【作者】 范良千；

【导师】 吴祖成； 张清宇；

【作者基本信息】 浙江大学 ， 环境工程， 2011， 博士

【摘要】 为指导流域污染负荷削减,实现流域水环境污染物总量控制,本文围绕水体富营养化程度识别,污染物及其来源定量分析技术和非点源污染物总量负荷分配体系进行了研究,从而为辨识流域污染控制方向,确定污染源,明确污染控制任务提供科学技术方法体系。首先,以指数型白化权函数替代线性白化权函数改进了灰色聚类综合评价模型。以改进模型、原模型和梯形白化权函数模型对太湖2001-2009年水质营养程度进行评价,并比较结果。其次,基于克里格插值分析水体水质空间分布情况后,利用聚类分析、因子分析及Unmix受体模型构建水体污染物来源解析技术体系。采用太湖局部区域水质数据,分析主要污染来源及贡献。最后,借鉴生态评价模型构建社会-经济-环境三方公平的污染物总量负荷二次分配技术体系。以苕溪流域为例,利用产污系数模型预测流域2015年农业非点源化学需氧量(COD)、总氮(TN)、总磷(TP)的总量负荷,在改进的灰色聚类模型确定COD、TN和TP的农业非点源总量负荷削控目标条件下,将削控的农业非点源总量负荷分配至10个不同产污行业。获得的主要结果如下：①基于改进的灰色聚类综合评价方法获取的示例水体2001-2009年富营养化评价结果均为富营养等级。原模型法获取的结果为中-富营养等级,只有2006年为富营养等级。梯形白化权函数的模型方法结果与改进方法相一致。参照原始监测数据,改进方法是有效的,且解决了原方法的零权重问题。②2005年示例区域克里格空间插值结果为2、5、8月CODMn、TP浓度均从湖岸向湖中心逐渐降低,11月其浓度从湖岸向湖中逐渐降低后又有所增加,2、5、8、11月TN、Chl-a浓度均从湖岸向湖中心逐渐降低。结果反映出水体主要富营养化指标存在浓度分布的时空差异。③经聚类分析可将示例区域监测点划分为4个类别,各类别监测点位CODMn、TP、TN、Chl-a浓度变化与空间插值获得的变化趋势一致,且各类别监测点位富营养化均属P控制；各类别监测点位监测数据经因子分析显示其水质均可用3方面主要污染因素进行解释,且所有类别点位的3方面主要污染因素均涉及农业非点源；Unmix受体模型解析结果显示农业非点源对在前3类监测点位的氮、磷营养盐贡献大(Ⅰ类点：TN、TP、氨氮(NH3-N)：81%、85%、79%；Ⅱ类点：TN、NH3-N、磷酸盐(Po43-)：53%、90%、89%；Ⅲ类点：TN. NO3-N:34％.84％).④以产污系数模型预测的2015年苕溪流域的COD. TN. TP农业非点源负荷总量分别为：35383.11、11337.83、705.36吨。由于产业及土地利用差异,流域内各区县农业非点源负荷总量最大来源产业并不一致,但最小来源产业均为家兔饲养。各行政区域中,COD.TN单位面积负荷量最高在长兴,最低在临安；TP单位面积负荷最高在德清,最低在临安。⑤经水体浓度阈值估算,结合苕溪流域水环境容量,为使下游湖泊受纳水体达中营养等级,2015年苕溪流域农业非点源需削控COD. TN. TP的负荷总量分别为：9150.29、5047.37、116.52吨。以此为目标下的流域一次地域分配中,COD、TN、TP在吴兴区削控率最大,分别为：49.80%、91.27%、40.84%,德清县COD、TP削控率最小,分别为：13.88%、5.78%,TN在长兴县削控率最小,为23.62%；二次分配中,COD在流域中的茶桑果园、农村生活、鸡鸭饲养、生猪饲养污染源分配最大,TN在农村生活污染源分配最大,TP旱地、水产养殖、水浇地、生猪饲养污染源最大。研究建立的方法体系能准确辨识水体富营养化程度,简便、快速的获取主要污染源及其对水体污染物的贡献,实现污染物总量负荷公平分配。更多还原

【Abstract】 To guide abatement of pollution load and achieve water environment total amount control of pollutants in watershed, this thesis focused on identification of the eutrophication degree, source apportionment of pollutants and waste load allocation of non-point source system, so as to provide scientific metods for clarifying the direction of pollution control, indentifying pollution source and clearing the pollution control tasks.Firstly, through replacing the linear whitening weight function with the exponential whitening weight function, the grey clustering comprehensive aseesment model was improved. The nutrition status of water quality of Lake Taihu from 2001-2009 was assessed with the improved model, the original model and the model based on the trapezoidal whitening weight function model and the assessment results were compared with each other. Secondly, after analyzing the spatial distribution of the main pollutatns in water body with the kriging interpolation method, clustering analysis, factor analysis and the Unmix receptor model were used to establish source apportionment technology system of water pollutants. Taking the local area of Lake Taihu as case study area, the main polltion sources and the contribution of the sourecs were analyzed. At last, drawing on ecological assessment model, the secondary technical system of waste load allocation based on socio-economic-environmental justice principles was established. Taking the watershed of Tiaoxi River which the agricultural non-point pollution of is serious as example, after predicting the total agricultural non-point source loads of chemical oxyen demand (COD), total nitrogen (TN) and total phosphorus (TP) and determinating control target of the total agricultural non-point source loads by deriving the concentration thresholds of COD, TN and TP with the improved grey clustering model, the total agricultural non-point source loads of COD, TN and TP which need to be abate were allocated in different 10 industries. The results showed as the following:a) For the water qulity of sample, the eutrophication assessment results with the improved grey clustering method were the eutrophication classification from 2001 to 2009. However, the eutrophication assessment results with the original method were the mesotrophic classification from 2001 to 2009, except the assessment result was the eutrophication classification in 2006. The assessment results with the grey clustering method based on the trapezoid whitening weight functions were consist with those with the improved method. Referring to the raw monitoring data, the improved method was more effective. Furthermore, the improved method completely solved the zero-weight problem in the general method.b) The interpolation results with the Kriging method in the example area were that the concentrations of CODMn and TP decreased from the shore to the center of the lake in February, May and August and they were decreased firstly and increased from the shore to the center of the lake in November and the concentrations of TN and Chl-a always decreased from the shore to the center in each month. The results show that the concentration distribution of the main eutrophicaiton indicators exist the spatial and temporal characters.c) Through cluster analysis, the routine monitoring points in the example area can be classified to four classifications. The concentration trends of CODMn, TP, TN and Chl-a were consistent with the aforementioned trends of spatial interpolation. In this monitoring points, the eutrophication was controlled by phosphorus concentration; Through factor analysis, the water quality of the four classifications can be explained by three main factors and agricultural non-point source were involved in the three main pollution factors of the four classifications; Analytical results with the the receptor model (Unmix) shows that agricultural non-point source had a greater contribution of nutrients (nitrogen and phosphorus) to the previous three monitoring sites of the four classifications (I classification:TN,TP and NH3-N:81%,85% and 79%,Ⅱclassification:TN, ammonia nitrogen (NH3-N) and phosphate (PO43-):53%, 90%, and 89%,Ⅲclassification:TN and nitrate (NO3-N):34% and 84%).d) The loads of COD,TN and TP from agricultural non-point source which were predicted with the pollutant production coefficient model in Tiaoxi River watershed are 35383.11,11337,705.6 tons in 2015 respectively. Because of the difference of industry and land use, the industry which generates the most load of COD,TN and TP respectively are differce in each county, but the industry which generates the least loads of COD,TN and TP respectively is the rabbit breeding. In all six counties which the Tiaoxi River watershed includs, Changxin county has the maxium loads of COD and TN in unit area, Linan county has the minium load of COD and TN in unit area; the maxium load of TP in unit area appears in Deqing county and the minium load of TP in unit area is in Linan county.e) In order to promote the water quality of the reveiving water body of downstream to the middle nutrition classification, through calculating the threshold concentration of CODMn,TN and TP and conbining water environment capacity of the Tiaoxi River watershed, the abatement load of CODMn,TN and TP from agricultural non-point source is 9150.29,5047.37,116.52 tons in 2015 respectively. Under the target, in the first allocation, Wuxing county is the county in which the abatement rates of CODMn, TN and TP are the maximum (the abatement rate of the three indexes is 57.77%,61.47% and 74.17% respectively), Deqing county is the county which the allocation load of CODMn and TP of is the minimum (13.88% and 5.78% respectively); the abatement rate of TN is in the Changxin county (23.62%). In the second allocation, the main pollutant sources which the COD allocation loads of is the maximum were plantations of tea, mulberry and orchards, rural life, poultry farming and hog raising respectively. The main pollutant sources which the TN allocation load of is the maximum was rural life. The main pollutant sources which the TP allocation load of is the maximum were plantation of dry land, aquaculture, plantation of irrigated land and hog raising respectively.The methodology established in the thesis can accurately identify the eutrophication state, simpley and fastly acquire the main pollution sources and their comtribution to water pollutants, and achieve fair waste load allocation.更多还原