Shrinking Cities And Their Spatial Shrinkage Patterns From The Perspective Of Natural Cities In Northeast China

In the context of China's decades of economic development and population growth,unique urbanization paths and patterns have developed.Localized urban shrinkage has brought a huge impact on the New Construction Urbanization in China.However,current research on urban contraction is mainly based on the framework of Western contexts and is not applicable to China's unique urbanization.Moreover,the definition of cities in China is significantly different from that of Western cities.In Chinese urban management and urban studies,many studies look at cities based on administrative perspectives and ignore urban space itself.All these have undoubtedly put China's urban contraction in a difficult position.Therefore,taking the natural city as the perspective,emphasizing the city itself,selecting typical districts,and conducting research on China's shrinking cities and their microscopic shrinkage patterns is an important path to construct a Chineseized shrinking city study.In this study,the Northeastern China was selected as the study area.The Northeastern China was once known as the "cradle of Chinese industry" and the "eldest son of the Republic".However,since the end of the last century,a large number of enterprises have been closed and workers have lost their jobs in the Northeastern China.Although the Northeastern China Revitalization Strategy proposed by the National government has promoted the socio-economic development of the Northeastern China to a certain extent in response to the special situation of the region,it is still declining compared to other regions in China.At present,there are serious problems such as brain drain and aging in the society of Northern China.Therefore,studying the urban shrinkage and its patterns in Northeastern China can provide a realistic and theoretical basis for the Chineseization of shrinking cities research.In this study,the natural urban space was extracted as the research object using land use data,and the micro-scale population change was depicted using World Pop data.Finally,the shrinking cities under each period were identified based on the population changes of the natural cities,and the spatial shrinkage patterns of the shrinking cities were classified by applying convolutional neural networks.The main research results and conclusions are as follows:(1)The natural cities were extracted and the dataset were formed.Based on the theoretical framework of urban entity space,the natural cities in Northeastern China were extracted using land use data combined with the actual situation.186 natural cities were obtained in 2000,233 natural cities were obtained in 2010,248 natural cities were obtained in 2020,and the final natural city dataset was formed.The results of natural city extraction show that the natural cities in Northeastern China have a significant "long-tail" effect,with large cities mainly concentrated in the axis of Harbin-Dalian railroad and some natural cities merging into larger cities in the process of urban expansion.(2)shrinking cities were identified from the perspective of natural cities.By combining World Pop data and analyzing the population changes within the natural cities,shrinking cities in Northeastern China were identified within the framework of China's urban shrinkage study.The urban population in the Northeastern China was growing slowly,and most cities even show a trend of population decline.From 2000 to 2010,126 shrinking cities were identified,accounting for 67.74% of the number of cities;from 2010 to 2020,146 shrinking cities were identified,accounting for 62.66% of the number of cities;from 2000 to 2020,123 shrinking cities were identified,accounting for 66.13% of the number of cities.(3)The shrinkage patterns were classified from the perspective of natural cities and the dataset were formed.Based on the theoretical framework of shrinkage pattern and the actual situation in the Northeastern China,the natural cities were used as the study area to observe the population changes within the natural cities.The shrinkage patterns were classified into seven categories in the theoretical framework: no shrinkage,complete shrinkage,local shrinkage,central shrinkage,peripheral shrinkage,unilateral shrinkage,and scatter shrinkage.Using convolutional neural networks,the shrinkage patterns of natural cities under each period were well classified and a shrinkage pattern dataset was formed.The results show that the population changes of each shrinkage pattern were very different,and the complete shrinkage dominated the urban shrinkage phenomenon in Northeastern China.(4)The spatial and temporal distribution and evolution of shrinking cities were analyzed.The spatiotemporal variation of shrinking and non-shrinking cities in Northeastern China is observed at the regional scale.The non-contracting cities were mainly concentrated in the south-central Liaoning region and the Harbin-Changchun urban agglomeration.The shrinking cities were mainly located in the "land edge" areas with poor natural conditions,i.e.,the Greater Khingan,Lesser Khingan,and Changbai Mountain areas.In addition,there was a "shrinkage zone around the central cities" on the periphery of the four central cities,and the existence of the four central cities attracted a large number of people from the surrounding small cities.(5)The formation mechanisms of shrinking cities and shrinkage patterns were explored.The driving mechanisms of shrinkage were explored,analyzed and discussed from macro,meso,and micro perspectives.The macroscopic perspective can be attributed to the homogeneous external and internal factors of cities in Northeast China.The mesoscopic perspective can be attributed to the polarization effect and the reorganization of population patterns resulting from the continuous development of urban clusters and metropolitan areas with central cities as the core.The microscopic perspective can be attributed to the lack of resilience of some cities.