Chaos in Air Pollutant Concentration (APC) time series

Abstract

Three chaotic indicators, namely the correlation dimension, the Lyapunov exponent, and the Kolmogorov entropy, are estimated for one-year long hourly average NO (nitrogen monoxide), CO (carbon monoxide), SO2 (sulfur dioxide), PM10 (particles with an aerodynamic diameter of approximately 10 μm or less), and NO2 (nitrogen dioxide) concentration to examine the possible chaotic characteristics in the air pollutant concentration (APC) time series. The presence of chaos in the examined APC time series is evident with the low correlation dimensions (3.42-4.71), the positive values of the largest Lyapunov exponent (0.128-0.427), and the positive Kolmogorov entropies (0.628-0.737). Since the existence of multifracial characteristics in the above time series has been confirmed in our previous investigations, the presence of chaotic behavior identified in the current study suggests the possibility of a chaotic multifractal approach for APC time series characterization. Some problems concerning the applicability of chaos analysis in air pollution are also discussed.