基于AnnAGPS模型的三峡库区典型小流域农业非点源污染模拟评价研究
文献类型：学位论文
| 作者 | 孙正宝 |
| 学位类别 | 硕士 |
| 答辩日期 | 2012-05 |
| 授予单位 | 中国科学院研究生院 |
| 授予地点 | 北京 |
| 导师 | 陈治谏 |
| 关键词 | 三峡库区 陈家沟小流域 农业非点源污染 AnnAGNPS模型 |
| 其他题名 | Simulting and Evaluting Agricultural Non-Point Source (AGNPS) Pollution Loads from a Typical Agricultural Small Watershed of Three Gorges Reservoir Area Based on AnnAGNPS |
| 学位专业 | 自然地理 |
| 中文摘要 | 三峡库区水土流失严重，生态环境脆弱，农业非点源污染形势严峻。有效的控制农业非点源污染已成为从根本上解决水质污染问题的关键。目前，定量化研究农业非点源污染最直接、最有效的方法是建立和运用分布式水文模型进行模拟研究。 本研究以三峡库区中部的重庆市万州区长岭镇陈家沟小流域为研究区，其地理范围介于30°43′30″-30°46′20″N，108°29′40″-108°31′40″E之间。流域面积8.32 km2，流域内最大相对高差580 m，流域出水口距汇入长江的入水口 15.45 km。该小流域属于典型的农业生产区，是在三峡库区开展农业非点源污染研究的理想区域之一。经过对目前主要 农业非点源污染模拟模型的对比，结合研究区的实际情况，本研究选用能够反映复杂地形地貌，对水文、氮磷迁移以及土壤侵蚀具有较强模拟能力的AnnAGNPS（Annualized AGricultural Non-Point Source）模型对陈家沟小流域的农业非点源污染进行模拟研究。 根据 AnnAGNPS模型原理和技术规范，本文以中国科学院水利部成都山地灾害与环境研究所万州生态站的野外观测、实验数据和调查资料为基础，在ArcGIS9.3、ArcView3.3以及ERDAS9.2等地理信息系统软件的辅助下，建立了陈家沟小流域农业非点源污染模拟研究数据库。对AnnAGNPS模型进行了参数敏感性分析、参数率定（calibration）与验证（validation），评价了该模型在陈家沟小流域的适用性。在此基础上，本文依次开展了两项相关研究：一是回顾评价了研究区2001——2010年的农业非点源污染状况；二是对研究区2011——2020年的农业非点源污染发展趋势做了2种不同情景的模拟研究和对比分析。研究结果表明： （1）在土壤可蚀性因子（K）、SCS曲线值（CN）、曼宁糙率系数（MA）、作物管理因子（C）、水土保持因子（P）、坡度（S）、坡长（L）和化肥施用量（FERT）等8个因子中，总氮模拟输出对化肥施用量（FERT）的变化响应最灵敏；总磷模拟输出对K值的变化响应最灵敏。按照它们的SI值由大到小排列，各参数针对TN和TP的敏感性排序分别为FERT > P > K > C > S > MA > CN > L和K > FERT > CN > C > P > MA >S >L。 （2）回顾评价发现：2001——2010年间，陈家沟小流域农业非点源氮磷负荷总量在时间上呈现出不规律动态变化的特点，年均氮磷输出量分别为8.55吨和3.02吨；在空间上，呈现出向流域中部和流域出水口聚集的趋势。此外，不同土地利用类型单位面积的氮磷负荷差异显著，但以旱地和水田的氮磷输出占主导地位，居民点及工矿用地、经果林等次之。 （3）情景模拟预测发现：情景2（模拟情景）模拟条件下，陈家沟小流域未来的农业非点源氮磷负荷总量呈现出不断消减的趋势；即按照情景2的退耕还林方案对小流域内坡度≥25°的耕地（包括旱地和水田）实施逐步退耕还林能够有效的消减研究区34.39%的总氮流失和25.50%的总磷流失。按照该方案，到2020年陈家沟小流域非点源氮磷负荷将分别下降到5吨以内和2吨以内。情景分析结果同时表明，研究区气象因素是小流域农业非点源氮磷负荷波动的主导因素，但是这种波动仅是围绕着某一均值做小幅的摆动，而非根本性的改变；小流域土地利用类型（尤其是耕地和林地）的改变是研究区域内农业非点源污染负荷发生根本性变化的主要原因。另外，无论是在对照情景下还是模拟情境下耕地（包括旱地和水田）都是小流域农业非点源污染负荷的主要来源。在空间分布上，陈家沟小流域的农业非点源污染负荷延续了过去10年（2001——2010年）的发展趋势，向流域中部和流域出口附近聚集。值得注意的是，随着退耕还林方案（情景2）的不断推进，流域边界附近区域（海拔相对较高、坡度较大）的非点源氮磷输出得到有效控制。 （4）本研究根据降雨及其引发土壤侵蚀是一个连续变化过程的特点，基于次降雨雨量（P）和最大30 min雨强（I30）提出了一种基于模糊逻辑下精细的表达和识别侵蚀性降雨的模型——模糊隶属度模型（Fuzzy Membership Model，FMM）。FMM采用隶属度向量描述侵蚀性降雨，与常规的基于精确逻辑下的表达有着本质的区别。在FMM中，次降雨是否属于侵蚀性降雨由其隶属度向量集合表达，隶属度（Fuzzy Membership Value）的取值范围介于0到1之间，刻画了降雨事件及其引发土壤侵蚀连续变化的客观事实。值得注意的是，FMM并不是直接给出侵蚀性降雨标准，而是提供了一种用于识别侵蚀性降雨的方法和实现降雨侵蚀精细化表达的基本途径，以作为对模型运用的修正。 （5）通过AnnAGNPS模型在陈家沟小流域的运用，本研究发现该模型在模拟运算过程中对研究区的土地利用类型数据进行了不同程度的概化，这种概化导致了未调整地类在相同条件下的输出不同。因此本文建议在后续的开发研究中以研究区土地利用类型数据为依据，增加次级分室单元（Sub-Cell）的划分，以改进模型在丘陵区和山区的适用性，改善模拟效果。 研究结论不仅为陈家沟小流域深化和完善现有的生态恢复措施和政策、开展新的生态恢复工程提供决策依据，同时也可为三峡库区类似小流域的规划与优化管理参考。 |
| 英文摘要 | Agricultural non- point source pollution (AGNPS) is a significant problem in Three Gorges Reservoir Area which is faced critical soil erosion problem and fragile ecology environment. It is the key to fundamental solve the problems of water pollution that effective control of agricultural non-point source pollution. At present, it is the most effective and the most useful approach of quantitative study on agricultural non-point source pollution that is estimating and simulating by distributed hydrological model. In this paper, Chen Jiagou small watershed in Three Gorges Reservoir Area is selected for study area, which located in 30°43′30″- 30°46′20″N，108°29′40″- 108°31′40″E(8.32 km2). It a typical agricultural catchment and that is the best choice to research for agricultural non- point source pollution in this area. AnnAGNPS (Annualized AGricultural Non-Point Source) model, an effective and useful model is used to simulate and evaluate agricultural non-point source pollution loads associated with Arc /info 9.3 and ERDAS 9.2 in Chen Jiagou small watershed. The paper first introduces the theories and structure of AnnAGNPS model. This is followed by a brief overview of current research efforts to explore and establish non- point source pollution models and distributed hydrological models. In the section 2, the detail information about study area is then presented, which is followed by a preliminary study about the law of agricultural non- point source pollution loss in Chen Jiagou small watershed to further illustrate the study area. In the section 3, the data set of Chen Jiagou small catchment is prepared which was conducted by the principles and technical specifications of AnnAGNPS, and then, the sensitivity analysis and estimation are carried out. The reviewed evaluation and scenarios prediction are presented in section 4 and section 5 respectively. Section 6 provides a summary of this study. Some meaningful and useful results are obtained by estimating and simulating in this paper. First of all, results of sensitivity analysis of parameters indicated that the eight parameters which have been selected have different effects on total nitrogen and total phosphorus. The amount of chemical fertilizer (FERT) and soil erodibility factor (K) has important effect total nitrogen (TN) and total phosphorus (TP) respectively. Secondly, reviewed evaluating showed that the Chen Jiagou small watershed agricultural non- point source nitrogen and phosphorus loads showing irregular dynamic changes in 2001-2010, the average annual output of nitrogen and phosphorus were 8.55 tons and 3.02 tons; in spatial, TN and loads showed a trend that gathered to the basin in the central and watershed outlet. In addition, the different land use has differences of nitrogen and phosphorus loads in per unit area significantly, but dry land and paddy fields played a major role in 2001- 2010. Thirdly, scenarios simulation and prediction indicated that the agricultural non-point source nitrogen and phosphorus loads showing abatement on different levels in 2011- 2020. In the scene 1（background scene）, the TN and TP will be reduced 15.32% and 4.64% respectively, and in the scene 2(assumed scene), the TN and TP will be reduced 34.39% and 25.50% respectively in 2020. It further illustrate that returning land for farming to forestry is a useful and effective approach for ecological restoration in Chen Jiagou small watershed. And in the scene 2, the TN and TP loads will be controlled less than 5 tons and 2 tons respectively in 2020. Fourthly, a fuzzy membership model (FMM) was developed to recognize the erosive rainfall based on the fact that rainfall and soil erosion often varies gradually. FMM describes rainfall by membership values of two physical parameters, precipitation (P) and maximum 30 min intensity (I30). Firstly, calculate the membership values of P and I30, and construct the fuzzy membership functions. Secondly, the membership values of P and I30 were combined to a new membership value of individual rainfall. Finally, the erosive rainfall was identified by the new membership values.FMM has the characteristics of simple structure and meaningful physical parameters, and all parameters can be derived from hydrological and meteorological observations. The model calibration and test were performed in Chenjiagou small watershed, and the mistaking degrees are under 20%. Finally, extracting Sub-cells based on Cells is suggested as an approach to improve the applicability of model in hilly region and mountains. These conclusions are not only helpful for carrying out the ecological restoration projects in Chen Jiagou small watershed, but also useful for the similar basins planning and optimization of management in the Three Gorges Reservoir Area. |
| 学科主题 | 自然地理学 ; 土壤学 |
| 语种 | 中文 |
| 公开日期 | 2012-07-26 |
| 源URL | [http://192.168.143.20:8080/handle/131551/4164]  |
| 专题 | 成都山地灾害与环境研究所_山区发展研究中心 |
| 推荐引用方式 GB/T 7714 | 孙正宝. 基于AnnAGPS模型的三峡库区典型小流域农业非点源污染模拟评价研究[D]. 北京. 中国科学院研究生院. 2012. |
入库方式： OAI收割
来源：成都山地灾害与环境研究所
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。