空间权重矩阵对空间自相关的影响分析——以我国肾综合征出血热疾病为例

Influence of Spatial Weight Matrices on Spatial Autocorrelation:a Cased Study of HFRS in China

  • 摘要: 应用空间自相关统计方法,分析了2008年我国肾综合征出血热(HFRS)发病率的空间分布。采用多种权重度量计算了全局和局部两种相关性指数,分析了自相关数值对空间权重矩阵的依赖性。分析结果表明:①空间距离矩阵比空间邻接矩阵能更好地度量HFRS的空间分布;②使用空间距离矩阵时,当距离阈值500km<δ<800km时,全国发病率数据显示出显著的空间自相关性;③从局域上看,吉林省高值显著聚集,新疆、西藏、青海、广西和海南省自身低值被高值包围聚集显著。

     

    Abstract: With spatial autocorrelation statistics,we reveal the spatial distribution of incidence of in 2008.Because spatial weight matrix greatly affects spatial autocorrelation,various measurement matrixes are adopted to conduct global and local analysis.The results indicate that: ① the distance matrix is more powerful to describe spatial distribution of HFRS than the adjacency matrix.② The incidence of HFRS implies significant spatial correlation when the distance threshold lies within 500 km and 800 km.③ From a local perspective,high incidences are clustered significantly around Jilin Province,while Xinjiang,Guangxi,Qinghai and Tibet with low incidences are surrounded by the provinces with high incidences.

     

/

返回文章
返回