| As a phenomenon related with society,economy,population and industry,urbanization has become a focusing question in urban research and also of vital importance to the development of human beings,increase of economy and ecological protection.Therefore,researches on urbanization,especially those related with urbanization development and dynamics,are quite important to the better understanding of urbanization process,monitoring the general trend of urban social development,clarifying the disadvantages of urban development,formulation of urban development policies and planning of urban space.Moreover,with the development of urban remote sensing and GIS,more and more researches related with urban development could be made from the spatial scales,especially the images of nighttime light,which make the detection much simpler,more direct and effective.However,due to the shortages of the different nighttime light images,there are still a few disadvantages related with the applications and technologies.Based on the thorough understanding of existing researches at home and abroad,in this study,we explored the relationship between nighttime lighting brightness and urbanization.According to the results,we further made some researches related with urban regions identification,dynamics monitoring,and spatial pattern detection and so on,details are bellows:(1)With the DMSP/OLS images from 1992 to 2013,we provided a pixel level detecting of the long-term dynamics and trends of urbanization of mainland China.The temporal variations of nighttime light were analyzed and the results showed that the regional development and urbanization of mainland China in the past two decades experienced drastic fluctuations with high slopes(with averaged CV of 0.1676 and averaged slope of 0.07).Such drastic fluctuations were mainly concentrated in eastern China,especially the coastal and metropolitan areas,while in regions such as NC,EC and SC,the total light variations,light dynamics and increasing trends were much more stable and significant than that in GNW and GSW.Moreover,with the sustainability estimating of nighttime light dynamics,the future distribution and trend of urban agglomerations were also detected.(2)We utilized a new urban region extraction strategy based on the edge-detection method which is widely applied in automatic digital images processing.Better urban regions extraction results were made which is more accurate and effective compared with other traditional methods(OA=93.1409%,Kappa=0.8755).The urban dynamics process and urban sprawl pattern of Zhejiang province from 1992-2013 were further detected by using the proposed method,and the new urbanization pattern of"center-axes" in Zhejiang was found,consisting of the center of Hangzhou,axes of Hangzhou to Ningbo,Hangzhou to Jinhua(Lishui),Hangzhou to Jiaxing,and Wenzhou to Taizhou.Moreover,with the detection of centers,rate and speed of urban sprawl,characteristics of urbanization of each citiy in Zhejiang province were further analyzed.(3)At the city level,we detected the inner structure and dynamics of urbanization.Based on the characteristics of VIIRS DNB nighttime images,we built an Urban Index model.With the established model,we further detected the inner structure of selected regions of Hangzhou,identified the urban centers and urban regions,as well as the spatial-temporal variation of it.We found that the urban cores in the study area present the form of "multi-center and scattered distribution",and the urban regions also present the overall spatial pattern of "multi-center,group-type and contiguous expansion".In addition,with a time series detection,more than 96%of the pixel were found with growing trend,which also indicating a significant increase of urban development of the selected region. |
PDF Full Text Request