基于动态聚类分析和盲数理论的综合营养状态指数评价模型  

Improved assessment model for comprehensive trophic state index based on dynamic cluster analysis and blind theory

作　　者：李镇镇[1,2] 李晓东[1,2] 李飞[1,2] 高智花 严文峰[1,2] 梁婕[1,2] 曾光明[1,2] 

机构地区：[1]湖南大学环境科学与工程学院,长沙410082 [2]湖南大学环境生物与控制教育部重点实验室,长沙410082

出　　处：《环境工程学报》2015年第4期2021-2026,共6页Chinese Journal of Environmental Engineering

基　　金：国务院三峡委员会项目(SX2010-026);国家自然科学基金资助项目(61173106;51009063;51039001);湖南大学青年教师成长计划

摘　　要：考虑到常用水质富营养化评价模型的单一、确定性缺点,尝试将动态聚类分析与盲数耦合理论引入水环境富营养化评价领域,建立了改进的综合营养状态指数评价模型,最后借助克里格插值法对评价结果的可信度进行空间可视化表征。利用改进模型对2009年洞庭湖区域的水环境空间的富营养化状况进行了评价,结果表明,东洞庭湖、扁山、鹿角富营养化程度相对严重,隶属于富营养化的可信度分别达到了0.5444、0.5318和0.5260。与确定性评价模型对比分析表明,所建改进模型更充分地考虑到被观察个体的差异,集成了更全面的样本信息,并可同时计算出了各采样点综合营养状态指数的区间数与其相对应的可信度水平,从而有效地提高了评价模型的可分辨性。Take consideration of shortcomings in the frequently-used eutrophication assessment model,an improved comprehensive trophic state index was developed based on blind theory and dynamic cluster analysis.Through the improved comprehensive trophic state index,eutrophication state of water environment from the Dongting Lake in 2009 was evaluated as a case,and the corresponding results were spatially expressed employing the kriging interpolation. The results showed that the east Dongting Lake,Bianshan and Lujiao were at relatively serious eutrophication degree with confidence level of eutrophication at 0. 5444,0. 5318 and 0. 5260,respectively. Compared with results of the deterministic assessment model,analysis indicated that the improved comprehensive trophic state index reduced the drawbacks of the procedure of certain assessment model efficiently. Improved assessment model could carry more information of samples,and the assessment results provided not only interval number of comprehensive trophic state index but also their corresponding confidence level which made a basis for the scientific decision-making.

关 键 词：动态聚类分析 盲数 综合营养状态指数 富营养化 

分 类 号：X824[环境科学与工程—环境工程]