Hierarchical Boundary Identification,pattern Analysis And Expansion Simulation Of Physical Cities

In the context of rapid global urbanization,cities have become the most important habitats for human beings.The various negative effects accompanying the urbanization make urban scientific research and the global urban agenda attract increasing attention.The spatial definition of a city is the foundation of urban scientific research and the global urban agenda,which is reflected in the fact that urban boundaries provide the necessary spatial basis for urban statistics,urban planning,and urban management,and is more related to the scientific understanding of urban development and evolution.At present,there are considerable differences in the definition of a city among different countries,which affects the unit comparability and result reliability for international urban studies.At the same time,rapid urbanization makes the spatial mismatch between the built-up boundary and the administrative boundary of the city more and more seriously,the overall and systematic understanding of the city has become increasingly fragmented and ambiguous.Without a unified and comparable urban spatial benchmark and a complete,hierarchical urban macro understanding,it is challenged to obtain reliable global knowledge of sustainable urban development.The physical city(PC)is defined according to the spatially continuous characteristics of the physical built environment of the city.It,together with administrative cities(AC,such as prefecture-level cities in China)and functional cities(FC,such as metropolitan statistical areas in the United States),are different ways of defining a city.PC can provide an objective and unified spatial benchmark for urban research,meanwhile,it can accurately capture the urban dynamics without being affected by administrative boundaries,and support the discovery of global urban scientific knowledge.However,the method for consistently identifying the boundary of PC worldwide is still unclear,and there is still a lack of systematic understanding of the differences in urban land characteristics and expansion patterns among cities in different countries.Therefore,this thesis focuses on the scientific issues of the unified delimitation of cities and the international comparison of their urban expansion.A systematic study is carried out on the boundary identification methods of PC,the current spatial differentiation of PC around the world,the historical expansion process of PC in major countries,and the future urban land simulation of Chinese PC,through a globally consistent perspective.The research seeks to discover the scientific knowledge of urbanization to support the development goals of sustainable urbanization.The main research work and conclusions are summarised as follows:1.A bottom-up method of hierarchical urban boundary identification for global physical cities,HUBM,is proposed.Based on HUBM,the vector data of hierarchical urban boundaries for global physical cities(i.e.,GHUB)from 1988 to2018 is produced and a dataset product is formed.The HUBM method mainly considers the density,size and spatial relationship of artificial impervious surfaces that characterize the physical built environment of a city.The boundary identification is implemented with the help of the remote sensing cloud computing platform Google Earth Engine(GEE)and the local GIS platform.Finally,three main urban boundaries of"Urban Extent(UE)","Urban Built-up area(UB)"and"Urban dense Centre(UC)"were identified,and so were three derived boundaries,i.e."Urban Open space(UO)","Urban Water body(UW)"and“Ex-urban built-up area(EB)”.The effectiveness and superiority of the HUBM method are proved through cross-product quantitative comparison.The product of global GHUB,is produced based on remote sensed-impervious surface product GAIA and water body product GSWL with a spatial resolution of 30 meters.The product includes hierarchical urban boundaries for five time slices in 1988,1990,2000,2010 and 2018,which provide the necessary spatial data foundation for a comprehensive,consistent understanding of global cities.2.The global distribution of physical cities presents significant regional differences and imbalances,with most physical cities concentrated in a few areas.There is a general spatial mismatch between the physical cities formed from the bottom-up and the spatial units that perform spatial management and control.The Rank-Size rule of the urban system shows universality.In 2018,a total of 10,242physical cities with an area of more than 5 square kilometres were identified around the world,and their distribution showed the characteristics of"more in the north"and"closer to the ocean".89.11%of the physical cities were in the northern hemisphere,and 39.18%of the physical cities were located within 100 kilometres distance from the coastline.The number of physical cities in Asia,Europe and North America accounts for 84.4%of the world.North America has the highest proportion of large cities,and the average area of European cities is the smallest.The United States(US)and China are the countries with the largest urban area(total UE area of 212,226 km~2)and the largest number of physical cities(2,175).More than half of the countries in the world have less than 10 physical cities.The results also show that UB in most countries conforms to the Rank-Size rule,and about two-thirds of them satisfy Zipf’s law at the same time.the urban polycentric structure in developed countries is more common.the urban built-up area in the US is more sprawling.3.Despite being at different stages of urbanization and economic development,the characteristics of urban land,population,and their coupling patterns of physical cities in China,the United States,and the European Union(EU)over the past three decades have shown some commonalities,e.g.,the per capita urban land consumption has generally increased,the polarization effect of population among physical cities increases gradually,and the sub-linear scaling law pattern of the built-up area remains stable.The pattern analysis of 2175 physical cities in China,1734 physical cities in the US and 1853 physical cities in the EU show that:(1)In terms of urban land,China’s physical cities are expanding rapidly,and they have caught up with the US to form a situation of“keep pace”.Although the centrality of urban land in China is higher than that of the US and the EU,but it is undergoing to be scattered,while the EU and the US are increasingly centralized.(2)In terms of urban population,the population density of China and EU has dropped rapidly,but China is mainly due to the faster growth of urban land,while EU is due to the slow population growth.The population ratio within UC of the EU and the United States continues to rise,and the phenomenon of"re-urbanization"deserves attention.(3)In terms of the coupling pattern of urban land and population,the physical cities in China and the EU are more compact than those in the US,while the EU cities have a more common polycentric structure than China.The new-formed physical cities in China are mainly low-density and entralised.It is necessary to focus on the"incremental development"of small and medium-sized cities in the future management and control of urban land.4.A national-scale urban expansion simulation framework NUES is proposed by distinguishing between the drivers and control units of urban expansion.Combining the historical and planning scenarios of urban expansion and global urbanization knowledge,the future simulation of Chinese physical cities is carried out,and it is found that the urban land scale gap between cities may tend to be polarized,and the urban spatial reconfiguration in plain areas may be intensified.Compared with the existing models,NUES considers the inconsistency of the bottom-up driving unit and the top-down control unit of urban expansion,and distinguishes different levels of spatial units for the calculation of land development potential and the simulation iterative process control at the national scale.It depicts the influence of different levels of spatial units in urban land expansion in more detail.The multi-scenario simulation results of China’s physical cities for 2030 show that the gap in urban land scale between cities may widen,the number of small cities will decrease,and the number of physical cities above the medium level will increase.At the micro level,the phenomenon of urban“spatial integration”in the plains area such as the Yangtze River Delta and Guangdong-Hong Kong-Macao Great Bay Area may be further intensified.In terms of policy implications,firstly,it is important to choose the optimal path for the scale distribution of the urban system of“coordinated development of large,medium and small cities”.Secondly,it is recommended to learn advanced experience and establish an urban spatial statistical zone system independent of administrative divisions to better implement urban spatial statistics and monitoring.