Identifying sinks and sources of human flows: A new approach to characterizing urban structures

TL;DR

This paper introduces a novel method using combinatorial Hodge theory to identify city centers as sources and sinks of human mobility, providing deeper insights into urban structures beyond traditional trip count statistics.

Contribution

It presents a new approach to analyze human mobility data by extracting scalar potential fields, enabling the detection of both origins and destinations of trips in urban areas.

Findings

Successfully identified sinks and sources in Tokyo's mobility data.

Classified city centers based on mobility patterns.

Applicable to various datasets for urban spatial analysis.

Abstract

Human flow data are rich behavioral data relevant to people's decision-making regarding where to live, work, go shopping, etc., and provide vital information for identifying city centers. However, it is not as easy to understand massive relational data, and datasets have often been reduced merely to the statistics of trip counts at destinations, discarding relational information from origin to destination. In this study, we propose an alternative center identification method based on human mobility data. This method extracts the scalar potential field of human trips based on combinatorial Hodge theory. It detects not only statistically significant attractive locations as the sinks of human trips but also significant origins as the sources of trips. As a case study, we identify the sinks and sources of commuting and shopping trips in the Tokyo metropolitan area. This aim-specific analysis leads to a combinatorial classification of city centers based on the distinct aspects of human mobility. The proposed method can be applied to other mobility datasets with relevant properties and helps us examine the complex spatial structures in contemporary metropolitan areas from the multiple perspectives of human mobility.