Prediction of Hourly Ozone Concentrations with Multiple Regression and Multilayer Perception Models

Prediction of Hourly Ozone Concentrations with Multiple Regression and Multilayer Perception Models

C. Capilla

Polytechnic University of Valencia, Spain

31 August 2016
| Citation



In this work ozone observations of an urban area of the east coast of the Iberian Peninsula, are analyzed. The data set contains measurements from five automatic air pollution monitoring stations (background suburban or traffic urban). The application of multiple linear regression and neural networks models is considered. These models forecast hourly ozone levels for short-term prediction intervals (1, 8, and 24 h in advance). The study period is 2010–2012. The input variables are meteorological observations, ozone and nitrogen oxides concentrations, and daily and weekly seasonal cycles. The performance criteria to evaluate the computations accuracy are the residual mean square error, the mean absolute error, and the correlation coefficient between observations and predictions. These criteria have better results for the 1-h and 24-h predictions in all the locations. The comparison of multiple linear regressions and multilayer perceptron networks indicates that the second approach allows to obtain more accurate forecast for the three prediction intervals.


multilayer perceptron networks, multiple linear regression, ozone, urban air quality


[1] Hartman, D.L., Kleintank, A.M.G., Rusticucci, M., Alexander, L.V., Brönnimann, S., Charabi, Y., Dentener, F.J., Dlugokencky, E.J., Easterling, D.R., Kaplan, A., Soden, B.J., Thorne, P.W., Wild, M. & Zhai, P.M., Observations: atmosphere and surface (Chapter 2). Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, eds. T.F. Stocker, D. Qin, G.–K. Plattner, M. Tigno, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex & P.M. Midgley, Cambridge University Press, Cambridge: United Kingdom and New York, NY, USA, pp. 159–218, available at (accessed 25 January 2016), 2013.

[2] Official Journal of the European Union, Directive 2008/50/EC of the European Parliament and of the Council of 21 May2008 on Ambient Air Quality and Cleaner Air for Europe, available at http:// L0050&qid=145371475043&rid=1 (accessed 25 January 2016).

[3] Boletin Oficial del Estado. Real Decreto102/2011 de 28 de Enero, Relativo a la Mejora de la Calidad del Aire, available at http// (accessed 25 January 2016).

[4] World Health Organization, Effect of Air Pollution on Childrens Health and Development. World Health Organization Regional Office for Europe, Copenhagen, Denmark, 2005, available at (accessed 25 January 2016).

[5] World Health Organization, Air Quality Guidelines. Global Update 2005. Particulate Matter, Ozone, Nitrogen Dioxide, and Sulfur Dioxide, World health Organization Regional Office for Europe, Copenhagen, Denmark, 2006, available at (accessed 25 January 2016).

[6] Dutot, A.L., Rynkiewicz, J., Steiner, F.E. & Rude, J., A 24-hour forecast of ozone peaks and exceedance levels using neural classifiers and weather predictions. Environmental Modelling & Software, 22, pp. 1261–1269, 2007.

[7] Raheem Abdul, A.M.O., Adekola, F.A. & Obioh, I.O., The seasonal variation of ozone, sulphur dioxide and nitrogen oxides in two Nigerian cities. Environmental Modeling & Assessment, 14, pp. 497–509, 2009.

[8] Ripley, B.D., Statistical aspects of neural networks (Chapter 2). Networks and ChaosStatistical and Probabilistic Aspects, eds. D.E. Barndorff-Nielsen, J.L. Jensen & W.S. Kendall, Chapman & Hall/CRC: London, UK, pp. 40–123,1993.

[9] Gardner, M.W. & Dorling, S.R., Artificial neural networks (the multilayer perceptron)- a review of applications in the atmospheric sciences. Atmospheric Environment, 32(4), pp. 2627–2636, 1998.

[10] Gardner, M.W., The advantages of artificial neural networks and regression tree based air quality models (Chapters 1 and 2). A dissertation submitted to the School of Environmental Sciences of the University of East Anglia (part of the requirements for the degree of Doctor of Philosophy), UK, 1999.

[11] Elkamel, A., Abdul-Wahab, S., Bouhamra, W. & Alper. E., Measurement and prediction of ozone levels around a heavily industrialized area: a neural network approach. Advances in Environmental Research, 5, pp. 47-59, 2001.

[12] Agirre-Basurko, E., Ibarra-Berastegui, G. & Madariaga, I., Regression and multilayer perceptron-based models to forecast hourly O3 and NO2 levels in the Bilbao area. Environmental Modelling & Software, 21, pp.430-446, 2006.

[13] R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, available at available at, 2014.

[14] Gómez-Sanchis, J., Martín-Guerrero, J.D., Soria-Olivas, E., Vila-Francés, J., Carrasco, J.L. & del Valle-Tascón, S., Neural networks for analysing the relevance of input variables in the prediction of tropospheric ozone concentration. Atmospheric Environment, 40, pp. 6173-6180, 2006.

[15] Castell-Balaguer, N., Téllez, L. & Mantilla, E., Daily, seasonal and monthly variations in ozone levels recorded at the Turia river basin in Valencia (Eastern Spain). Environmental Science & Pollution Research, 19,pp. 3461-3480, 2012.

[16] Castell-Balaguer, N., Téllez, L., Luján, A. & Mantilla, E., Informe Final Previozono 2010. Programa Especial de Vigilancia de las Concentraciones de Ozono Troposférico en la Comunidad Valenciana. (accessed 28 January 2016), 2010.

[17] Ballester, F., Iñíguez, C. & García, F., ENHIS-1 project: WP5 health impact assessment. Local city report Valencia. (accessed 4th May 2006), 2005.

[18] The Mediterranean Centre for Environmental Studies Foundation, Informe Final Previozono 2012, available at (accessed 28 January 2016), 2012.