随着减贫救助难度的加大，贫困人口日益呈现出空间集聚并向长期贫困发展的态势。而当下贫困的空间研究却显现得异常滞后；与此同时国内自然灾害事件的频发又成为致贫的关键影响因素。在这样的双重背景与视角下，本研究分别从空间数据分析、空间建模及时空动态预测与干预三个层面构造了贫困分析的空间取向研究框架，提出了扶贫与减灾空间一致性理论；利用多指标贫困占优MPDM方法构造了空间灾变贫困指数，并结合空间建模技术和Coupla分析方法等在省域尺度下对理论进行科学验证和问题挖掘，研究灾害风险驱动因子与贫困动态间的作用机制和影响过程；并利用神经网络BPNN方法和拟开发的Markov-Clue方法模拟与预测相应的灾害风险干预机制，从而为贫困和灾害的综合治理提供科学依据。本项目的研究是进行区域政策干预的基础，它对改进贫困瞄准机制、防灾减灾管理工作以及救灾物资的配置等都具有重大的理论和现实意义。

With the relief of poverty becoming more and more difficult, the current situation of poor population in China tends to show the charateristic of spatial agglomeration and long-term involvement. On one hand, spatial studies of poverty-related issues appear to be surprisingly lacking; but on the other hand, frequent natural disasters have already become one of the most important contributing factors of poverty. Against the above dual backgrounds, this study was designed to construct a spatial-oriented research framework (including spatial data analysis, spatial modeling and spatial dynamic prediction and intervention), put forward and prove the theoretical coherence of poverty alleviation and disaster reduction. Furthermore we want to have an empirical test of the related theory through inter-provincial spatial modelling techniques, multivariate poverty dominance methods（MPDM） and Coupla analysis method. By doing this, the mechanisms of poverty induced by Disaster and influencial process of contributing factors of poverty can be found out. In the end, simulations through neural networks ( BPNN ) and the Markov-Clue method may help us to comprehensively deal with both poverty alleviation and disaster reduction. And through regional intervention policy, it may carry great theoretical and practical significance of improving the aiming mechanism of poverty-leading facors, preventing and relieving disaster and the allocation of disaster-relieving materials.