Advertisement

Air Quality Forecasting in Madrid Using Long Short-Term Memory Networks

  • Esteban PardoEmail author
  • Norberto Malpica
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10338)

Abstract

European and Spanish legislation set hourly limits for Nitrogen Dioxide, \(NO_2\), that are enforced with traffic restrictions. In this context it is important to warn the citizens in advance, which can only be done if the \(NO_2\) levels are forecasted. In this paper we propose a deep learning based air quality forecasting system that uses air quality and meteorological data to produce \(NO_2\) forecasts up to 24 h with a root mean squared error, RMSE, of 10.54 \(\upmu {\text {g}}/{\text {m}}^3\). We also compare our results with the model based system CALIOPE.

Keywords

Air quality forecasting Long short-term memory 

Notes

Acknowledgments

This work was partially funded by Banco de Santander and Universidad Rey Juan Carlos in the Funding Program for Excellence Research Groups ref. “Computer Vision and Image Processing (CVIP)”. We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Tesla K40 GPU used for this research. We would also like to thank the Barcelona Supercomputing Center and the Madrid city council for their support during our research.

References

  1. 1.
    Real Decreto 102/2011, de 28 de enero, relativo a la mejora de la calidad del aireGoogle Scholar
  2. 2.
    PROTOCOLO DE MEDIDAS, A ADOPTAR DURANTE EPISODIOS DE ALTA CONTAMINACIN POR DIXIDO DE NITRGENO (2016). http://www.madrid.es/UnidadesDescentralizadas/UDCMedios/noticias/2016/01Enero/21Jueves/Notasprensa/Nuevo%20protocolo%20contaminaci%C3%B3n/ficheros/Protocolo_NO2.pdf. Accessed 5 Mar 2017
  3. 3.
  4. 4.
  5. 5.
    Baldasano, J.M., Jorba, O., Gass, S., Pay, M.T., Arevalo, G.: CALIOPE: sistema de pronstico operacional de calidad del aire para Europa y EspaaGoogle Scholar
  6. 6.
    Gong, B., Ordieres-Mer, J.: Prediction of daily maximum ozone threshold exceedances by preprocessing and ensemble artificial intelligence techniques: case study of Hong Kong. Environ. Model. Softw. 84, 290–303 (2016)CrossRefGoogle Scholar
  7. 7.
    Zheng, Y., Yi, X., Li, M., Li, R., Shan, Z., Chang, E., Li, T.: Forecasting fine-grained air quality based on big data. In: Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 2267–2276. ACM, August 2015Google Scholar
  8. 8.
    Michalakes, J., Dudhia, J., Gill, D., Klemp, J., Skamarock, W.: Design of a next-generation regional weather research and forecast model. Towards Teracomput. 117–124 (1998)Google Scholar
  9. 9.
    Guevara, M., Martnez, F., Arvalo, G., Gass, S., Baldasano, J.M.: An improved system for modelling Spanish emissions: HERMESv2. 0. Atmos. Environ. 81, 209–221 (2013)CrossRefGoogle Scholar
  10. 10.
    Byun, D.W., Ching, J.K.S. (eds.): Science algorithms of the EPA Models-3 community multiscale air quality (CMAQ) modeling system, p. 727. US Environmental Protection Agency, Office of Research and Development, Washington, DC (1999)Google Scholar
  11. 11.
    Prez, C., Nickovic, S., Pejanovic, G., Baldasano, J.M., Özsoy, E.: Interactive dustradiation modeling: a step to improve weather forecasts. J. Geophys. Res.: Atmos. 111(D16) (2006)Google Scholar
  12. 12.
    Li, X., Peng, L., Hu, Y., Shao, J., Chi, T.: Deep learning architecture for air quality predictions. Environ. Sci. Pollut. Res. 23(22), 22408–22417 (2016)CrossRefGoogle Scholar
  13. 13.
    Sainath, T.N., Vinyals, O., Senior, A., Sak, H.: Convolutional, long short-term memory, fully connected deep neural networks. In: 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 4580–4584. IEEE, April 2015Google Scholar
  14. 14.
    City council of Madrid open data portal. http://datos.madrid.es/portal/site/egob. Accessed 5 Mar 2017
  15. 15.
  16. 16.
    Dunlea, E.J., Herndon, S.C., Nelson, D.D., Volkamer, R.M., San Martini, F., Sheehy, P.M., Allwine, E.J.: Evaluation of nitrogen dioxide chemiluminescence monitors in a polluted urban environment. Atmos. Chem. Phys. 7(10), 2691–2704 (2007)CrossRefGoogle Scholar
  17. 17.
    Unidata, 2015: THREDDS Data Server [Version 4.6.2 - 2015–06-09T15:16:47–0600]. Boulder, CO: UCAR/Unidata Program Center. doi: 10.5065/D6N014KG
  18. 18.
  19. 19.
    CALIOPE evaluation reports. http://www.bsc.es/projects/earthscience/v_FCST_EVAL/files/new/NRT/KF/. Accessed 5 Mar 2017
  20. 20.
    Zaremba, W., Sutskever, I., Vinyals, O.: Recurrent neural network regularization. arXiv preprint arXiv:1409.2329 (2014)
  21. 21.
    Cooijmans, T., Ballas, N., Laurent, C., Gülçehre, Ç., Courville, A.: Recurrent batch normalization. arXiv preprint arXiv:1603.09025 (2016)
  22. 22.
    Parascandolo, G., Huttunen, H., Virtanen, T.: Recurrent neural networks for polyphonic sound event detection in real life recordings. In: 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 6440–6444. IEEE, March 2016Google Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  1. 1.Universidad Rey Juan CarlosMadridSpain

Personalised recommendations