Spatiotemporal trip profiles in public transportation reveal city modular structure

Abstract

Understanding urban mobility patterns within public transportation (PT) systems is key for cities to improve services and promote sustainable mobility. Exploring daily PT riders’ traffic flows using anonymized big data is a first step to analyze when and to where people travel. Nevertheless, assessing the degree to which PT usage patterns correlate with the spatial distribution of points of interest (POI) and subsequently affect the city structure is less attempted. Classic approaches use questionnaires and survey data that are costly and often yielding limited statistical significance. The purpose of this research is to understand associations between travel patterns of urban commuters and the functional organization of a city. To this end, we propose a network-constrained temporal distance measure for modeling PT rider travel patterns from smart card data; and further introduce a fully autonomous approach to describe the span of services available at catchment areas around metro stations. The end result is a detailed analytical prescription of spatiotemporal commuting patterns in the city of Lisbon as well as an analytical contextual information that enables us to understand the functional modular structure of urban facilities. Using the city of Lisbon as the guiding case study, the gathered results confirm the hypothesis that PT rider’s daily flows along the PT network of metro stations reveal a city modular structure.

Keywords

• Travel behavior
• Intelligent Transportation Systems
• Clustering
• Catchment areas
• Open data