【作者】 高银超；

【导师】 刘秀华；

【作者基本信息】 西南大学 ， 土地资源管理， 2011， 硕士

【摘要】 农业非点源污染成为水质污染的重要原因之一,这一环境问题受到全球范围的关注。非点源污染监测难、定量分析难,基于数学模型的计算机模拟法在解决这些问题上优于其他非点源污染研究方法。AnnAGNPS等大型连续分布式模型成为估算非点源污染负荷、识别污染关键源区、研究农业管理措施的重要手段。三峡库区农业非点源污染日趋严重,而地形复杂、地块破碎、农业经营分散、耕作措施多样,野外监测和人工降雨等研究方法都很难满足该地区农业面源污染研究的需求,数学模型模拟为定量评价流域非点源污染负荷提供了一种有效手段。AnnAGNPS模型在三峡库区的应用较少且多集中在极小流域范围研究,缺乏典型代表性,本文选取距离三峡坝址247km的三峡库区一级支流小江流域作为研究对象,运用AnnAGNPS模型建立数据库：分析流域输入参数的敏感性；模拟小江流域径流量、输沙量、氮、磷污染量；识别面源污染关键源区;模拟三种农业最佳管理措施(BPMs)对流域地表径流量、泥沙量、氮、磷负荷量的影响,为模型在该地区面源污染控制方面的研究提供借鉴意义,主要研究结论如下：(1)对三峡库区小江流域AnnAGNPS模型参数进行分析,掌握构建AnAGNPS模型所需气象、地形、土地利用、土壤、作物、非作物、作物管理、施肥、农药、径流曲线等参数的方法。(2)通过敏感性分析确定影响流域非点源污染负荷输出的8个主要因子,包括SCS曲线值CN、土壤可蚀性因子K、水土保持因子P、作物管理因子C、坡度S、坡长L、曼宁糙率系数MA和化肥施用量PERT。径流量、输沙量、总氮、总磷污染负荷对8个输入参数变化的响应程度不同。(3)运用AnnAGNPS模型对三峡库区小江流域2002-2004年的径流量、输沙量、氮磷负荷量进行模拟估算,并分别采用宝塔窝水文站(控制范围占流域面积的62%)2002-2004年实测数据对径流和泥沙模拟值进行校准和验证,采用小江流域山口2002-2004年实测年均数据对总氮、总磷模拟负荷进行校准和验证。结果表明模型模拟结果相对可靠,适合于三峡库区流域非点源污染负荷评价。(4)根据模型预测结果,模型对径流输出的模拟精度高于对泥沙和养分的输出模拟,2002-2008年小江流域径流量年均模拟值为30.63亿m3,输沙量为277.81万t,总氮负荷为7949.38t,总磷负荷为463.74t。泥沙和养分输出具有相似的规律性,均呈现季节性变化,与降雨量呈现显著的正相关,且受到施肥、地表状况及耕作条件等影响。(5)从空间上看,非点源污染负荷在空间分布不均匀,不同地类污染负荷差别较大。单位面积负荷大小依次是水域、建设用地、耕地、林地、草地,总量负荷大小均依次是耕地、林地、建设用地、水域、草地。(6)从模型模拟的结果可以看出,流域内产生氮磷负荷的主要来源是水土流失和化肥损失等造成的农业非点源污染,其土地利用方式主要是旱坡地,因此,加强旱坡地的农业非点源污染控制对控制流域非点源污染至关重要。通过三种不同农业管理措施模拟研究表明,退耕还林和保护性耕作等措施对流域的污染负荷有明显的削减作用。更多还原

【Abstract】 The agricultural non-point source pollution as the main cause of water pollution which became the focal point of worldwide attention. It is difficult for monitoring and quantitative analysis. The computer simulation based on mathematic models is a good approach to solve these problem. The distributed-parameter, batch process, continuous simulation model such as AnnAGNPS model became the important method to simulate the pollution load,identify the critical pollution sources, study the agricultural management measures. Because of its serious environmental pollution, complex topography, mini land, scattering management, diversified farming measures, it’s better to use mathematic models to simulate pollution loads, not the field monitoring or artificial rainfall in the Three-gorge Reservoir Region.Little research has been used in the AnnAGNPS model,the study amost focus on small-scale. Present research mainly study Xiao Jiang watershed in the Three Gorges reservoir region,it has 250m from the Three Gorges dam. It can give us the implication to control the pollution sources when use the AnnAGNPS model to build databases, analyse sensitiveness, simulate the runoff, silt, nutrient output, identify the critical pollution sources, study the three agriculture management how to influence pollution loads,all of these is This rearch provides reference value for the follow study. Presentresearch’conclusion to be as follows:(1) Present research introduces methods of constructing data set for AnnAGNPS in the case of Xiaojiang River watershed located in the Three Gorges Region. The data set mainly focuses on climate, topography, landuse, soil, crop, non-crop, crop management, fertilizer, pesticide information and runoff curve number data.(2)Main 8 factors determining non-point pollution output were identified through sensitivity analysis, including curve number, K, conservation of water and soil factor, crops management factor, slope, slope length, Manning and the chemical fertilizer. This parameters have different effects to runoff yields, sediment yields, total nitrogen and total phosphorus.(3) AnnAGNPS was applied to estimate runoff, sediment yields, N and P loads from Xiaojiang River Watershed in the Three Gorges reservoir region from 2002 to 2004. Time series data of the same time periods on runoff, sediment yields from Baotawo gauge station (which covers an area of 62% of the whole catchment) and nutrient loads from the outlet of Xiao Jiang catchment were used to calibrate and validate simulation datas. It shows that the results are reliable to use.(4) It shows that the result of runoff simulation is more precise than sediment and nutrition loads simulation. Annual (2002-2008) output of sediment from watershed outlet is 277.81×104 tons, while annual loads of total nitrogen and total phosphorous are 7949.38 tons and 463.74 tons, respectively. All non-point source pollutants are of similar generation trends, which are varying with season, being positive correlative with precipitation, and determined by fertilization, cultivation as well as land surface condition.(5) The results shows that the non-point source pollution loads are different for spatial division and land-use types. The unit area loads in turn are waters, construction land, cultivated land, forest land, grass land, the total quantity loads in turn are farming, forest land, construction land, waters, grass land.(6) The results shows that the non-point source pollution loads are mainly from the soil erosion and the chemical fertilizer loss by dry lands on slopes, so it’s important to control this areas’pollution.Though the study of three agriculture management practices, it indicates that some methods have obviously function to reduce the pollution loads, such as returning farmland to forest and change the slope.更多还原