Revealing the dynamic relations between traffic and crowding using big data from mobile phone network

In this work, we use three sources of mobile phone data to monitor traffic between three neighboring small areas - three “Aree di Censimento” (ACE) - in the Province of Brescia. Two indicators aimed at capturing crowding and traffic intensity are defined and their rela- tionship is analyzed. Then, their impact on traffic flows is investigated. To this scope, traffic flows between pairs of ACEs are estimated by means of a vector autoregressive model with crowding and traffic intensity indicators as regressors. Moreover, seasonal components are included in the model as exogenous variables modeled as dynamic harmonic regressions. We find that our model always satisfactorily captures traffic flows, although the effect of the seasonal components varies considerably among the pairs.