The importance of statistical analyses for spatial data has grown in various scientific fields.Spatial lattice data is comprised of measurements or observations taken at specific locations or within specific regions.Cluster analysis based on some types of measures such as a similarity or defined distance has been performed in many fields.However,there are few approaches for cluster analysis of spatial lattice data.GIS (geographic information system) provides powerful tools to study the spatial lattice data,but it is very difficult to describe the spatial clustering based on its spatial structure.For example,a statistical map with shading is used to show how quantitative information varies geographically,but we can only find contiguous clusters in this map with the poor accuracy of visual decoding.