通过沉积岩芯元素含量垂直分布和富集因子分析,结合多元统计方法如系统聚类和模糊聚类分析,研究了太湖梅梁湾沉积岩芯元素地球化学记录的湖泊环境演化过程.以系统聚类分析和模糊聚类分析为基础,将梅梁湾地球化学记录的环境过程划分为如下几个层段:0-6cm为人类活动强烈干扰的阶段,25—50cm和55—72cm层段都是自然过程的表现,反映了流域物源的影响,而两层段元素记录的差异反映了不同自然作用驱动下的湖泊环境变化;6-25cm和50—55cm层段则是不同环境特征的过渡阶段,因此,综合应用多元统计方法可以更直观更精确地量化影响湖泊环境的因素,有助于恢复湖泊环境历史演化过程.
To evaluate the environmental change of Meiliang Bay of Lake Taihu, the elemental geochemical records of the sediment core were analyzed by the element enrichment factors and multivariate statistical methods like hierarchical cluster and fuzzy cluster. The results of element enrichment factors showed the enrichment of heavy metals in surface layers. Hierarchical cluster analysis made it possible to separate several groups of elements with different sources or geochemical behaviors. Moreover, fussy C-means clustering method was applied to divide geochemical records of the sediment into several segments and separate natural factors from anthropogenic. The enrichment of heavy metals and nutrients above 6cm depth was dominated by human activities. Both lower levels between 25 and 50cm depth and higher levels between 55 and 72cm depth of most elements had been driven by natural processes and been affected by catchment sources, and that discrimination between them indicated lacustrine environmental change driven by different natural processes. The transition periods on environment were reflected in other segments. Integrating these multivariate statistical methods is an efficient tool in achieving better understanding on the quantitative analysis for the complex factors that influence lakes, and helpful to reconstruct the history of environmental change in lakes.