Research On Spatiotemporal Evolution And Driving Factors Of Built-up Areas In Changsha-Zhuzhou-Xiangtan Urban Agglomeration Based On Multi-source Dat

Since the reform and opening up,China’s urban construction has developed rapidly,and the continuous expansion of urban area has influenced the natural environment and people’s way of life.Therefore,the study of urban built-up area can not only help understand the state of urban expansion,but also help guide urban planning and construction,so as to realize the harmonious coexistence of human,city and nature.Changsha-zhuzhou-xiangtan urban agglomeration,as an important supporting point of the strategy of the rise of central China,is the key development core growth pole of Hunan Province,as well as the key reform test area that the country pays attention to.Taking it as the research object,the urban built-up area is extracted and its space-time evolution and driving factors are discussed,which can provide reference for the overall planning and development of Hunan Province.In view of the shortcomings of traditional statistical data and remote sensing data extraction of urban built-up areas,this paper takes Changsha-Zhuzhou-Xiangtan urban agglomeration as the research area and takes 2013-2019 as the time span to study the extraction and expansion rules of changsha-Zhuzhou-Xiangtan urban agglomeration supported by multi-source data.This paper firstly analyzes the advantages of three kinds of data and proposes a method of multi-source data combined with high-precision extraction of built-up area,and extracts the built-up area of Changsha-Zhuzhou-Xiangtan urban agglomeration from 2013 to2019.Secondly,the temporal and spatial evolution of the built-up area of Changsha-Zhuzhou-Xiangtan urban agglomeration is discussed from the perspectives of time series characteristics and spatial migration.Finally,by integrating the analysis results of single driving factors and multiple driving factors,the representative quantitative indexes with high correlation degree were selected to construct GM(1,N)model and BP neural network model to invert the built-up area of Changsha-Zhuzhou-Xiangtan urban agglomeration and evaluate its accuracy.The main contents and conclusions are as follows:(1)Propose a method of high-precision extraction of urban built-up area based on multi-source data.Based on the light solid area extracted from NPP-VIIRS nighttime light data,the light solid area extracted from POI data was combined to fill the light solid area.In addition,the main body range of urban buildings extracted from Landsat 8-OLS image data was introduced to carry out intersection filtering operation.Finally,the results of multi-source data combined with built-up area were obtained,and their accuracy was verified.The results show that the multi-source data extraction results not only coincide with the actual spatial location of the built-up area,but also have the dark area where night lights are missing and the economic weak area where POI is missing.The inner details of the built-up area are retained while the noise outside the built-up area is removed,which effectively improves the extraction accuracy.In addition,the area of the built-up area extraction results is highly correlated with the statistical area,which confirms the effectiveness of the built-up area extraction method based on multi-source data proposed in this paper.(2)Analyze the spatio-temporal evolution of the expansion of built-up areas in changsha-Zhuzhou-Xiangtan urban agglomeration at multiple scales.Based on the multi-source data and extraction results,this paper evaluated the expansion of the Changsha-Zhuzhou-Xiangtan built-up area from 2013 to 2019 from the perspectives of temporal characteristics and spatial migration,and explored the spatio-temporal evolution of the expansion of the built-up area at multiple scales.The results show that in terms of urban agglomeration scale,the built-up area of Changsha-Zhuzhou-Xiangtan urban agglomeration increased by 35.92 km~2 per year from2013 to 2019,showing a polynomial growth trend.The expansion rate and expansion intensity reached the peak value from 2017 to 2019,and the lowest value from 2015 to 2017.From 2013to 2019,the center of gravity migration direction of Changsha-Zhuzhou-Xiangtan urban agglomeration was northwest,and the migration speed was increasing year by year.The standard deviation ellipse was positively distributed from north to south,and the change of the long axis was small,but the short axis increased more.In terms of prefecture-level city scale,The built-up area of Changsha increased the most among the three cities from 2013 to 2019,and the speed and intensity of changsha’s expansion kept increasing,with the center of gravity moving to the northwest.In addition,the distribution of gravity center migration direction and standard deviation ellipse direction of Xiangtan city tends to be consistent with urban agglomeration.(3)The research system of driving factors for the expansion of built-up areas in Changsha-Zhuzhou-Xiangtan urban agglomeration is constructed.Through the analysis of single driving factors,several driving factors with high correlation and easy quantification were selected,representative quantification indexes were introduced,and the representative quantification indexes with high correlation were obtained by combining the grey correlation model.Then,the built-up area inversion model of Changsha-Zhuzhou-Xiangtan urban agglomeration was constructed by using GM(1,N)model and BP neural network model.The results show that traffic factors,economic factors and population factors have a direct impact on the development of the built-up area in the changsha-Zhuzhou-Xiangtan urban agglomeration in the short term.Economic factors are the main driving force of the expansion of the built-up area in the Changsha-Zhuzhou-Xiangtan urban agglomeration,while traffic and population are the secondary driving force.At the same time,the proportion of the tertiary industry in GDP,civil vehicle ownership,GDP,fixed asset investment and the expansion of the built-up area of the Changsha-Zhuzhou-Xiangtan urban agglomeration are all more than 0.7.In the subsequent established models,the BP neural network model has better simulation effect than GM(1,N)model,and the relative error of the BP neural network model predicted the built-up area from2013 to 2019 is generally less than 3%,with higher correlation coefficient and better stability.