پیش بینی غلظت آلاینده های هوای تهران بر اساس متغیرهای هواشناسی با استفاده از شبکه عصبی مصنوعی و رگرسیون خطی در فصول گرم و سرد

نوع مقاله : مقاله پژوهشی

نویسندگان

1 کارشناسی ارشدعلوم و مهندسی محیط زیست، دانشکده علوم محیطی، دانشگاه حکیم سبزواری، سبزوار، ایران

2 نانوتکنولوژی

3 استادیار، دانشکده محیط زیست، دانشگاه تهران

4 استادیار، دانشکده ریاضیات و علوم کامپیوتر، دانشگاه حکیم سبزواری

چکیده

تهران آلوده‌ترین شهر کشور محسوب می‌شود که این آلودگی می‌تواند آثار دراز مدت و کوتاه مدتی بر سلامت انسان داشته باشد. از این‌رو پیش‌بینی غلظت آلاینده‌ها می‌تواند در برنامه‌ریزی‌های پیشگیری و کنترل مفید واقع شود. روشهای متفاوتی برای پیش‌بینی وجود دارد و دراین میان سالها، روش‌های شبکه‌ی عصبی پیشرفت قابل توجهی در پیش‌بینی آلودگی هوا داشته است. در این مطالعه، از شبکه‌ی عصبی مصنوعی پرسپترون سه‌لایه به‌منظور پیش‌بینی غلظت آلاینده‌های PM10، CO و شاخص کیفیت هوا (AQI) در هوای شهر تهران استفاده شد. داده‌های غلظت آلاینده‌ها از اداره‌ی کنترل کیفیت هوای تهران جمع‌آوری شد و داده‌های هواشناسی از اداره‌‌ی کل سازمان هواشناسی کشور طی سال‌های 1392 و 1393 جمع‌آوری شد. بیشترین ضریب همبستگی (R2) برای آلاینده PM10 با مقدار 0.83 در فصول گرم بود و بیشترین ضریب همبستگی آلاینده CO مربوط به فصول سرد بود (R2=0.74). در نهایت بیشترین ضریب همبستگی AQI در فصل سرد (R2=0.57) بود. در مدل رگرسیون خطی بیشترین ضریب همبستگی با مقدار 0.58 برای آلاینده PM10 در فصول گرم بود. بیشترین ضریب همبستگی در این مدل برای آلاینده CO با مقدار 0.33 در فصل سرد بود. درنهایت بیشترین ضریب همبستگی AOI (R2=0.31) در فصل گرم بود. این به این معنی است که با تغییرات متغیرهای هواشناسی، غلظت CO و ذرات معلق و مقادیر شاخص AQI تغییر می‌کند به گونه‌ای که افزایش باد باعث پراکنش آلاینده و کاهش غلظت آن می‌شود و افزایش درجه حرارت باعث افزایش غلظت آلاینده می‌شود. بنابراین بین آنها ارتباط وجود دارد.

کلیدواژه‌ها

عنوان مقاله [English]

Prediction of Air Pollutants Concentration Based on Meteorological Factors in Warm and Cold Season by Artificial Neural Network and Linear Regression, Case Study: Tehran

نویسندگان [English]

  • Razieh Farhadi 1
  • mojtaba hadavifar 2
  • Mazaher Moeinaddini 3
  • Mahmood Amintoosi 4

1 Department of environmental Science of Hakim Sabzevari University, Sabzevar, Iran

2

3 Assistant professor, environmental Science Department, University of Tehran, Tehran, Iran

4 Assistant professor, Faculty of Mathematics and Computer Science, Hakim Sabzevari University, Sabzevar, Iran

چکیده [English]

Tehran is the most polluted city in the country, which can have long-term and short-term effects on human health. Therefore, predicting the concentration of pollutants can be useful in prevention and control planning. There are different methods for predicting, and over the years, neural network methods have made significant progress in predicting pollution. In this study, an artificial neural network of three-layer perceptron was used to predict the concentrations of PM10, CO and air quality index (AQI) in air in Tehran. The concentrations of pollutants were collected from the Tehran Air Quality Control Department and the weather data collected from the Office of the Iranian Meteorological Organization during 2013 and 2015. The highest correlation coefficient (R2) for PM10 pollutant was 0.83 in warm seasons and the highest CO emission factor correlated with cold seasons (R2 = 0.74). Finally, the highest correlation coefficient of AQI was in cold season (R2 = 0.57). In linear regression model, the highest correlation coefficient with 0.58 for PM10 pollutant was in hot seasons. The highest correlation coefficient in this model was for the CO pollutant (0.33) in the cold season. Finally, the highest correlation coefficient was AOI (R2 = 0.31) in the warm season.

کلیدواژه‌ها [English]

  • Air pollution forecast
  • Carbon monoxid
  • Particulate matter
  • Air quality index
Agirre-Basurko, E., G. Ibarra-Berastegi, and I Madariaga. 2006. Regression and Multilayer Perceptron-Based Models to Forecast Hourly O3 and NO2 Levels in the Bilbao Area. In Environmental Modelling and Software, 21(4): 430–46.
Anderson, H. R. 2009. Air Pollution and Mortality: A History. Atmospheric Environment 43: 142–52.
Biancofiore, F., Verdecchia, M.C., Piero,T., Barbara, A., Eleonora, B., Marcella, B., Sebastiano, D., Tommaso, S., Colangeli, C. 2015. Analysis of Surface Ozone Using a Recurrent Neural Network ed. Edward A Keller. Science of the Total Environment 514(4): 379–87.
Brian, G., Michae, K., Bruce, D., David, C., Karen, M., Michael, D. 2007. Comparison of Lead Isotopes with Source Apportionment Models, Including SOM, for Air Particulates. Science of The Total Environment 381(1–3): 169–79.
Chelani, A.B., Chalapati Rao, C.V., Phadke, K.M., Hasan, M.Z. 2002. Prediction of Sulphur Dioxide Concentration Using Artificial Neural Networks. Environmental Modelling & Software 17(2): 159–66.
Demuth, H., Beale, M. 2002. Neural Network Toolbox Users Guide. Copyright 1992-2002, by the Math Works, Inc, Version 4, PP, 840.
Demuzere,  M., Trigo, R. M., Vila-Guerau de Arellano, J., van Lipzig, N. P. M. 2009. The impact of weather and atmospheric circulation on O3 and PM10 levels at a rural mid-latitude site. Atmospheric Chemistry and Physics.  9. 2695–2714.
Demuzere, M., Trigo, R.M., Vila–Guerau de Arellano, J., van Lipzig, N.P.M. 2009. The Impact of Weather and Atmospheric Circulation on O3 and PM10 Levels at a Rural Mid–latitude Site. Atmospheric Chemistry and Physics 9: 2695–2714.
Fausett, L. 1994. Fundamental of Neural Network: Architecture, Algorithms, and Applications, Prentice. Hall Press.
García,  P.J., Sánche Lasherasa, F., García-Gonzaloa, E., de Cos Juez, F.J. 2017. PM10 Concentration Forecasting in the Metropolitan Area of Oviedo (Northern Spain) Using Models Based on SVM, MLP, VARMA and ARIMA: A Case Study. Science of The Total Environment, 621: 753–61.
Giorgi, F. and Meleux, F. 2007. Modelling the regional effects of climate change on air quality, Comp. Rend. Geosci., researchgate, 339 (11), 721–733.
Giustolisi, O., Doglioni, A., Savic, D. A., Webb, B. W. 2007. A Multi-Model Approach to Analysis of Environmental Phenomena. Environmental Modelling and Software 22: 674–82.
Gratani L, Varone L. 2005. Daily and Seasonal Variation of CO2 in the City of Rome in Relationship with the Traffic Volume.  Atmospheric Environment 39: 2619–2624.
Harrison, JI., Ping, SHI., ROY, M. 1997. Regression modelling of hourly NOx and NO2 concentration in urban air in Londen. Atmospheric Environmen, 31(24): 4081–94.
Jian, L., Zhao, Y., Zhu, YP., ZhangM, B., Bertolatti, D. 2012. An Application of ARIMA Model to Predict Submicron Particle Concentrations from  Meteorological Factors at a Busy Roadside in Hangzhou, China.” Science Total Environment, 426: 336–345.
Jiang, D., Zhang, D., Hu, Y., Zeng, X., Tan, Y., Jianguo, T., Demin, S. 2004. Progress in Developing an ANN Model for Air Pollution Index Forecast.  Atmospheric Environment,  38(40 SPEC.ISS.): 7055–64.
Karaca, F., Nikov, A., Alagha, O. 2006. NN-Airpol: A Neural-Networks-Based Method for Air Pollution Evaluation and Control. International Journal of Environment and Pollution, 28(3/4): 310–25.
Mamtimin, B., Meixner, FX. 2011. Air Pollution and Meteorological Processes in the Growing Dryland City of Urumqi (Xinjiang, China). Science of the Total Environment 2011, 409(7): 199–226.
Masoudi,  M. and Gerami, S. 2017. Status of CO as an air pollutant and its prediction, using meteorological parameters in Esfahan, Iran. Pollution. 3 (4). 527-537
McCulloch, W.S., Pitts, W. 1943. A Logical Calculus of the Ideas Imminent in Nervous Activity. B.Math. Biophys,  8: 115–33.
McKendry, I.G. 2015. Evaluation of Artificial Neural Networks for Fine Particulate Pollution (PM10 and PM2.5) Forecasting.” Journal of the Air & Waste Management Association 52(9): 1096–1101.
Mohammadhassani, J., Dadvand, A., Khalilarya, Sh., Solimanpur, M. 2015. Prediction and Reduction of Diesel Engine Emissions Using a Combined ANN–ACO Method. Applied Soft Computing, 34: 139–50.
Nejadkoorki, F and and Baroutian, S. 2012. Forecasting Extreme PM10 Concentrations Using Artificial Neural Networks. Statewide Agricultural Land Use Baseline 2015, 1(1): 277–84.
Orr, Mark J.L. 1996. Introduction to Radial Basis Function Networks. Time: 1–67.
Park,  S., Kim, M., Namgung, H.G., Kim, K. T., Cho, K. H., Kwon, S.B. 2018.  Predicting PM10concentration in Seoul Metropolitan Subway Stations Using Artificial Neural Network (ANN). Journal of Hazardous Materials 341: 75–82.
Patricia, M., Jonathan, A., Fevrier, V., Oscar, C. 2014. A New Neural Network Model Based on the LVQ Algorithm for Multi-Class Classification of Arrhythmias. Information Sciences, 279: 483–97.
Salvador, s and Salvador, El. 2012. Air Quality Index (AQI). 16-17.
Seinfeld, J. H. and Pandis, S. N. 1998. Atmospheric chemistry and physics from air pollution to climate change, New York, John Wiley & Sons, Inc, 1113 pp.
Shariepour, Z. 2010. Seasonal and Daily Variation of Air Pollutants and Their Relation to Meteorological Parameters. Earth and Space Physics,  35(2): 119–137 (In Persian).
Sharma, N., Chaudhry, K., Rao, CC. 2005. Vehicular Pollution Modeling Using Artificial Neural Network Technique: A Review. Journal of Scientific and Industrial research. 64(9): 637.
Shepherd, G.M. 1990. The Synaptic Organization of the Brain,. third edition, Oxford university press.
Shi, Dan., Hongjian, Z., Liming, Y. 2003. Time-Delay Neural Network for the Prediction of Carbonation Tower ’ s Temperature. 52(4): 1125–28.
Yi, J and Prybutok, V.R. 1996. A Neural Network Model Forecasting for Prediction of Daily Maximum Ozone Concentration in an Industrialized Urban Area. Environmental Pollution, 92(3): 349–57.
Zannetti, P. 1990. Air Pollution Modeling, Theories, Computational Methods and Software’s, Computational Mechanics Publication.
Ziomas, I. C., Melas, D., Zerefos, C.S., Bais, A.F., Paliatsos, A.G. 1995. Forecasting Peak Pollutant Levels from Meteorological Variables. Atmospheric Environment , 29(24): 3703–11.