The Identification And Spatiotemporal Pattern Research Of Urban Polycentric Point In China Based On NPP-VIIRS Nighttime Light Data

City is the product of the human settlement developing into an advanced stage,and it is an important and productive area in modern society.Thus,over the last half century,rapid increase was a typical evolution characteristic of urbanization.During the 41 years of reform and opening-up,China's urbanization rate had increased from 17.9% in 1978 to 58.5% in 2017,with an average annual growth rate of 1%.It is expected that in 2030,China's urban population will reach 1 billion.As the improvement of urbanization level has brought the improvement of productivity,economic prosperity,the advancement of science and technology and the development of cultural education,a series of ecological environment and social problems have become increasingly noticeable,and there are even some 'urban diseases' in the congested central urban areas.The city is the fastest growing region in the transition period in China,and its construction can directly affect China's sustainable development process.Thus,the urban disease problem must be solved.After many years of theoretical and empirical research,it has been found that polycentric urban development models have a good effect in improving economic level and social cohesion,relieving the pressure of crowded population,land and transportation and improving the urban ecological environment.Therefore,with the demand of polycentric urban structure model,it is necessary to carry out synchronous monitor towards China's urban centers in a large spatial scale,analyze the dynamic evolution and spatial expansion pattern of different urban centers at different spatial scales,and quantify the characteristics of specific types of centers,and in the end,provide reference for China to formulate reasonable urban polycentric development strategy.Most of current research on urban polycentric region identification is based on statistical and point of interest(POI)data with good accuracy and representativeness.The high quality of these data results in high acquisition difficulty.Therefore,the researches based on such data are mostly limited to single-phase and individual large cities,so it is difficult to conduct dynamic and spatial comparative analysis.The brightness information captured by Suomi National Polar-Orbiting Partnership's Visible Infrared Imaging Radiometer Suite(NPP-VIIRS)sensor is highly correlated with human activities at night,and can reflect the spatial distribution and human activities in a large spatial scale,so it effectively compensate the shortcomings of statistics and POI data.However,as the China's urban polycentric region identification research based on NPP-VIIRS data is still in initial stage,it focuses more on the single year and lacks dynamic evolution analysis on time series.The threshold segmentation method identify the top 10% of the pixels with the highest brightness in the city as the urban center regions;however,the urban centers with the brightness of the light which is not belong in this threshold range is difficult to be detected.On the spatial scale,it just focuses on individual large cities,and lack some researches on a large number of medium and small sized cities.On the time scale,it prefers a particular single year,and lack dynamic evolution analysis in time series.Tailoring to the above three scientific problems,this dissertation bases on NPPVIIRS nighttime light data,and proposes a method to identify China's urban center points in a large spatial scale by spatial proximity and nighttime light intensity relationship.And it also analyzes the spatiotemporal pattern of urban centers through three scales between 2012 and 2017.The main research contents and conclusions of this dissertation are as following:(1)Proposes an urban polycentric region location identification method for massive cities in large spatial scale,based on NPP-VIIRS nighttime lighting data.Firstly,this research corrects the NPP-VIIRS data in China from 2012 to 2017 with reference to the previous research results.Secondly,learn from the method of extracting the peak point based on the Digital Elevation Model(DEM)data in the terrain analysis,the maximum neighborhood and difference algorithm is applied to the NPP-VIIRS data,for extracting the nighttime light peak pixels that identify potential location of the city centers.Thirdly,the method is used to determine the 25×25 appropriate window size for maximum neighborhood algorithm of all cities in national scale.Then,the peak pixels and urban land cover data of Moderate Resolution Imaging Spectroradiometer(MODIS)are superimposed and filtered,and the screening results are time-series corrected.Finally,this research obtained all the central points of county,prefecture and provincial level cities in 2012-2017.The results show that until 2017,there are 7,239 urban centers in 2,200 cities across the country,with an average of 3.3 urban centers per city,and about 68% of cities have significant polycentric structures.Through cross-validation with predecessor extraction results and high-precision remote sensing images,it is found that this proposed urban center identification method can easily and effectively identify the quantity and location of urban centers in large spatial scale.(2)On the three scales of economic region,grade and polycentric type cities,this research conduct a comparative analysis of the function and morphology of 2200 cities in China from 2012 to 2017,and describe the spatiotemporal pattern of urban center development at different scales.Describe the urban center's development level,regional differences,and functionality through the average intensity,coefficient of variation and population density,in order to analyze the multi-center characteristics of the city's functions.The multi-center characteristics of the city's morphology are analyzed by the degree of dispersion and aggregation of spatial distribution in the city center.The results show that the development level of urban centers in China had increased steadily,and the development differences between different urban centers have gradually narrowed.As for the development level and growth rate of the urban center,the pattern of eastern overweigh that of western,high-grade cities overweigh low-grade cities,multi-center cities overweigh single-center cities.In terms of central functionality,the pattern of regional scale and multi-center type urban scale is consistent with its development level and growth rate.However,for different grades of cities,the centers of small and medium-sized cities with fewer center quantities have stronger central functions than those with larger centers,showing a pattern of “small but fine”.In the northeastern region,most grade cities and multi-center cities,the urban centers in the northeast direction have a higher level of development.In addition,2013 is a year worthy of attention.The difference in the level of development of urban centers in the central and northeastern regions had suddenly increased during the year,and a large number of small city centers also had monitored significant human activity enhancements during the year,even at a rate that exceeds that of large cities.(3)The vacant problem of urban edge emerging center in 2013 is explored.Based on the inequality of land and population urbanization process,a method for assessing the severity of urban vacancy and its spatial distribution is proposed.And the spatial morphological characteristics of vacant urban areas are discussed at regional and urban scales.It is found that in 2013,China's urban vacancy situation is widespread,and around one-third of the urban built-up areas are vacant to varying degrees.The most vacant cities are mainly located in the northeast,north and western inland areas of China.The vacant urban areas in the eastern region are small and scattered,while the western regions are large and concentrated,with large areas of vacant land.Through comparison and verification,it is found that the proposed method for estimating the vacancy degree of urban areas is consistent with the previous research results.Finally,a comparative analysis is conducted between the vacant center and the transportation center,which are also located on the edge of the urban built-up area,but human activities are significantly higher than others.At the same time,the reasons for the formation of vacant centers are discussed through economic factors,city level,government planning,construction cycle and if main functions are clear.