Multisource methodology for traffic analysis zone definition based on the fusion of Remote Sensing, OpenStreetMap, and Floating Car Data

The choice of an appropriate Traffic Analysis Zone (TAZ) system is a critical step of travel demand modelling that is often overlooked. Studies have approached this process with the goal of favouring the homogeneity of socio-economic and geographic characteristics of the zones, also possibly taking into account the minimisation of the number of intrazonal trips. However, beyond the application of general guidelines and individual experience to specific case studies, the definition of a formal approach is still an unsolved issue. Nevertheless, the rapid ICT development and the novel big data sources allow to enhance traditional models by exploiting additional land use and spatio-temporal mobility features. This paper proposes a multisource data-driven method to support TAZ definition by identifying, through a clustering approach, zones that are homogeneous from the point of view of activities, network characteristics, and land-cover. To this end, in the proposed approach, satellite remote sensing image segmentation, OpenStreetMap layers, and Floating Car Data (FCD) are jointly exploited to define a TAZ configuration, which can be directly used as a support for planning purposes. The procedure is experimentally validated with a case study associated with the EUR district of the city of Rome, Italy, using satellite Sentinel-2 imagery, the corresponding OpenStreetMap data, and an FCD set containing more than 1.500.000 trips. • This paper proposes a multisource methodology to identify Traffic Analysis Zones (TAZs). • Remote Sensing data are used to identify areas with homogeneous distribution of the spatial features. • OpenStreetMap data are used to cluster areas sharing common characteristics (e.g., activities and network characteristics). • Floating Car data are used to populate the origin-destination matrix corresponding to the output TAZ. • The proposed methodology is experimentally validated on a case study in the city of Rome, Italy.