Abstract: Discriminant model of urban zonal vegetation types in Northeast China was established with principal component analysis(PCA)and multiple group discriminant analysis(MGDA). The results showed that 210?7 matrix was established with variables of 210 cities(towns)and 7 climate factors. According to classification and ordination from principal component analysis, there were obviously 7 city(town)zonal vegetation types.Temperature, moisture and their integrate situation the main climate factor influencing distribution of city(town)zonal vegetation types. Discriminant function model of urban zonal vegetation types in Northeastern China was established with multiple group discriminant analysis, and predicted urban zonal vegetation types. Urban zonal vegetation types were validated, and 3 urban forest ecological construction models were put forward.The study provided scientific basis for urban forest ecological construction.

Key words: Fishery resources, Marine geographic information system, Spatial heterogeneity, Semivariogram, Spatial autocorrelation