农业小流域面源氮负荷与源汇变异的定量化及其不确定性

Nonpoint source pollution (NSP) from agriculture is one of the most broad and disguised pollution among NSP. Nevertheless, it is difficult to quantify nonpoint N loads and variations of sources and sinks, which greatly hinder the realization of the goal of allowable total discharge amount of NSP. The possible reason is the poor modeling of N discharge and removal according to references. Therefore, the main purpose of this study is to quantify nonpoint N loads and variations of sources and sinks by improving the modeling precision of N discharge and removal. This study will apply the new toolbox of geographical information system (GIS) to identify the pathway of N transportation and summarize distance of transportation, and use membrane inlet mass spectrometer (MIMS) method to accurately quantify N removals and its related parameter in cannels, ponds, buffers, and river along N pathway. Applying the new method of hierarchical Bayesian published recently by us, we will optimize N discharge and its related parameters. With the improvement of the modeling of N discharge and removal, we will then construct an NPS model which links N discharge, removal, and loading. Combined with Monte Carlo method, we will finally quantify N loads to rivers and variations of sources and sinks including uncertainties using the improved model, and provide scientific decisions for NSP controls in watershed. This study will include N pathway in transportation and its removal into simulation for the first time, construct an unbiased NPS model links N discharge, removal, and loading, and potentially contribute to more precise apportionment of the NPS in lack-data region.

农业面源污染是面源污染中覆盖面最广、隐蔽性最强的污染类别。然而，由于种类多、过程复杂，目前还很难量化面源氮负荷和源汇变异，极大阻碍了农业面源氮总量控制目标的实现。文献调研表明，发生量和消纳量模拟的不准确，是面源氮负荷和源汇变异量化过程中最主要的不确定性来源。本研究拟应用新一代地理信息系统工具，识别各面源氮迁移路径和距离，结合先进的膜进样质谱分析方法，准确获取面源氮沿途（沟渠-池塘-缓冲带-河流等）消纳系数与消纳量；应用申请人最新提出的层次贝叶斯方法，优化发生量和发生系数；最终构建面源氮“发生量-消纳量-水质负荷量”的“三量”模型，应用蒙特卡罗方法反演农业流域面源氮负荷和源汇的时空变异，并量化其不确定性，为农业面源污染控制提供科学决策。这一研究将1）首次在模型中考虑了氮素迁移路径及其消纳特点，构建无偏的面源氮“三量”模型；2）有望为缺资料地区提出更准确的面源污染源汇解析方案。

农业面源污染是面源污染中覆盖面最广、隐蔽性最强的污染类别。然而，由于种类多、过程复杂，目前还很难量化面源氮负荷和源汇变异，极大阻碍了农业面源氮总量控制目标的实现。在稻作流域，多小型水体（沟、塘、河等）的存在加剧了面源污染过程模拟的困难性和不确定性。迁移路径识别和沿程消纳量是面源氮负荷和源汇变异量化过程中最主要的不确定性来源。本项目首次解析了面源污染沿程路径几何特征（水体长度、面积）与位置的重要性。研究发现，通过位置和面积的三个统计指标（湿地离河岸的均值，位置的均匀度和平均面积）可以解释流域面源污染负荷均值的70%以上，而四个统计指标（位置和距离的均匀度、变异系数）可以解释流域面源负荷标准差的65%以上。基于此，开发新一代地理信息系统工具，识别并量化流域内每个栅格污染物沿程沟渠、池塘、河流、湖泊等的距离或面积；结合先进的膜进样质谱分析方法，准确获取面源氮沿途（沟渠-池塘-河流等）消纳系数与消纳量，发现不同类型的湿地反硝化速率均表现出春夏季高、秋冬季低的特点。沿氮素迁移路径，水体反硝化速率先升高后降低，源头沟渠的消纳速率明显高于末端河流的消纳速率，其主要影响因素为硝氮浓度和水体温度。基于上述面源氮路径的识别与消纳系数，开发了参数少、操作简单、结果可视化、模拟结果可靠的全新流域面源污染模拟系统。该系统以污染物路径识别为核心、工具条式功能开发为特色，污染物流失量-消纳量-负荷量分布式模拟为根本，为流域面源污染的智慧管理如生态湿地选址、农场选址、污染物路径追踪、减排策略分析、风险评估、水质目标实现等提供新的途径。