Acta Geographica Sinica ›› 2021, Vol. 76 ›› Issue (10): 2477-2488.doi: 10.11821/dlxb202110010

• Urban and Regional Development • Previous Articles     Next Articles

Classification of shrinking cities in China using Baidu big data

MENG Xiangfeng1(), MA Shuang1, XIANG Wenyi2, KAN Changcheng2, WU Kang3, LONG Ying1,4,5()   

  1. 1. School of Architecture, Tsinghua University, Beijing 100084, China
    2. Times Technology (Beijing) Co., Ltd., Beijing 100085, China
    3. Beijing Key Laboratory of Megaregions Sustainable Development Modeling, School of Urban Economics and Public Affairs, Capital University of Economics and Business,Beijing 100070, China
    4. Hang Lung Center for Real Estate, Tsinghua University, Beijing 100084, China
    5. Key Laboratory of Eco-Planning & Green Building, Ministry of Education, Beijing 100084, China;
  • Received:2020-08-05 Revised:2021-07-25 Online:2021-10-25 Published:2021-12-25
  • Contact: LONG Ying;
  • Supported by:
    National Natural Science Foundation of China(51778319);National Natural Science Foundation of China(42171216);Beijing Nova Programme Interdisciplinary Cooperation Project(Z191100001119017)


In the context of rapid urbanization, shrinking cities cannot be ignored. The classification of shrinking cities according to the patterns of urban population changes is the key to reveal the "hollow" and changing characteristics of the population. Researchers classified the shrinking cities in Western developed countries as "perforated", "doughnut" and "peripheral" types. However, there is a lack of evidence showing the types of shrinking cities in China. This paper takes the redefined physical cities in China (excluding Hong Kong, Macao and Taiwan) as the study area, and employs Baidu Huiyan's 200 m × 200 m grid cell permanent population data from November 2016 to November 2018 to identify different types of shrinking cities in the country. Our results show that 1506 physical cities have experienced a population reduction between 2016 and 2018, accounting for 49.8% of the total number of physical cities (3022). Furthermore, in order to accurately identify the types of shrinking cities, 126 densely populated and severely shrinking cities during the study period were selected as the research samples, mainly located to north of the Qinling Mountains and Huaihe River of China. Our analysis shows that the shrinking types of the sample cities can be classified into five categories: "perforated", "complete", "local", "doughnut", and "peripheral". The number of cities with perforated shrinkage is the largest, which numbered 47 (37.3% of the sample cities), most of which are found in Heilongjiang Province, Inner Mongolia Autonomous Region and Gansu Province. Complete shrinkage ranks the second, including 33 cities, which accounts for 26.2% of the samples, and mainly distributed in Gansu and Shaanxi provinces. The numbers of shrinking cities with local, doughnut and peripheral types are 25, 18 and 3, respectively. These cities are mainly distributed in Heilongjiang and Inner Mongolia. This is because the governments paid more attention to the development speed over quality during the first half of China's urbanization process, and adopted a growth-oriented urban planning strategy. Traditional industries declined due to resource depletion or economic transformation, and the transfer of industries under active government adjustment, all of which contribute to the emergence of perforated shrinkage. This study could provide scientific evidence for micro management of urban development, optimization of land use pattern, improvement of urban livability, and conduct of urban planning and design in the context of urban shrinkage.

Key words: physical city, urban shrinkage, population change pattern, Baidu Huiyan, permanent resident, China