Multidimensional poverty measurement and spatial–temporal pattern analysis at county level in the arid area of Xinjiang, China

To quick and accurate identify poverty-stricken areas in the context of poverty eradication is a challenge that has troubled researchers. Based on the nighttime light data of NPP-VIIRS, this study takes three prefectures in South Xinjiang in China as the study objective to analyze the development and changes of concentrated contiguous poverty-stricken areas from 2012 to 2018. The multidimensional poverty index of administrative region of the study area and county is constructed, with the optimal light index selected through correlation analysis with different nighttime light indexes. Spatial visualization analysis of poverty differences in the study area from 2012 to 2018 was conducted. The landscape ecological index was used to analyze the nighttime light patches classified by DN (Digital Number) interval. The results show that: (1) There are significant differences in poverty among different administrative regions in the study area. From 2012 to 2018, the multidimensional poverty index (MPI) of all county-level administrative regions showed a significant growth trend, with corresponding decrease in poverty to varying degrees. (2) The fitting analysis of MPI and the three different light indexes show that MPI has the highest correlation with nighttime light area, and the administrative area with a larger nighttime light area develops better while the poverty level is lower in the administrative area with a smaller nighttime light area. (3) From 2012 to 2018, the nighttime light area in the study area increased steadily, and the growth area was mainly in villages and towns, indicating that the development level of rural and township areas where poverty mainly clustered was improved, and the possibility of poverty decreased.