Research On The Spatialization Of Economy Of Typical Areas In " Belt And Road " China Based On Night Light Data

As one of China's three major strategies,the “Belt and Road” is of great significance to China's economic development.Effectively understanding the economic development of the regions along the “Belt and Road” in China will contribute to economic strategic deployment and will be important for the “Belt and Road”investment and construction significance.GDP is an important indicator to show the economic development of a region,but traditional GDP research has many limitations,and at the same time can not express the intra-regional differences.With the development of geographic information technology,the spatialization of GDP provides a fast and accurate new research method and method.In this paper,based on the night light image data,an improved spatial model of GDP is constructed.Through the experimental study of the long-term sequence of the “Belt and Road” typical regions,the validity and accuracy are verified,and the GDP space of the “Belt and Road” typical regions distribution analysis.The main work of the thesis research is reflected in the following four aspects:(1)Correction of night light image data.The night light data mainly used in this article are DMSP / OLS and NPP / VIIRS night light data.Among them,for DMSP images due to their data discontinuities and saturation,the image correction method based on the constant target area method is used to perform mutual correction and desaturation Timely sequential correction;for VIIRS images,threshold extraction based on the economic intensity method is used and DMSP stabilized light images are used as masks to filter out background noise;finally,the two image corrections are tested and the correction results are better.(2)Extraction of built-up area based on night light images.Extraction of built-up areas is not only very important for solving urban-related problems,but also has important significance for the targeted research of the economy.This paper uses built-up area extraction to make the optimal light index and GDP parameters to make the GDP spatialization model have higher accuracy.The iterative threshold method is used to extract the built-up area,and the extracted area and its change diagram are obtained.Then,the optimized combination of the overall classification accuracy,Kappa coefficient,and landscape index method is used as the verification index of the extraction accuracy.It shows that the experiment of extracting built-up area based on night light data in this part of this paper is feasible and accurate.(3)Establishment of improved GDP spatialization model.In the past,the spatialization of GDP was mainly based on the regression model of a certain light index and GDP data.In this paper,the optimal light index and GDP parameters are selected,and an improved GDP spatialization model based on geographic regression weighting combined with the population grid data set is used for The GDP inversion obtained the grid spatial distribution density maps of GDP in each study area from 2000 to 2017,which solved the problems of spatial heterogeneity and GDP distribution in areas without light brightness in the traditional GDP spatialization research.The combined superiority and relative error model is used to test the accuracy of the improved model.Finally,the model was used to complete the spatialization of key regions along the“Belt and Road”(Fujian,Guangdong,Zhejiang)and analyze the spatial distribution of GDP.(4)Analysis of the spatial distribution of GDP in the study area.The analysis of the spatiotemporal change of GDP density in the study area shows that for Guangdong Province,regardless of the GDP density value and the growth rate of GDP density,the Pearl River Delta region is the highest,the northern Guangdong region is the lowest,and the eastern and western regions It is in the middle of the two.For Zhejiang Province,the economic level is decreasing from northeast to southwest,and the economic level in the eastern coastal areas is much higher than that in inland areas.It has performance between cities and within cities.For Fujian Province,it indicates GDP Contributed areas of GDP density are mainly distributed in coastal areas and form coastal economic belts,and their economic development levels are in a relatively unbalanced position among the three provinces.Finally,based on the long-term GDP spatialization research in each research area,the common conclusions about the spatial distribution and change of GDP were obtained.