Forecasting of Air Pollution at Unmonitored Sites

  • Sofia Lopes
  • Mahesan Niranjan
  • Jeremy Oakley
Conference paper


We address the problem of forecasting air-pollution at a site where there is no monitoring station by constructing data-driven models. We assume synchronous measurements of pollution are available at other sites in the vicinity, and that the spatial correlation carries information relevant for prediction. A Gaussian Process type spatial model is assumed, for interpolating pollution, and the time variation of the hyperparameters of the GP model is considered. An illustration of the method on synthetic data is presented.


Gaussian Process Street Canyon Spatial Interpolation Markov Chain Monte Carlo Method Time Series Prediction 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    K. V. Mardia and C. R. Goodall, “Spatio-temporal analysis of multivariate environment al monitoring data” in Multivariate Environmental Statistics (G. P. Patil and C. R. Rao, eds.), North-Holland, 1993.Google Scholar
  2. [2]
    T. Follestad and G. Høst, “Spatial interpolation of ozone exposure in Norway from space-time data” in Geostatistics for Environmental Applications (J. Gomez-Hernandez, A. Soares, and R. Froideveaux, eds.), Kluwer Academic Publishers, 1999.Google Scholar
  3. [3]
    P. D. Sampson and P. Guttorp, “Nonparametric estimation of nonstationary spatial covariance structure” J. Am. Statist. Assoc, vol. 87, pp. 108–119, 1992.CrossRefGoogle Scholar
  4. [4]
    D. Higdon, J. Swall, and J. Kern, “Non-stationary spatial modelling” in Bayesian Statistics 6 (J. M. Bernardo, J. O. Berger, A. P. Dawid, and A. F. M. Smith, eds.), Oxford: University Press, 1999.Google Scholar
  5. [5]
    A. M. Schmidt and A. O’Hagan, “Bayesian in-ference for nonstationary spatial covariance structure,” Tech. Rep. 498/00, Department of Probability and Statistics, University of Sheffield, 2000. Submitted to J. Am. Statist. Assoc.Google Scholar
  6. [6]
    M. C. Kennedy and A. O’Hagan, “Bayesian callibration of complex computer models,” tech. rep., Department of Probability and Statistics, University of Sheffield, 2000. To appear in J. Roy. Statist. Soc. Sero B. Google Scholar
  7. [7]
    W. J. Krzanowski, Principles of Multivariate Analysis, a User’s Perspective. Oxford: University Press, 1988.Google Scholar
  8. [8]
    G. Matheron, “Principles of geostatistics” Economic Geol., vol. 58, pp. 1246–1266, 1963.CrossRefGoogle Scholar
  9. [9]
    R. Neal, “Regression and classification using gaussian process priors” in Bayesian Statistics 6 (J. M. Bernardo, J. O. Berger, A. P. Dawid, and A. F. M. Smith, eds.), pp. 69–95, Oxford: University Press, 1999.Google Scholar
  10. [10]
    J. Oakley and M. Niranjan, “Sequential interpolation of pollution monitoring data with particle filters” tech. rep., Department of Computer Science, University of Sheffield, 2000.Google Scholar
  11. [11]
    N. J. Gordon, D. J. Salmond, and A. F. M. Smith, “Novel approach to nonlinear/nostationary Gaussian Bayesian state estimation” IEE-Proceedings F, vol. 140, pp. 107–113, 1993.CrossRefGoogle Scholar
  12. [12]
    J. S. Liu and R. Chen, “Sequential Monte Carlo methods for dynamic systems” J. Am. Statist. Assoc., vol. 93, pp. 1032–1044, 1998.CrossRefMATHGoogle Scholar
  13. [13]
    A. Doucet, S. J. Godsill, and C. Andrieu, “On sequential Monte Carlo sampling methods for Bayesian filtering” Statist. Comp., vol. 10, pp. 197–208, 2000.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Wien 2001

Authors and Affiliations

  • Sofia Lopes
  • Mahesan Niranjan
  • Jeremy Oakley
    • 1
  1. 1.Department of Computer ScienceThe University of SheffieldSheffieldEngland

Personalised recommendations