next up previous

Nonlinearity and prediction of air pollution

Milan Palus, Pelikan, E., Eben, K., Krejcir, P., Jurus, P.
Institute of Computer Science, Academy of Sciences of the Czech Republic
Pod vodárenskou vezí 2, 182 07 Prague 8, Czech Republic


A presence of nonlinearity in time series of concentrations of air pollutants and in their relations to time series of meteorological variables is tested using information-theoretic functionals and the surrogate data approach. The results are discussed in relation to predictability of the pollutant concentrations aimed to alert smog episodes.

In: Artificial Neural Nets and Genetic Algorithms. Proceedings of the International conference. (Ed.: Kurkova V., Steele N.C., Neruda R., Karny M.) - Wien, Springer 2001, pp. 473-476 (ISBN: 3-211-83651-9).

Milan Palus 2002