摘要
为寻求适用于城市小型景观水体富营养程度的评价方法,基于对世纪公园、中山公园等15个位于上海市域的小型景观水体的水质和营养状况分析,探讨了城市小型景观水体富营养化程度综合评价的因子及标准,应用MATLAB软件建立了基于T-S算法的模糊神经网络评价模型。该模型选用叶绿素a(Chla)、总氮(TN)、总磷(TP)、高锰酸盐指数(CODMn),透明度(SD)5个评价因子,将城市小型景观水体的营养程度细分为8个等级,评价因子的数据都比较易得,使模型具可推广性。应用模型对75组水体样本进行评价,得到它们在各营养等级的分布状况如下:超富占3%,重富占4%,中富占20%,轻富占33%,中-轻富占19%,中营养占12%,贫-中营养占9%,贫营养水体为0。通过与其他3种常用评价方法进行对比,证明了该模型评价法的可靠性和优越性。该研究建立的评价模型能快速有效地判定水体的富营养化等级,操作简单,实用性强,较适用于城市小型景观水体的富营养化评价。
In order to make eutrophication evaluation for small landscape waters properly,and based on the analysis of waters and nutrition state of small sightseeing waters in Shanghai City such as Century Park,Zhongshan Park etc,the factors of comprehensive evaluation of urban small sightseeing waters were discussed.A fuzzy neural network model was established by using MATLAB.In this model,five evaluation factors（Chla,TP,TN,CODMn,SD） had been chosen and eight nutrition levels had been divided.The evaluation factors were easy to get,which was good for the extension of the model.Then,the model was used to evaluate 75 water samples for verification.The result showed that the distribution of the 75 samples in different nutrition levels was as follow： super euthrophic 3%,hyper-euthrophic 4%,meso-euthrophic 20%,light eutrophic 33%,meso-light eutrophic 19%,mesotrophic 12%,oligo-meso trophic 9%,oligotrophic 0%.Through comparison,this model evaluation method was proved to be reliable and better than other three methods.It can determine the eutrophication level for waters quickly and effectively.It was a simple and useful method for eutrophication evaluation for small landscape waters in cities.
出处
《安徽农业科学》
CAS
北大核心
2011年第23期14321-14325,共5页
Journal of Anhui Agricultural Sciences
基金
上海市科技兴农重点攻关资助项目[沪农科攻字(2007)第1-4号]
关键词
模糊神经网络
小型景观水体
富营养化评价
主成分分析
Fuzzy neural network
Small landscape waters
Eutrophication evaluation
Principle component analysis
