City limits in the age of smartphones and urban scaling
Urban planning still lacks appropriate standards to define city boundaries across urban systems. This issue has historically been left to administrative criteria, which can vary significantly across countries and political systems, hindering a comparative analysis across urban systems. However, the wide use of Information and Communication Technologies (ICT) has now allowed the development of new quantitative approaches to unveil how social dynamics relates to urban infrastructure. In fact, ICT provide the potential to portray more accurate descriptions of the urban systems based on the empirical analysis of millions of traces left by urbanites across the city. In this work, we apply computational techniques over a large volume of mobile phone records to define urban boundaries, through the analysis of travel patterns and the trajectory of urban dwellers in conurbations with more than 100,000 inhabitants in Chile. We created and analyzed the network of interconnected places inferred from individual travel trajectories. We then ranked each place using a spectral centrality method. This allowed to identify places of higher concurrency and functional importance for each urban environment. Urban scaling analysis is finally used as a diagnostic tool that allowed to distinguish urban from non-urban spaces. The geographic assessment of our method shows a high congruence with the current and administrative definitions of urban agglomerations in Chile. Our results can potentially be considered as a functional definition of the urban boundary. They also provide a practical implementation of urban scaling and data-driven approaches on cities as complex systems using increasingly larger non-conventional datasets.