吉林省德惠市土壤特性空间变异特征与格局

摘要

Based on regionalized variable theory, semivariograms of geo-statistics were used to research the spatial variation of soil properties quantitatively. The results showed that the semivariogram of soil organic matter is best described by spherical model, the best model for semivariograms of soil total N and available K is exponential models and that of available P belongs to linear with sill model. Those soil properties have different spatial correlations respectively, the lag of organic matter is the highest and that of available P is the lowest, the spatial correlation of N and available K belongs to moderate degree. Spatial heterogeneities are different too, the degree of organic matter and total N are higher, the degree of available K is in the next place and that of available P is the lowest. Influenced by the shape, topography and soil of the study area, all isotropies of available P are obvious in all directions while anisotropies of others are manifested. According to the analytical results, supported by GIS, Kriging and IDW methods are applied to describe and analyze the spatial distribution of soil properties. The results indicate that soil organic matter, total N and available K are distributed regularly from northeast to southwest, while available P is distributed randomly.