基于遥感影像及在线房租数据的城市内部贫困空间测度研究——以广州市内城核心区为例

学界以社会经济指标（人口普查、问卷调查等）为主导测度城市内部贫困空间，取得较好的研究成果；但是普查数据周期长、贫困群体社会经济数据可获得性低，如何制定全覆盖、多方位、易获取的测度指标是该项研究的主要挑战之一。本文尝试使用遥感影像、在线房租等公众可获取的数据，采用FETEX2.0、WEKA等软件，利用三个指标（土地覆盖指数、复杂度、单位房租）建构基于大数据的贫困指数，测度广州市内城核心区718个居委会的贫困得分。再将测度结果与利用第六次全国人口普查数据测度的贫困空间对比分析，探究两种测度方法的区别和适用性。结果显示：①两者的测度结果具有较强的一致性，尤其对前5%最贫困的居委会具有较高重合度；②本文建构的大数据贫困指数对城中村、传统工业区等典型贫困空间识别效果更好。本研究使用易获取、更新周期短的数据，有利于城市贫困空间分布的实时监测，对引导精准分配扶贫资源、有效实施贫困社区更新规划具有重要意义。

POVERTY MEASUREMENT OF URBAN INTERNAL SPACE BASED ON REMOTE SENSING IMAGES AND ONLINE RENTAL INFORMATION: A CASE STUDY OF THE CITY CORE OF GUANGZHOU

Academia has achieved remarkable results in poverty measurement mainly based on socioeconomic statistical data, such as census data and survey data. The development of wide coverage and multi-dimensionalpoverty indicators is one of the main challenges in urban poverty studies. By integratingonline rental information and remote sensing images from Google Earth and Landsat8, thisstudyproposesa BDPI (Big data poverty index), which is composed of three indicesin cluding land cover index, surface texture index and unit rent, to measure the spatial distribution of poverty in the city core of Guangzhou. 718 communities were included and a series of software such as FETEX2.0 and WEKA was applied in processes. Then the paper compares the result with the spatial distribution of multiple deprivation based on the Sixth national population census data, and study the differences between the two methods. The result shows that there is a high consistency between the two results, especially for the top 5% communities of poverty.