论文部分内容阅读
The possible scale-invariant behavior in air pollutant concentration time series (APCTS) was investigated by means of some statistical tools: histogram, spectral analysis, fractal, and chaos.The time series exhibited the characteristic right-skewed fiequency distri bution, which could be well described with the lognormal distribution.Long-term memory and the possibility of scale invariance were first roughly identified through the analysis of autocorrelation.Scale invariance was clearly revealed with the aid of box-counting method and the box dimension was shown to be a decreasing function of the threshold magnitude level.