应用聚类分析划分孢粉带(区)的实例

APPLICATION OF CLUSTER ANALYSIS TO PALYNOLOGICAL ZONATION WITH EXAMPLES

一、序言随着计算机的普遍使用,孢粉分析学者对于数值方法在第四纪孢粉分析中的应用有了长足的认识。聚类分析、全成分分析、因子分析、回归分析、时间序列分析等多变量统计分析已用于处理第四纪孢粉学问题。应用这些方法划分孢粉带、孢粉分布区,进而探讨孢粉谱和植被、气候间的关系,定量解决第四纪孢粉学和古生态学的问题。在我国,虽有少数孢粉学者做了这方面的工作,但工作开展得还不够充分。本文是应用Q型聚类分析方法的一个实例。

Based on the results of Q-system cluster analysis, two pollen stratigraphical sequences from the cores in the Zhujiang Delta and the Hanhai Lake in Nei Monggol are divided into zones, while the surface samples from the South China Sea are separately attributed to different distributive areas, with a summary of the basical procedure of the system cluster analysis for polynological zonation. The zones derived from the system cluster analysis correspond closely to the concept of an assemblage zone.Using the system cluster analysis in palynological zonation can help us hold and balance large amounts of information, and gain insights into the structure of the data. In terms of mathematical criteria, this method avoids the possibility of bias on the part of an investigator; the similarity and difference among the zones in a sequence can be found by the system cluster analysis. However, due to the various decisions associated with the system cluster analysis with different threshold values, and the inevitable uncertainty about whether the decisions taken are the appropriate ones, it is suggested that one should not pitch too high any claim for the 'objectivity' of the system cluster analysis. But when several methods such as system cluster, fuzzy cluster and correspondence analysis are applied to the same data, similar results are obtained, thus providing the investigators with more confidence that the results suggested are not solely an artifact of the method employed.