Multiscale multifractal analysis on air traffic flow time series: A single airport departure flight case.

Multiscale multifractality of air traffic flow volume time series characterizes the operation and the evolution characteristics of the air traffic system in the temporal dimension, which are helpful in gaining a better understanding of the dynamical evolution mechanism of this system. We identify the airport departure air traffic flow volume time series is of multiscale multifractality via the investigation on the Hurst surface obtained by the multiscale multifractal analysis. The resulting Hurst surface is of distinct local fluctuation characteristics and shows distinct fractal characteristics on different scales. We confirm that the two types of multifractality both present in the time series, with the long-range correlation being the main cause of the multifractality. We extend multifractal spectra to multifractal spectrum surfaces to provide a common multiscale framework for exploring the multifractal strength, the dominant fluctuation type and the insensitivity of the multifractality. Furthermore, we find that we can use the time series of only 101 days, saving us 116 days, to obtain a quite similar Hurst surface whose difference to the original one is only 2.11662−7. • We identify the air traffic time series of the airport is of multifractality. • The main cause of the multifractality is the long-range correlation. • We obtain multifractal spectrum surfaces of regular shapes. • The minimum length of time series for effective multifractal analysis is identified. [ABSTRACT FROM AUTHOR]