中文摘要：

不同视角下多维贫困精准度量与分析是国内外扶贫领域的研究热点，本研究针对国家“中国农村扶贫开发纲要”中精准扶贫的要求，考虑到河池市当地的具体情况，在尽可能全面反映研究区现状的原则下选择了评价指标并建立评价模型，运用“双临界值”法和GIS地统计方法，通过维度“加和-分解”剖析了不同层面、不同自然与社会经济条件下的多维贫困空间分异特征。结果表明：①河池市多维贫困综合指数MPI呈现“周边高中间低”趋势，其中凤山、东兰、环江MPI最大，金城最小；②研究区主要致贫因素为危房、家庭健康、成年人文化程度低；③研究区贫困程度表现为西部聚集，中部和东部地区空间异质性明显；就致贫因素而言，研究区南部的大化、巴马，北部的南丹呈现显著聚集；④不同分类体系下，同一类型的类间多维贫困特征差异大，类内差异小；⑤研究区的MPI、A、H值与地表破碎程度呈正相关关系；H、A、MPI3项综合贫困指数随地表石漠化程度增大而增大，石漠化和贫困之间存在地理耦合性。

英文摘要：

The precise measure and analysis of multidimensional poverty from diverse perspectives have become a research hotspot in the field of poverty alleviation at both of the domestic and abroad levels. In order to fulfill the needs of precise poverty alleviation which were set by the national poverty alleviation＇ s ＂new compendium＂, we designed and completed this study. Combined with the specific conditions of Hechi city, we systematically designed a measure indicator system and evaluation template for the county-level multidimensional poverty from the perspective of rights poverty. This paper used the ＂double critical value＂ method and the GIS statistical method to analyze the spatial distribution characteristics of multi-dimensional poverty at different levels and different natural and social economic conditions through the ＂summation/decomposition＂ of dimensions. The results show that： first of all, MPI, which is the composite index of the multidimensional poverty in Hechi city, indicates a trend of ＂high in the peripheral area and low in the central area＂, in which Fengshan, Donglan and Huanjiang bear the highest index values and Jincheng being the lowest. Secondly, the major contributing factors of poverty in the research area include the dilapidated houses, the low family health condition and the low literacy of adults. Moreover, in terms of the poverty level, the western part of the research area indicates a significant agglomeration effect, while the middle and eastern parts indicate a significant heterogeneous spatial distribution; in terms of the causes of poverty, Dahua and Bama located in the southern part of the research area and Nandan located in the northern part show a significant agglomeration pattern. In addition, under different classification systems, the same-type inter-class multidimensional poverty reveals a huge discrepancy in its characteristics while the intra-class multidimensional poverty presents a smaller discrepancy. Lastly, there is a positive correlation between MPI