云贵高原湖区湖泊营养物生态分区技术方法研究

提出一种基于主成分分析、聚类分析、判别分析和空间自相关的分区模型,用于对云贵高原湖区湖泊营养物生态分区的研究.首先运用主成分分析方法对众多指标进行了降维综合处理,产生彼此互补相关又能综合反映湖区情况的4个新指标,累计贡献率达到93.69%,具有充分的代表性,构建了云贵高原湖区湖泊营养物生态分区指标体系.在此基础上,根据各区域新指标值,结合聚类模型初步将湖泊流域分为5类,再利用判别分析完成非湖泊流域的类别判别,最后运用空间自相关分析方法,按聚类结果进行全局自相关分析,Moran’sI为0.320,且检验值Z值为68.2,远大于临界值(显著水平1%所对应的临界值2.58),表明聚类结果与空间位置具有显著相关性,之后运用局部自相关对区域各因素的主成分综合得分的空间分布格局进行了量化分析,揭示了零散分类区块在空间地域分布上的关联和差异,根据关联结果完成最终的分区.结果表明,利用此分区模型可以尽量避免人为因素,得到更为客观的分区结果,具有良好的适应性和可行性,可为探索适合我国湖泊营养物生态分区的指标体系和分区技术方法,完成全国湖泊营养物生态分区和科学地制定我国湖泊营养物基准和富营养化控制标准提供技术支持.

Study on Ecoregion Techniques of Lake Nutrients in Yunnan-Guizhou Plateau Lake Regions

The primary objective of this research is to present an ecoregion model based on the principle components analysis (PCA), the cluster analysis (CA), the discriminant analysis (DA) and the spatial autocorrelation analysis (SAA), which were applied to establish the ecorigion of lake nutrients in Yunnan-Guizhou Plateau Lakes. First, the principle components analysis method was used not only to reduce the dimensionality of the feature space, but also to deal with new indexes. The accumulation contribution ratio of the four new indexes achieved 93.69%, which had the full representation. These complementary indexes which reflect the conditions of lakes were helpful to establish the index system of ecorigion of lake nutrients in Yunnan-Guizhou Plateau Lakes. Second, initial classification of lake drainage area was accomplished by new indicator values and cluster analysis which contained five classes, and then the categories of non-lake drainage area could be identified by the discriminant analysis. Finally, the cluster result was carried on the global statistics, Moran's I was 0.32 which was higher than 0, and the examination value Z was 68.28 which was higher than marginal value (remarkable level 1% correspond 2.58), it was indicated that the cluster result and the space position had the remarkable relativity, afterward spatial distribution pattern of the region was achieved by the quantificational analysis about over score of principle components and application of local spatial autocorrelation method. Then the ecorigion of lake nutrients was accomplished by the association and difference between different districts of the scattered regions' s spatial character. These results indicated that the ecoregion model initiated in the paper was appropriate and feasible to receive objective results and avoid artificial factors. Also it has provided a new method to explore the index system and techniques about the ecoregion of lake nutrients. Scientifically, the model was in favor of the foundation of lake nutrient criteria and eutrophication control standards in China.