基于逐步聚类分析的水库浮游藻类生长预测.pdfVIP

中国环境科学 2015,35(9) ：2805~2812 China Environmental Science 基于逐步聚类分析的水库浮游藻类生长预测 常 淳 1*,冯 平 1,孙冬梅 1,张 凯 2 (1.天津大学水利工程仿真与安全国家重点实验室,天津 300072 ；2.天津 市水利科学研究院,天津 300061) 摘要：采用逐步聚类分析方法预测水库浮游藻类的生长情况, 以于桥水库为例,选取了 1999~2006 7 个水质和气象因子的 56 组数据建立 逐步聚类分析模型,其结果用于预测 2007~2010 年的叶绿素 a 含量.结果表明,实测值与预测值的变化趋势基本一致,相关系数 R 达到 0.94, 线性相关性较好.预测值的平均绝对误差为-0.0007,平均相对误差为21.66%.逐步聚类分析法可以快速准确的对水库的叶绿素a 含量进行有 效预测.对逐步聚类分析模型的敏感度分析表明,影响于桥水库藻类生长的主要因素是水体的 pH 值、溶解氧以及总磷,因此控制这 3 个因 素是预防藻类爆发的重要途径. 关键词：逐步聚类分析；水库；浮游藻类；预测 中图分类号：X524 文献标识码：A 文章编号：1000-6923(2015)09-2805-08 1* 1 Prediction of the alga growth in a reservoir based on the stepwise cluster analysis. CHANG Chun , FENG Ping , 1 2 SUN Dong-mei , ZHANG Kai (1.State Key Laboratory of Hydraulic Engineering Simulation and Safety, Tianjin University, Tianjin 300072, China ；2.Tianjin Hydraulic Science Research Institute, Tianjin 300061, China). China Environmental Science, 2015,35(9)：2805~2812 Abstract ：The stepwise cluster analysis (SCA) was employed to predict the chlorophyll a content of alga in Yuqiao Reservoir. A SCA model was firstly developed based on 56 data sets of 7factors related to water quality and meteorology from 1999 to 2006, which was used to predict the chlorophyll-a content from 2007 to 2010. The predicted values matched the measured values well with the average absolute error of 0.0007 and the average relative error of 21.66%, and the correlation coefficient is 0.94. It showed that SCA was effective to predict the chlorophyll-a content in Yuqiao Reservoir. The sensitivity analysis were also performed, the results showed that pH value, total phosphorus content and dissolved oxygen had more significant impacts on t