功能区视角下城市夜光与夜间“两抢一盗”犯罪的空间相关性——以H市市辖区为例

以H市市辖区为例，基于POI数据识别城市功能区，并结合NPP-VIIRS夜光遥感数据与夜间“两抢一盗”犯罪数据，构建双变量空间自相关分析模型，从整体及不同功能区角度探讨城市夜光与夜间“两抢一盗”犯罪的空间关系。结果表明：1）整体上，H市夜间“两抢一盗”犯罪形成“一主一副”的犯罪热点，且夜光与夜间“两抢一盗”犯罪存在空间异质性关系。2）大部分功能区在全局Moran's Ι上表现为正相关。在局部LISA聚类上，只有居住用地、商业服务业用地和道路与交通设施用地呈现显著的空间异质性。商业用地的高-高聚类多数分布在一环及二环以内，低-低聚类多数分布在城市外围及边缘地带，高-低聚类与低-高聚类则零散分布在二者之间的区域；居住用地的聚类情况呈现南北区域差异；道路与交通设施用地则由于报警地点与实际案发地点可能有所出入，因而结果可信度降低。

Spatial Correlation Between Urban Nighttime Light and Nighttime Robbery,Snatching, and Theft Crimes from the Perspective of Functional Areas: A Case Study of H City

With accelerated urbanization, urban nighttime crimes—especially robbery, snatching, and theft—have received increasing attention from scholars as the highest incidence of local criminal crimes. Related studies have pointed out that the phenomenon of nighttime robbery, snatching, and theft crimes in most cities show the characteristics of spatial clustering, which is similar to the distribution characteristics of urban nighttime lights. Simultaneously, the correlation between urban lighting and nighttime robbery, snatching, and theft varies in different functional areas. Recently, the Suomi National Polar-orbiting Partnership Satellite(NPP-VIIRS) nighttime light remote sensing data have greatly improved the accuracy of nighttime light remote sensing, providing a more accurate basis for the study of urban nighttime social behavior. The quantitative identification method of urban functional areas based on POI and programming languages has been gradually improved, as well. In view of this, a bivariate spatial autocorrelation model was constructed based on the 110 police data from October to November 2017 in H city. The correlation between urban lights and robbery, snatching, and theft crimes at night was explored from the perspective of the main urban area as a whole and different functional areas. According to the model, (1) from the perspective of the main urban area, nighttime robbery, snatching, and theft crimes in H city form a fixed crime hotspot of "one main and one vice," and the correlation between lights and nighttime crimes varies spatially. The results further confirm that the correlation between urban nighttime light and nighttime crime varies according to the nature of urban functions. (2) Both single and mixed sites show clustering characteristics on global Moran's Ι, indicating that the correlation between urban nighttime light and nighttime robbery, snatching, and theft crimes is real. (3) The results show that urban night light and nighttime robbery, snatching, and theft crimes show obvious spatial correlations in commercial and business facilities, and such correlations show large spatial differences, with high-high clusters distributed in the central area of the city and low-low clusters in the peripheral area; urban night light and nighttime robbery, snatching, and theft crimes in residential facilities show north-south differences. Road, street, and transportation facilities are characterized by high crime rates inside public transportation and high mobility, which may lead to discrepancies between the location of the alarm and the actual crime location, ultimately making the results less credible. The correlation between urban night-light and robbery, snatching, and theft crimes at night shows that an increase in urban economic vitality brings about the social problems of robbery, snatching, and theft crimes in cities, which is contrary to the original intention of urban development. Therefore, to limit robbery, snatching, and theft crimes at night from the perspective of urban development, the correlation between nighttime crime and urban vitality, and the analysis of the influencing factors, need to be explored in greater depth.