly smaller influence, but certainly significant. Bearing in mind the importance of PM2.5 as an atmospheric pollutant, the existence of alternative models, in addition to the existing ones, for estimating the emissions of this pollutant is of great importance. Based on the results of the model, it can be said that ANN can serve as an alternative model for estimating PM2.5 emissions.
References
Antanasijević, D., Pocajt, V., Perić-Grujić, A., & Ristić, M. (2018). Multiple-input–multiple-output general regression neural networks model for the simultaneous estimation of traffic-related air pollutant emissions. Atmospheric Pollution Research, 9(2), 388–397. https://doi.org/10.1016/j.apr.2017.10.011
Bera, B., Bhattacharjee, S., Sengupta, N., & Saha, S. (2021). PM2.5 concentration prediction during COVID-19 lockdown over Kolkata metropolitan city, India using MLR and ANN models. Environmental Challenges, 4, 100155. https://doi.org/10.1016/j.envc.2021.100155
Cabaneros, S. M., Calautit, J. K., & Hughes, B. R. (2019). A review of artificial neural network models for ambient air pollution prediction. Environmental Modelling & Software, 119, 285–304. https://doi.org/10.1016/j.envsoft.2019.06.014
Database—Eurostat. Retrieved 18 November 2022, from https://ec.europa.eu/eurostat/data/database
Dong, L., Hua, P., Gui, D., & Zhang, J. (2022). Extraction of multi-scale features enhances the deep learning-based daily PM2.5 forecasting in cities. Chemosphere, 308, 136252. https://doi.org/10.1016/j.chemosphere.2022.136252
Environment—OECD Data. Retrieved 26 June 2023, from The OECD website: http://data.oecd.org/environment.htm
Kim, Y., Park, S.-B., Lee, S., & Park, Y.-K. (2023). Comparison of PM2.5 prediction performance of the three deep learning models: A case study of Seoul, Daejeon, and Busan. Journal of Industrial and Engineering Chemistry, 120, 159–169. https://doi.org/10.1016/j.jiec.2022.12.022
Masood, A., & Ahmad, K. (2020). A model for particulate matter (PM2.5) prediction for Delhi based on machine learning approaches. Procedia Computer Science, 167, 2101–2110. https://doi.org/10.1016/j.procs.2020.03.258
Park, S.-Y., Woo, S.-H., & Lim, C. (2023). Predicting PM10 and PM2.5 concentration in container ports: A deep learning approach. Transportation Research Part D: Transport and Environment, 115, 103601. https://doi.org/10.1016/j.trd.2022.103601
Peng, J., Han, H., Yi, Y., Huang, H., & Xie, L. (2022). Machine learning and deep learning modeling and simulation for predicting PM2.5 concentrations. Chemosphere, 308, 136353. https://doi.org/10.1016/j.chemosphere.2022.136353
Sobri, N. M., Yaacob, W. F. W., Ismail, N. A., Malik, M. A. A., Rahman, R. A., Baser, N. A., & Sukhairi, S. A. M. (2021). Predicting Particulate Matter (PM2.5) in Malaysia using Multiple Linear Regression and Artificial Neural Network. Journal of Physics: Conference Series, 2084(1), 012010. https://doi.org/10.1088/1742-6596/2084/1/012010
Stamenković, L. J., Antanasijević, D. Z., Ristić, M. Đ., Perić-Grujić, A. A., & Pocajt, V. V. (2016). Estimation of NMVOC emissions using artificial neural networks and economical and sustainability indicators as inputs. Environmental Science and Pollution Research, 23(11), 10753–10762. https://doi.org/10.1007/s11356-016-6279-z
Stamenković, L. J., Antanasijević, D. Z., Ristić, M. Đ., Perić-Grujić, A. A., & Pocajt, V. V. (2017). Prediction of nitrogen oxides emissions at the national level based on optimized artificial neural network model. Air Quality, Atmosphere & Health, 10(1), 15–23. https://doi.org/10.1007/s11869-016-0403-6
Su, X., Gough, W., & Shen, Q. (2016, June 20). Correlation of PM 2.5 and meteorological variables in Ontario cities: Statistical downscaling method coupled with artificial neural network. 215–226. Crete, Greece. https://doi.org/10.2495/AIR160201
Wu, F., Min, P., Jin, Y., Zhang, K., Liu, H., Zhao, J., & Li, D. (2023). A novel hybrid model for hourly PM2.5 prediction considering air pollution factors, meteorological parameters and GNSS-ZTD. Environmental Modelling & Software, 167, 105780. https://doi.org/10.1016/j.envsoft.2023.105780
Wu, S., & Li, H. (2022). Prediction of PM2.5 concentration in urban agglomeration of China by hybrid network model. Journal of Cleaner Production, 374, 133968. https://doi.org/10.1016/j.jclepro.2022.133968
Xia, X., Yao, L., Lu, J., Liu, Y., Jing, W., & Li, Y. (2022). Observed causative impact of fine particulate matter on acute upper respiratory disease: A comparative study in two typical cities in China. Environmental Science and Pollution Research International, 29(8), 11185–11195. https://doi.org/10.1007/s11356-021-16450-5
Yang, H., Liu, Z., & Li, G. (2022). A new hybrid optimization prediction model for PM2.5 concentration considering other air pollutants and meteorological conditions. Chemosphere, 307, 135798. https://doi.org/10.1016/j.chemosphere.2022.135798
Yang, H., Wang, W., & Li, G. (2023). Prediction method of PM2.5 concentration based on decomposition and integration. Measurement, 216, 112954. https://doi.org/10.1016/j.measurement.2023.112954
