流域污染负荷模型的构建及其在滦河流域的应用
文献类型：学位论文
| 作者 | 杜新忠 |
| 学位类别 | 博士 |
| 答辩日期 | 2014-10 |
| 授予单位 | 中国科学院研究生院 |
| 授予地点 | 北京 |
| 导师 | 李叙勇 |
| 关键词 | 流域模型 WPLE 污染负荷 非点源污染，watershed model WPLE pollution load non-point source pollution |
| 其他题名 | The construction of watershed pollution load model and its application in Luanhe basin |
| 学位专业 | 生态学 |
| 中文摘要 | 水资源短缺与水环境污染已经严重制约了我国社会经济的可持续发展，水环境污染已成为亟需解决的环境问题。流域污染负荷模型是污染负荷定量化的重要工具，可 以对污染物的产生及迁移过程进行模拟。由于我国特有的区域特色及数据缺乏等原因，国外的复杂模型难以得到广泛的应用，而过于简单的模型无法对污染物的输移 进行机理性模拟。如何平衡流域过程复杂性与模型结构简单性，建立反映我国区域条件和特色，模型结构相对简单，所需输入数据相对较少，同时能对流域污染物输 出过程进行机理性模拟的模型，对流域污染控制与水环境保护有重要的理论意义和应用价值。 以平衡流域复杂性与模型结构简单性之间的矛盾为原则，考虑半干旱半湿润流域水文及污染物输出过程的特征，构建了流域污染负荷模型（Watershed Pollution Load Estimation-WPLE模型）的结构。以柳河流域为研究对象，模拟了流域径流、泥沙及氮磷负荷，并应用流域实测数据进行了模型参数率定与验证。以 柳河流域数据率定的模型参数为输入，模拟预测了滦河上游流域其他7个子流域的非点源污染负荷。以WPLE模型为基础，进行了柳河流域氮磷负荷污染源解析及 其季节与年际差异分析。基于WPLE模型氮磷负荷模拟结果，在子流域尺度识别了柳河流域非点源污染关键源区，并通过BMPs（Best Management Practices）措施效果及费用对比选取了BMPs措施组合方案。论文的主要结论如下： (1) 以平衡流域复杂性与模型结构简单性为原则，考虑半干旱半湿润地区水文及污染物输出过程特征，构建了流域污染负荷模型WPLE的结构，模型可以对流域水文、 泥沙及污染负荷过程进行模拟。WPLE模型结构相对简单，所需输入数据相对较少，同时包括了对流域水文水质主要过程的模拟。此外，在WPLE模型中引入了 人类活动对流域水质水量过程影响的模拟。 (2) 利用2012年汛期的3场洪水事件的水质-水量监测数据，分析了营养盐浓度在洪水事件中的变化，营养盐浓度与流量之间的关系，不同降雨事件特征及水文条件 对营养物浓度变化的影响，并对影响不同形态营养盐输出特征的因素及流失机制进行了探讨。洪水事件下的氮磷浓度变化特征是由降雨量级、强度、水文情势及前期 状况等因素所共同影响的。从营养盐浓度与流量的相关分析及浓度过程的滞后效应两个方面来分析水质与水量关系。浓度与流量即存在正相关关系，又存在负相关关 系，绝大多数表现出正相关关系。不同营养盐形态在洪水事件中表现出不同的滞后效应，没有统一的模式。流域内的磷主要被泥沙所吸附，在洪水事件中主要以颗粒 态流失。与之相反，氮主要通过溶解态的形式流失，淋洗机制是氮的主要流失途径。 (3) 应用WPLE模型对柳河流域的水文过程、泥沙及氮磷污染负荷进行了模拟，并利用流域实测的径流、泥沙及污染负荷数据进行模型的参数率定与验证。总体上，模 型对于月流量序列的模拟效果较好，率定期的R2和NSE均超过了0.85，而验证期的R2和NSE也都超过了0.75。模型对泥沙的模拟中，率定年份的模 拟值与实测值误差较小，相对误差为13.5%，验证年份的模拟误差较大，但模拟值与实测值在同一数量级上。总体上，WPLE模型对氮、磷负荷的模拟效果良 好，月负荷序列率定期的R2和NSE均超过了0.65，而验证期的R2和NSE也都超过了0.6。以柳河流域数据率定的模型参数为输入，预测了滦河上游流 域其他7个子流域的非点源污染负荷，结果表明各子流域的非点源污染负荷输出强度差异明显，土地利用类型组成是导致差异的主要因素。 (4) 基于WPLE模型氮磷负荷模拟结果，分析氮磷负荷的污染源组成及其季节性与年际差异。首先，分析了总氮与总磷年均负荷的污染源组成，非点源为总氮与总磷主 要污染源，分别占总氮及总磷负荷的75%及88.4%。其中，土壤侵蚀负荷、农村地表径流及壤中流为主要的非点源污染。总氮与总磷负荷污染源组成季节性差 异明显，非汛期主要污染源为点源污染，汛期以各种形式的非点源污染为主，主要的非点源污染源在汛期占总负荷的比例要高于非汛期。总氮与总磷负荷污染源比例 年际差异显著，从枯水年到丰水年，点源占总负荷的比例减少，而主要的非点源污染源占总负荷的比例增加。为探究污染源组成的季节及年际差异的原因，分析了流 域水量平衡及其季节与年际变化，结果表明水量平衡变量的季节及年际变化是造成污染源组成年际及季节性差异的主要原因。 (5) 基于WPLE模型氮磷负荷模拟结果，在子流域尺度进行柳河流域非点源关键源区识别。综合考虑总氮与总磷非点源污染，子流域11，14，15及16被识别为 柳河流域的氮磷非点源污染关键源区，4个子流域的面积占流域总面积的19.4%，占总氮非点源污染负荷的30.4%，占总磷非点源污染的29.4%。统计 了关键源区的土地利用组成，结果表明关键源区子流域的农用地及居住用地面积比例更高。将WPLE模型氮磷负荷模拟结果与PRedICT工具相结合以进行非 点源BMPs措施的选择。初选7种BMPs措施应用于关键源区子流域，对比BMPs措施的负荷削减效果和实施费用，选择水土保持耕作及河岸植被缓冲带为柳 河流域非点源污染控制措施组合方案。 |
| 英文摘要 | The water resources shortage and water pollution has estricted the sustainable development of social economy in our country, and water pollution has become an envirometal problem desperately needs to be solved. Watershed pollution load model was the useful tool to quantify the pollution load, and it can simulate the processes of pollution generation and migration. Due to unique regional characteristics and data shortage in our country, the foreign complex model can not be widely applied. However, the simple model can not simulate the mechanism of the pollution transport. Under the principle of balancing the complexity of watershed processes and the simplicity of model structure, establishing a watershed model reflecting the regional characteristics and conditions in our country, with simpler model structure, modest data requirements and at the same time including the simulation of mechanism the watershed pollution export process, has theoretical significance and application value to watershed pollution control and water environment protection. The structure of Watershed PollutionLoad Estimation (WPLE) model was established based on the principle of balancing the complicity of watershed process and the simplicity of model structure and the characteristics of hydrology and pollution export process for semi-arid watershed. The streamflow, sediment and nutrient loads of Liuhe watershed were simulated based on the WPLE model, and the parameters were calibrated and validated using the watershed observed data. The non-point source pollution loads of other sub-watersheds of upper Luanhe basin were predicted using the parameters calibrated by the data of Liuhe watershed. The results of WPLE model were used to investigate nutrient load contributions from different sources and temporal variations of source apportionments in Liuhr. The critical source areas (CSAs) of non-point source pollution of Liuhe watershed were indentified at sub-watershed scale based on the simulation results of nutrient load by WPLE model, the combination scheme of Best Management Practices (BMPs) was selected by evaluating the effectiveness and cost of the BMPs. The main research results included: (1) The WPLE model was established based on the revision of GWLF mode and the principle of balancing the complicity of watershed process and the simplicity of model structure, and the model can simulate the processes of watershed hydrology, sediment and pollution load. WPLE has relatively simple model structure and modest data input requirement, but at the same time, is sufficiently complex to provide descriptions of the main watershed hydrology and water quality processes. In addition, the simulations of the impact of human activities on the processes of watershed water quantity and quality were added in the WPLE model. (2) The water quantity and quality monitoring data were used to investigate nutrient concentration during flood events, the relationship between nutrient concentration and discharge, the impact of storm event characteristics and hydrological conditions on concentration changes, potential factors influencing nutrient export and nutrient loss mechanism during flood events. The results indicated that rainfall magnitude, rainfall intensities, hydrological conditions and antecedent conditions jointly influenced the nutrient concentration change patterns during flood events. The hysteresis effect and correlation between nutrient concentration and flow discharge were used to investigate the nutrient concentration-stream discharge relationships. Correlation coefficients between nutrient concentrations and flow discharge varied from positive to negative for three different events, and most were positive. There were different hysteresis effects for three events, and no consistent pattern was found. Phosphorus is mainly absorbed by sediment in Liuhe watershed, and mainly exported by particulate form during flood events. Conversely, nitrogen is mainly exported by dissolved form, and leaching mechanism is the main nitrogen transport pathway during flood events. (3) The streamflow, sediment and nutrient loads of Liuhe watershed were simulated using the WPLE model, and the model parameters were calibrated and validated based on the watershed observed data. The monthly streamflow simulation of WPLE model was generally good. The R2 (coefficient of determination) and NSE (Nash-Sutcliff coefficient) of the streamflow simulation during calibration period exceeded 0.85, and they exceeded 0.75 during validation period. For the sediment load simulation, the error of the simulated and the measured value for the calibration year is very small, the relatively error is 13.5%. The error of sediment simulation for the validation year is larger, whearas the simulation and observed value were still in the order of magnitude. There was acceptable agreement between the monthly simulated and observed data for TN and TP loads.The R2 and NSE of the nutrient loads simulation during calibration period exceeded 0.65, and they exceeded 0.6 during validation period. The non-point source pollution loads of other sub-watersheds of upper Luanhe basin were predicted using the parameters calibrated by the data of Liuhe watershed. The results indicated that the pollution load intensities for different sub-watersheds showed significant differences, and the land use composition is the main factor. (4) The simulation results of nutrient loads for WPLE model were used to investigate nutrient load contributions from different sources and temporal variations (seasonal and inter-annual) of source apportionments for TN and TP loads. Firsrt, the source apportionments of annual loads were analyzed, the non-point sources were the main source of the total load which accounted for 75% and 88.4% of the TN and TP loads. Soil erosion, surface runoff of rural area and interflow were the main non-point source pollution. There was significant seasonal varations of source apportiments of nutrient loads: the main source of nutrient load was non-point source during flood season, whereas the main source was point source during non-flood season. The main non-point sources, such as soil erosion, rural surface runoff and interflow, had higher contribution ratios to the total load during the flood season than the non-flood season. There was significant inter-annual varations of source apportiments of nutrient loads. From low flow year to high flow year, point source had lower contribution ratio to the total load, whereas the main non-point sources had higher contribution ratios to the total load. In order to investigate the reseason for the varations of source apportiments of nutrient loads among different seasons and hydrological years, the watershed water balance components and their seasonal and inter-annual varations were analyzed based on the WPLE hydrology simulation results. The results indicated that the varations of source apportiments of nutrient loads among different seasons and hydrological years were mainly caused by the seasonal and inter-annual varations of watershed water balance components. (5) The non-point sources CSAs of Liuhe watershed were identified at the sub-watershed scale based on the nutrient loads simulation results of WPLE model. Considering both TN and TP non-point source pollution, sub-watershed 11, 14, 15 and 16 were identified the CSAs. The four CSAs sub-watersheds accouted for 19.4% of the whole watershed area, 30.4% of the TN non-point source pollution load and 29.4% of the TP non-point source pollution load. The analysis of land use composition of CSAs sub-watersheds indicated that CSAs had more area proportions of agricultural and residential land than other sub-watersheds. The combination of WPLE nutrient load simulation results and the PRedICT tool was used to select the non-point source pollution BMPs. The primarily selected 7 BMPs were applied in the CSAs sub-watershed, and the load reduction efficiencies and costs were compared to choose the appropriate BMPs. Finally, the combination of conservation tillage and riparian vegetation buffer was selected as the combining scheme of non-point source pollution control measures of Liuhe watershed. |
| 源URL | [http://ir.rcees.ac.cn/handle/311016/34491]  |
| 专题 | 生态环境研究中心_城市与区域生态国家重点实验室 |
| 推荐引用方式 GB/T 7714 | 杜新忠. 流域污染负荷模型的构建及其在滦河流域的应用[D]. 北京. 中国科学院研究生院. 2014. |
入库方式： OAI收割
来源：生态环境研究中心
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。