Entangled communities and spatial synchronization lead to criticality in urban traffic

Abstract

Understanding the relation between patterns of human mobility and the scaling of dynamical features of urban environments is a great importance for today's society. Although recent advancements have shed light on the characteristics of individual mobility, the role and importance of emerging human collective phenomena across time and space are still unclear. In this Article, we show by using two independent data-analysis techniques that the traffic in London is a combination of intertwined clusters, spanning the whole city and effectively behaving as a single correlated unit. This is due to algebraically decaying spatio-temporal correlations, that are akin to those shown by systems near a critical point. We describe these correlations in terms of Taylor's law for fluctuations and interpret them as the emerging result of an underlying spatial synchronisation. Finally, our results provide the first evidence for a large-scale spatial human system reaching a self-organized critical state.

Spatiotemporal Context Awareness for Urban Traffic Modeling and Prediction: Sparse Representation Based Variable Selection

“…In [ 12 ], the experiments based on the traffic data achieved from thousands of sensors in Twin cities (a city in US) for one year show that the number of the sensors relevant to the prediction task at a given site is in general over 100, which indicates a global context for modeling and predicting the traffic flow at a given site. Shortly, it is further confirmed in the sense of complex network based experiments that traffic fluctuations in London are correlated over the whole city [ 13 ]. This means that we have to reconsider the problem of spatiotemporal correlations among traffic data from a big data point of view for the sake of prediction.…”

“…However, the rising of big data analytics brings in a new chance to revisit city dynamics from a novel point of view due to the massive data of human mobility available from taxi GPS traces [ 7 ][ 8 ][ 9 ], locations inferred from mobile phone positioning [ 10 ], web logs with geotags [ 11 ], and measurements from traditional traffic meters like loop detectors [ 12 ]. Recently, it is realized that traffic flows are coupled on a road network via mutual interactions such that the evolution of the traffic flow at each site is not independent to but constrained by those of the others at the whole city scale [ 12 ][ 13 ]. Correspondingly, the recent trend is shifted to making use of spatial-temporal correlations among observations at multiples sites to improve prediction.…”

The importance of antipersistence for traffic jams

“…To this end, universal characteristics of road networks and the according traffic patterns [2] can help to identify systemic bottlenecks [3]. While local traffic time series are best characterised with identifying different traffic states and state transitions [4,5], network aspects are well represented with fractal scaling laws [2,6,7]. Both aspects are grounded on empiric evidence in very diverse situations and are well understood with microscopic models.…”

Fluctuation Scaling, Taylor’s Law, and Crime

“…A TL relationship has been reported for Measles cases in the UK and the scaling was observed to change as the extent of vaccination increased [13] . Related scaling relationships have been observed outside of biology including such things as human interactions, stock trades, measures of firm size, and urban automobile traffic [14] , [15] , [16] , [17] , [18] . The origin and interpretation of the TL relationships has remained of interest over time, with particular effort directed to understanding the meaning of the exponent and the underlying mathematics [16] , [19] , [20] , [21] , [22] .…”

“…The origin and interpretation of the TL relationships has remained of interest over time, with particular effort directed to understanding the meaning of the exponent and the underlying mathematics [16] , [19] , [20] , [21] , [22] . Depending on the system studied, the size of the exponent and the model applied, it has been interpreted to indicate synchronization [17] , randomness and aggregation [8] , species interaction [22] , and multiplicative population growth [19] .…”