Scaling Properties of Traffic-flow Data

By computing the probability distributions of the velocity difference between cars a time-delay τ apart, the scaling properties of traffic flow can be analysed. These data display scaling behaviour thus confirming earlier results that have found 1/ fα -noise in traffic flow. Furthermore, the applicability of the scaling analysis for describing the two-point statistics of traffic flow is demonstrated, leading to an additional test for the dynamical properties of microscopic and macroscopic traffic flow models.