Construction Of Index System Of Regional Social And Economic Grid Data Set And Study On Spatialization Of Typical Indexes

As the most basic information resource to a country or a region, the socio-economic statistical data is the message body for the public to understand the national conditions and strength and social development. And, it plays an important role in making development policies on government management, economic development, scientific and education. However, as a traditional method, saved based on administrative region, the socio-economic statistical data has been gradually showing its limitations both in theory and in the application concept. With the development and maturity of3S technology,"spatialization of social data" has become a hot issue for geography and social science. Statistical data come from a certain statistical unit, but also can be allocated on a certain region of another space resolution. As geographical grid technology has been applied to space redistribution processing of socio-economic statistical data, it creates the conditions for the integration of natural geographic data and socio-economic data, and provides a new research ideas and patterns for the comprehensive study of natural and cultural. Construction of the regional socio-economic monitoring indicator system, based on geographical grid, will further broaden the application depth and breadth of statistics. It is an important research direction to spatialization of socio-economic data, that we explore the grid transformation technique, with strong operating, to statistical data. In this thesis, the research status quo of grid transformation of socio-economic data was reviewed and summarized. And this thesis did some research about, how to build the index system and the grid processing methods to the index, in the Loess Plateau of socio-economic indicators of the grid database system. The mainly content and conclusions are as follows:1) The socio-economic indicator of grid database on regional scaleThis study has summarized the system needs of research institutions and regulatory agencies on the regional socio-economic statistics. Based on some principles, it used a combination of qualitative and quantitative method, and reduced the dimension of the complex statistical indicators. An exploration of the regional socio-economic indicators based on a grid system unit has been built, the main design criteria includes population, resources and environment, economic activity and social life, and each criterion layer contains varying amounts of the element index.2) The treatment scheme of the spatialization of socio-economic statisticsBased on the analysis of the concept and requiring on the spatialization of socio-economic statistics, this study comes up with some conclusions on the method, that collect and grid transformation of socio-economic statistics. Each indicator data based on grid sets, should follow the priority order of "national source data-provincial and municipal sources of data-county-level source data", to be collected and collated. And it analyzed the general principles and processes of grid transformation of statistics.3) Demonstration of grid transformation model of demographic dataBased on previous studies, basic multi-source data fusion, in combination of demographic data, land use data, the settlements, roads and other strong indication factor, using spatial analysis and multiple regression analysis technique, the establishment of the study area population grid model were built at a scale of Ikm2grid-cells. In this study, the influence of each factor was analyzed by the qualitative analysis technique, the regression coefficient of each factor is corrected, so that the regression model is found in certain geographic extent rules. Then, according to demographic data at county level and demographic data weight above-mentioned, the construction of the population grid data set, with zero error at county scale, was achieved. According to the necessity of testify the precision of the result, the correlation analysis shows that the linear correlation coefficient between simulate result and demographic data at townships level is0.9177. The average error is5.35%, and the proportion is72%within the error range from-30%to30%. The results demonstrate that the result of grid transformation of demographic data generated in this study has a high precision.4) Demonstration of grid transformation model of GDP dataThe GDP was selected as key indicator of the economy class indicators. In combination of the nighttime satellite imagery (DMSP/OLS) and land use data, in accordance with three industry classifications, based on the multi-source remote sensing data, grid transformation model of GDP statistics of the Loess Plateau was constructed. The result of the first, second, third and total GDP of Loess Plateau at a scale of1km2grid-cells, has much more detail information of the spatial distribution. And according to the necessity of testify the precision of the result, the linear correlation coefficient between simulate result and GDP statistics at townships level is0.9, its accuracy could meet the needs of economic development, planning and other research fields at mid-scale. It makes the typical grid construction data set of GDP statistics come true.