应用多元统计分析中的因子分析 ,以中国西部地区各县市级行政单元为例 ,阐述了土地利用结构类型之间的关系 ,主要因子在表征土地利用集约性上的物理意义 ,以及土地利用结构的综合数值表征方法。利用 GIS的空间分析技术 ,展示了因子分析综合得分值极为丰富的结构性信息 ,并分析了这些信息在宏观分类、分区中的作用。最后 ,结合土地利用率与土地利用结构的信息熵 ,探讨了综合数值方法在分析土地利用问题以及防治土地退化方面的应用。

Based on detailed investigation data of land use structure in 1996, taking 974 counties or county level cities in West China as examples, indicating method of synthetic value for land use structure was studied by using factor analysis and information entropy analysis. Factor analysis not only could be used to classify the types of land use structure, but also provided synthetic values on land use structure and information entropy. Factor analysis showed that factor 1 was closely related to the proportions of arable land, garden land, residential and factory mining land, and traffic used land; factor 2 was determined by the proportions of forestry and grassland; while factor 3 had close relations to the proportion of water area. Containing more information, factor 1 showed that proportions of arable land, garden, residential and factory mining land, and traffic used land were changing synchronously. However, the mentioned situation would be changed in future or in some local places at present. In other words, high proportions of garden land and construction land would reduce the proportion of arable land. After analyzing the relationship among synthetic value of factor analysis, land use degree and information entropy, division on land use structure and degree was made, which is helpful to analyze the problems in regional land use.