Social Sensing of the Imbalance of Urban and Regional Development in China Through the Population Migration Network around Spring Festival

Ruoxin Zhu, Diao Lin, Yujing Wang, Michael Jendryke, Rui Xin, Jian Yang, Jianzhong Guo, Liqiu MengSustainability 2020 · DOI

TL;DR

This paper uses social media data to build a network of population movement during China's Spring Festival travel period. By analyzing this network, they reveal significant imbalances in urban and regional development across China, identifying key "rich-club" cities that drive national connectivity and highlighting areas in need of targeted development.

What Problem Does This Paper Address?

China's rapid economic growth has led to significant disparities in development between its regions. This imbalance fuels large-scale migration, particularly during the Spring Festival, the world's largest annual human migration. While traditional census data offers snapshots, it's infrequent. This paper aims to use novel "social sensing" techniques to capture real-time, dynamic population movements and use this data to quantitatively understand and visualize these regional development differences. The core challenge is to move beyond static economic indicators and leverage the rich information embedded in people's everyday movements to map development disparities.

Methods

The researchers propose a "geospatial network analytical framework." Here's a breakdown of their key methods:

  • Population Migration Network (PMN) Construction: They transform anonymized Sina Weibo user location data into a directed, weighted network where cities are nodes and the volume of people moving between them are edge weights. This is done for different periods of the Spring Festival travel (leaving, returning, and ordinary).
  • Urban Development Difference Analysis:
    • Relative Flow Difference Ratio (RFDR): This measures the imbalance of population inflow versus outflow for a city. A positive RFDR means more people are coming in, while a negative RFDR means more are leaving.
  • Attractiveness Index: This combines RFDR values from the leaving and returning periods to quantify how attractive a city is to migrants.
  • PageRank Algorithm: Adapted from web page ranking, this algorithm measures the "importance" of a city within the migration network, reflecting its influence and connectivity.
  • Spatial Interaction Analysis:
    • Community Detection (Infomap Algorithm): This method groups cities into "communities" that have denser migration flows within them than between them, revealing regional clusters.
  • Rich-Club Coefficient: This identifies "rich-club" cities – a small group of highly connected and influential cities that tend to interact more with each other than with less connected cities. This helps understand the backbone of the migration network.
  • Regional Importance and Attractiveness: They aggregate city-level metrics to evaluate the importance and attractiveness of these identified communities.

Datasets

  • Sina Weibo Data: The primary dataset used is anonymized geotagged posts from Sina Weibo users during the 2015 Spring Festival period. This includes data for approximately 360 cities across China. The authors note that this is a subset of available data due to access restrictions, but they validate its representativeness by showing a high correlation with urban population data.
  • China Statistical Yearbook 2015: Used for comparison with urban GDP data.
  • Administrative Division Data of China: Used for spatial analysis and mapping.

Key Insights and Ideas

  • Spring Festival as a Development Mirror: The annual Spring Festival migration, driven by family reunions, acts as a powerful, cyclical indicator of underlying regional economic disparities. The direction and volume of movement reveal where people are going for opportunity.
  • Social Sensing for Dynamic Urban Analysis: Using social media data offers a more dynamic and granular view of population movement than traditional census data, allowing for real-time analysis of urban development trends.
  • The "Rich-Club" Phenomenon in Migration: A small number of highly influential cities (like Beijing, Shanghai, Chengdu, Guangzhou, and Chongqing) form a core network that drives a significant portion of inter-regional migration. These cities act as hubs, connecting diverse regions.
  • Attractiveness is Contextual: A city's attractiveness isn't just about its own development level but is also influenced by the development of its neighbors. This highlights the interconnectedness of regional economies.
  • Community Structure Reflects Geography and Culture: The identified city communities largely align with geographical proximity and administrative divisions, but also show how cultural factors (like family reunions) can influence migration patterns during specific events.

Key Results

  • RFDR Analysis: During the leaving period, central, northeast, and northwest China show positive RFDR (more inflow), while developed areas like the Yangtze River Delta and Pearl River Delta show negative RFDR (more outflow). The returning period shows the opposite trend.
  • Attractiveness: Only about 17.8% of Chinese cities show positive attractiveness, indicating a severe imbalance. Three distinct attractive urban clusters are identified: the Yangtze River Delta, Pearl River Delta, and Beijing-Tianjin-Hebei region.
  • Importance (PageRank): Cities with high PageRank values (e.g., Beijing, Shanghai, Chengdu, Guangzhou, Shenzhen) are strongly correlated with urban GDP, suggesting this metric effectively captures economic development levels.
  • Rich-Club Cities: Five core rich-club cities (Beijing, Shanghai, Chengdu, Guangzhou, and Chongqing) were identified, accounting for 23.2% of total migration flow during Spring Festival. These cities are crucial for national connectivity.
  • Community Structure: 25 communities were identified for the ordinary period, largely mirroring provincial boundaries. During Spring Festival, some communities merge, reflecting increased long-distance travel.
  • Regional Imbalance: The study confirms a general east-to-west gradient of development, with western and northwestern regions showing lower influence and attractiveness, though some areas like Tibet maintain stable population flow due to cultural factors.

Limitations and Future Work

  • Data Limitations: The authors acknowledge that the Sina Weibo data is a subset of available data due to access restrictions. They also note that LBS data from other platforms might have different characteristics and limitations (e.g., hourly vs. daily granularity, lack of passenger information).
  • Temporal Resolution: While better than census data, the data is still tied to specific periods. The authors suggest that longer time series and higher spatial-temporal resolution data would be beneficial for a more complete understanding of urban development.
  • Attribute Information: The current analysis focuses on movement patterns. Future work could incorporate user attribute information (age, gender, education) and social media content to better understand the reasons behind migration (e.g., work, study).
  • Broader Applicability: The proposed methodological framework is presented as universal and could be applied to other fields like social networks, business trade, and traffic networks, as well as to analyze migration patterns in other regions (e.g., North America).
  • Targeted Poverty Alleviation: The study highlights the need for targeted government support in economically backward regions with limited external connections, such as Tibet.

Summary updated

Abstract

Regional development differences are a universal problem in the economic development process of countries around the world. In recent decades, China has experienced rapid urban development since the implementation of the reform and opening-up policy. However, development differs across regions, triggering the migration of laborers from underdeveloped areas to developed areas. The interaction between regional development differences and Spring Festival has formed the world’s largest cyclical migration phenomenon, Spring Festival travel. Studying the migration pattern from public spatiotemporal behavior can contribute to understanding the differences in regional development. This paper proposes a geospatial network analytical framework to quantitatively characterize the imbalance of urban/regional development based on Spring Festival travel from the perspectives of complex network science and geospatial science. Firstly, the urban development difference is explored based on the intercity population flow difference ratio, PageRank algorithm, and attractiveness index. Secondly, the community detection method and rich-club coefficient are applied to further observe the spatial interactions between cities. Finally, the regional importance index and attractiveness index are used to reveal the regional development imbalance. The methods and findings can be used for urban planning, poverty alleviation, and population studies.
BibTeX
@inproceedings{zhu2020social,
  author = {Ruoxin Zhu and Diao Lin and Yujing Wang and Michael Jendryke and Rui Xin and Jian Yang and Jianzhong Guo and Liqiu Meng},
  title = {{Social Sensing of the Imbalance of Urban and Regional Development in China Through the Population Migration Network around Spring Festival}},
  booktitle = {Sustainability},
  year = {2020},
  doi = {10.3390/su12083457},
}
Figure