Time-evolving distribution of time lags between commercial airline disasters

Abstract

We have studied the time lags between commercial line airplane disasters and their occurrence frequency till 2002, as obtained from a freely available website. We show that the time lags seem to be well described by Poisson random events, where the average events rate is itself a function of time, i.e. time-dependent Poisson events. This is likely due to the unsteady growth of the industry. The time lag distribution is compared with a truncated Tsallis distribution, thereby showing that the ''phenomenon'' has similarities with a Brownian particle with time dependent mass. We distinguish between ''other causes'' (or natural causes) and ''terrorism acts", the latter amounts to about 5 percents, but we find no drastic difference nor impact due to the latter on the overall distribution.

24 refs, 22 pages, 9 figures to be published in physica a