Taylor's power law for fluctuation scaling in traffic

TL;DR

This study applies Taylor's power law to traffic data in Minnesota, revealing variable fluctuation scaling exponents that depend on location, time, and traffic conditions, and compares these findings with cellular automaton models.

Contribution

It demonstrates the applicability of Taylor's power law to real traffic networks and explores its potential as an indicator of traffic phases.

Findings

The fluctuation scaling exponent varies with location, season, and traffic intensity.

Taylor's law can distinguish different traffic phases like free flow and jams.

The law's exponent is not a fixed characteristic of a network.

Abstract

In this article, we study transportation network in Minnesota. We show that the system is characterized by Taylor's power law for fluctuation scaling with nontrivial values of the scaling exponent. We also show that the characteristic exponent does not unequivocally characterize a given road network, as it may differ within the same network if one takes into account location of observation points, season, period of day, or traffic intensity. The results are set against Taylor's fluctuation scaling in the Nagel-Schreckenberg cellular automaton model for traffic. It is shown that Taylor's law may serve, beside the fundamental diagram, as an indicator of different traffic phases (free flow, traffic jam etc.).