地理学报 ›› 2020, Vol. 75 ›› Issue (4): 769-788.doi: 10.11821/dlxb202004008

• 农业与乡村地理 • 上一篇    下一篇

中国农村深度贫困的空间扫描与贫困分异机制的地理探测

潘竟虎1, 冯娅娅1,2   

  1. 1. 西北师范大学地理与环境科学学院,兰州 730070
    2. 中国科学院西北生态环境资源研究院,兰州 730000
  • 收稿日期:2018-05-10 修回日期:2019-12-07 出版日期:2020-04-25 发布日期:2020-04-22
  • 作者简介:潘竟虎(1974-), 男, 甘肃嘉峪关人, 教授, 博士生导师, 中国地理学会会员(S110011899M), 主要从事空间经济分析与遥感研究。E-mail: panjh_nwnu@nwnu.edu.cn
  • 基金资助:
    国家自然科学基金项目(41661025);甘肃省高等学校科研项目(2016A-001);西北师范大学青年教师科研能力提升计划(NWNU-LKQN-16-7)

Spatial distribution of extreme poverty and mechanism of poverty differentiation in rural China based on spatial scan statistics and geographical detector

PAN Jinghu1, FENG Yaya1,2   

  1. 1. College of Geography and Environmental Science of Northwest Normal University, Lanzhou 730070, China
    2. Northwest Institute of Eco-Environment and Resources, CAS, Lanzhou 730000, China
  • Received:2018-05-10 Revised:2019-12-07 Online:2020-04-25 Published:2020-04-22
  • Supported by:
    National Natural Science Foundation of China(41661025);Project of Educational Commission of Gansu Province of China(2016A-001);Research Ability Promotion Project for Young Teachers of Northwest Normal University(NWNU-LKQN-16-7)

摘要:

从自然和社会经济因素中选取贫困的影响因子,建立评价指标体系,通过Pearson相关分析确定了贫困的主要影响因素,利用GIS空间分析和BP人工神经网络,分别模拟了省域、市域和县域3个尺度下的自然致贫指数与社会经济消贫指数,进一步计算获得了3个尺度下的贫困压力指数,对其空间分布格局进行分析;借助Flexible空间扫描探测识别出深度贫困县,在此基础上运用地理探测器揭示其主导致贫因素,并提出差别化的减贫对策建议。结果表明:① 省域、市域、县域3个尺度的自然致贫指数和贫困压力指数在空间分布上具有较高的一致性,呈现出明显的“东低西高”的空间分布格局;社会经济消贫指数的空间分布一致性不高,县域尺度更为破碎。以黑河—百色线为界,县域贫困压力指数在空间上表现出“大分散、小聚集”的分布状态。识别出的全国贫困县共计655个,主要分布在重点生态功能区和农产品主产区。② 3个尺度下,空间扫描识别出的贫困高风险区主要分布在西北、西南少数民族和边疆地区。县域尺度下识别出208个深度贫困县,多位于省际边缘区。③ 地理探测器识别出深度贫困县的7个致贫主导因素,按照主导因素将深度贫困县划分为地形要素制约型、区位交通制约型、经济收入制约型和生态环境制约型4类。

关键词: 深度贫困, 空间贫困, 空间扫描探测, 地理探测器, 分异机制, 贫困识别

Abstract:

Poverty has appeared as one of the long-term predicaments facing human development in the 21st century. The essence of extreme poverty is absolute poverty, where individuals experience long-term shortages of essential resources or suffer from harsh environment. Extreme poverty is the priority for poverty alleviation and the tough row to hoe. We select the major poverty influencing factors from natural and social factors to build an evaluation index system based on spatial poverty and related theories. First, we use Pearson correlation analysis to differentiate poverty impoverishing and alleviation factors. Then, we use GIS and back propagating neural networks to define a natural impoverishing index (NII) and social economic poverty alleviation index (SEPAI), respectively, at provincial, municipal, and county levels. We then calculate a poverty pressure index (PPI) at provincial, municipal, and county levels by combining NII and SEPAI, and explore poverty spatial characteristics. We used the flexible spatial scanning statistical method to identify the severely impoverished counties among the poverty-stricken counties with PPI>1.63, which had higher poverty rate and difficulty in poverty alleviation. Finally, we diagnose dominant factors that differentiate severely impoverished counties, and identify the dynamic mechanism of regional extreme poverty differentiation using the geodetector model. Besides, we construct a theoretical basis for anti-poverty in rural China. The results show that NII and PPI spatial distributions are highly consistent at provincial, municipal, and county levels, with a significant distribution pattern: high in eastern China and low in western China. In contrast, SEPAI has relatively low spatial consistency at provincial, municipal, and county levels. The PPI poverty distribution pattern tends toward large dispersion, small aggregation dividing across the Heihe-Bose Line. A total of 655 poverty-stricken counties were identified, mainly distributed in major ecologically functional and agricultural production areas in China. High risk areas identified by spatial scanning are mainly distributed in the northwest, southwest minority, and border areas. We also identified 208 severely impoverished counties, mostly located in inter-provincial fringe areas. The geodetector model identified seven dominant impoverishing factors, with significant differences between the four identified rural extreme poverty types: terrain detail oriented, location traffic dominated, economic income leading, and ecoenvironment constrained regions.

Key words: extreme poverty, spatial poverty, spatial scan detect, geodetector, differentiation mechanism, poverty identification