基于风险普查建筑物隐患数据和夜间灯光数据的云南省盈江县人口分布精细化研究

(云南省地震局,云南 昆明 650224)

自然灾害风险普查; 灯光遥感影像; 人口空间化; 精细化; 盈江县

Refinement Study of Population Distribution in Yingjiang County Based on Hidden Danger Data and Nighttime Lighting Data of Risk Census
ZHENG Chuan,JIA Zhaoliang,XU Ruijie,LI Zhaolong,ZHUANG Yan

(Yunnan Earthquake Agency,Kunming 650224,Yunnan,China)

natural disaster risk census; nigthtime remote sensing images; population spatialization; streamline; Yingjiang County

DOI: 10.20015/j.cnki.ISSN1000-0666.2023.0045

备注

现有的在地震应急中人口空间化数据因制作方法、数据来源不同,导致人口空间化产品存在较大差异,人口空间分布数据无法满足震后地震应急期间数据需求。以云南省盈江县为研究区,基于第七次全国人口普查盈江县数据、第一次自然灾害风险普查数据和夜间灯光遥感影像数据,利用空间叠加法计算人口分布权重,结合面积权重得到盈江县100 m×100 m格网的人口空间化结果。经精度评定,研究区所有乡镇人口空间化相对误差绝对值均小于0.6%,与2020年乡镇街道人口统计数据的相关系数R2接近1。结果表明,结合乡镇尺度人口统计、夜间灯光影像和重点隐患调查等数据所构建的人口空间化模型,所获100 m×100 m格网的人口密度数据精度得到了显著提高。
Due to different production methods and data sources,the existing spatialized data of the earthquake emergency population have great differences in the consistency of spatialized products,and the spatial distribution data of the population cannot meet the data needs of the post-earthquake emergency period.This paper takes Yingjiang County as the research area,based on the seventh national population census,the first natural disaster risk census across China and light image data,uses the spatial superposition method to calculate the population distribution weight,and combines the area weight method to obtain the population of Yingjiang County with a resolution of 100 m spatialized results.This product better reflects the details of the actual distribution of the population in Yingjiang County.According to the accuracy evaluation,the absolute value of the relative error of the spatialization of the population of all townships in the study area is less than 0.6%.The results show that the population spatialization model constructed by combining multi-source data such as township-scale demographic data,night light image data,and the survey data of key hidden dangers can significantly improve the accuracy of the obtained 100 m-resolution population density dataset.