基于不确定性的流域最大日负荷总量分配及控制措施研究

Water pollution is an important factor in restricting the sustainability of socio-economic and environment in China now. Environment system in river basins is a complicated system with quite a lot of uncertainties. This kind of uncertainty is an important factor to affect the ability to satisfy water environment target. If this uncertainty was ignored, it may result in the fact that water environment management strategy wouldn’t meet the anticipated objective to protect water body. Watershed Total Maximum Daily Loads (TMDL) program is one of the main methods to control water pollution in U.S.A. It pays attention to the uncertainty which is integrated into TMDL as Margin of Safety (MOS). It reduces the risk of exceeding water quality standard due to uncertainty. However, the scientific estimation method for MOS has not been developed due to the short studies on quantitative estimation of MOS. Therefore, the MOS estimation needs to be investigated systematically.. The Xiangxi River Watershed lies in the northeast of the Three Gorges Reservoir Area and its upper reach was selected as the study area in this project. Based on the TMDL of US, the sketch of TMDL of Xiangxi River Watershed will be established. Next, the uncertainty of model parameters, input data and model structure will be analyzed to calculate the MOS based on the SWAT model. Then, propose some rational management measures of controlling the local agricultural non-point source pollution to help local farmer and government make more effective and efficient agriculture management measures, such as accurate fertilizer. Finally, the control scheme of Total Nitrogen (TN) and Total Phosphorus (TP) will be discussed and the impacts of management measures on agriculture economy and pollution abatement were studied. This study has important practical significance in controlling agricultural non-point source pollution and the protection of social, environmental and economic sustainable development in Xiangxi River Watershed.

目前，在我国社会经济迅速发展的同时，水环境问题逐渐成为制约社会经济和环境可持续发展的重要因素。流域水环境系统是一个充满不确定性的复杂系统，这种不确定性是影响水环境目标实现的重要因素。如果忽略不确定性的存在，那么流域水环境管理策略无法达到预定的水体保护目标。本项目选取香溪河流域为研究区域，借鉴美国最大日负荷总量(TMDL)方法，初步构建符合研究区实际的TMDL框架；结合SWAT模型探讨TMDL中污染负荷模拟中的不确定性，科学计算安全余量(MOS)，将污染负荷合理分配；考虑农民的初期损失和后期收益，综合考虑非点源污染控制中的环境效益和经济效益，优化选择非点源污染控制措施；最终形成基于降低风险的流域污染控制方案。本研究对于有效控制农业非点源污染，保障区域社会、环境和经济的可持续发展具有重要的实际意义。

本研究以香溪河流域为研究区域，在SWAT模型基础上，对非点源污染的模拟和控制进行了研究。基于多种因素识别了香溪河流域面源污染关键源区，探讨了DEM不确定性对关键源区识别的影响，研究结果表明，DEM不同分辨率和来源对关键源区识别影响较大，重采样方法对其影响可以忽略，TN和TP关键源区的识别对于DEM 数据敏感性不同，相比较而言，TN对DEM不同分辨率更敏感，而TP则对DEM不同来源和重采样方法更敏感。同时，基于新建肥料数据库模拟了畜禽养殖对非点源污染的影响，结果表明使用新建数据库与模型自带数据库模拟得到的总磷负荷结果差距很大，其中，有机磷受影响最大，空间上，流域下游比上游受粪肥施加的影响要大。不同粪肥施加导致的污染物质的空间分布差异性并不大。此外，探讨了土地利用动态变化对非点源污染模拟的影响，结果表明，采用动态的土地利用输入能够显著地提高模型的模拟效果，尤其对于氮磷等污染物的模拟，而在4中静态土地利用输入条件中，采用中间年份的土地利用输入可以取得相对较好的模拟结果。在污染控制中，结合SWAT模型与差分进化算法对香溪河流域的动态水环境容量进行了计算和分析，结果表明，丰水期氮磷的水环境容量明显低于枯水期的容量，这可能是由于丰水期的流量大，虽然在一定程度上增加了水体对污染物的容纳能力，但是更增加了氮磷通过地表径流进入河道的量，因此，导致了河道对氮磷容纳能力的下降。