关键词: 贫困, 时空数据, 时空统计, 空间自相关, 空间异质性, 空间贫困陷阱, 时空格局, 时空变化

Abstract:

Eliminating poverty is a common goal of human society. Poverty has the characteristics of spatial heterogeneity and spatial autocorrelation. Spatiotemporal statistical methods dealing with georeferenced or spatiotemporal data have been widely employed for analyzing spatiotemporal poverty data. This paper reviews the applications of spatiotemporal statistical methods in spatiotemporal poverty analysis and classifies the applications into four categories: (1) exploratory analysis of poverty, mainly to identify and quantitatively analyze the spatiotemporal distribution pattern of poverty; (2) identification of spatial determinants of poverty, to analyze the influencing factors of poverty by constructing a model of the relationship between poverty and various geographical elements; (3) spatial mapping of poverty, to obtain the distribution of poverty in the entire region using sampling data; and (4) spatiotemporal analysis of poverty, to reveal the spatiotemporal changes of poverty and their driving factors. On the basis of explaining the principles of these methods, we give examples of recent applications to illustrate how specific spatiotemporal statistical methods are applied to spatial poverty research. On this basis, the shortcomings of current spatiotemporal poverty research and potential development on future poverty research are also summarized.

Key words: poverty, spatiotemporal data, spatiotemporal statistics, spatial autocorrelation, spatial heterogeneity, spatial poverty traps, spatiotemporal pattern, spatiotemporal change