空间统计学在食品污染物分布研究中的应用

Application of spatial statistics in studying the distribution of food contamination

目的以2017年某省食品安全监测大米中砷含量数据为例,探讨空间统计学方法在食品污染物分析中的应用价值。方法采用空间点模式估计、核密度分析,全局以及局部自相关性分析等空间统计学方法,在县级空间尺度下,对某省大米中砷含量进行探索性空间数据分析。结果空间点模式分布图显示,该省大米砷污染的空间分布比较分散,核密度分析结果显示污染热点区域主要在该省中东部地区。全局自相关Moran’s I指数值为0.11,有统计学意义,大米样品中砷污染呈现出低度空间聚集性。有1个"高-高"聚集区,2个典型的"低-低"聚集区。结论空间统计学运用于食物污染物分布研究上,可以很好地可视化展示、识别污染分布规律、热点地区和聚集区,为基于问题的监测工作的开展提供技术支持。

Objective Based on data related to arsenic contents in paddy rice, as part of the food safety monitoring programs in 2017, to discuss and explore the application of spatial analysis used for food safety risk assessment. Methods One province was chosen to study the spatial visualization, spatial point model estimation, and kernel density estimation. Moran’s I statistic of spatial autocorrelation methods was used to analyze the spatial distribution at the county level. Results Data concerning the spatial point model estimation showed that the spatial distribution of pollution appeared relatively dispersive. From the kernel density estimation, we found that the hot spots of pollution were mainly located in the central and eastern regions. The global Moran’s I values appeared as 0.11 which presented low spatial aggregation to the rice arsenic contamination and with statistically significant differences. One "high-high" and two typical "low-low" clustering were seen in this study. Conclusion Results from our study provided good visual demonstration, identification of pollution distribution rules, hot spots and aggregation areas for research on the distribution of food pollutants. Spatial statistics can provide technical support for the implementation of issue-based monitoring programs.