山地丘陵区社会经济数据空间化模型构建及应用
Building of a Spatialization Model of Socioeconomic Data in Mountainous and Hilly Regions and Its Application

社会经济数据的空间化，是地学研究中解决数据结构不一致、空间单元不匹配和数据在空间单元内均一化问题的有效方法.本文基于多因素回归分析建立了社会经济数据空间化处理方法，选取重庆市丰都县为研究区，以人口密度和GDP规模为对象进行了回归建模，为社会经济数据的空间精细化研究提供借鉴.结果表明，采取向后剔除法准则的多元线性回归分析数据融合方法，较好地模拟了研究区社会经济数据的空间分布格局，拟合方程修正后的可决系数均达到0.80以上；将以乡镇为单元拟合的县域社会经济数据融合模型，应用于村域尺度社会经济数据的空间精细化表达，经验证数据吻合度普遍在80%以上，能够较好地实现村域尺度社会经济数据的空间化.因此，就研究区而言，基于多因素回归分析的社会经济数据空间化模型构建，较好地实现了对社会经济数据的反演，为创建区域范围内村域尺度连续的社会经济数据表面提供了支撑.
In geoscientific research, spatialization of socioeconomic data is an effective way to solve such problems as inconformity of data structure, mismatching of space units and data homogenization in the space units. In a case study, this paper develops a method based on multiple regression analysis and makes regression modeling with population density and GDP of Fengdu County of Chongqing as research objects, so as to provide a reference for spatio-detailed research of socioeconomic data. The results show that in the study area the data fusion method based on multiple regression analysis with the criterion of 'backward regression' can well simulate the spatial distribution pattern of the socioeconomic data, and the revised coefficients of determination of all the fitted equations are more than 0.80. When the data fusion model fitted with town as the unit are applied to spatio-detailed presentation of socioeconomic data under the village-scale domain, spatialization of village's socioeconomic data is satisfactorily realized, the goodness of fit of data between the predicted value and the actual value being generally more than 80%. Therefore, for this study area, the spatialization model of socioeconomic data based on multi-factor regression analysis has a relatively precise inversion to socioeconomic data, and supplies a new way to build regional continuous socioeconomic data surface under the village-scale