磷是引起水体富营养化的重要限制因子，当前总磷（total phosphorus，TP）已经成为 长江流域的首要水污染因子。因此，针对长江流域磷污染时空分布不清、负荷来源不明等 问题，选择岷江流域开展磷污染负荷特征及其驱动机制研究，具有重要的理论意义和应用 价值。本文探索建立了“水体磷污染时空分析—陆源正向计算—断面通量反向检验—重点 驱动专项识别—单元负荷定量限制”的流域磷污染防治研究框架，优化了磷污染负荷计算 和来源识别方法，建立了人口城镇化对磷的污染负荷驱动模型，并在岷江流域进行案例研 究，分控制单元提出了对策措施。主要研究成果如下： （1）结合历史资料和现场监测结果分析岷江水体中磷浓度特征。研究结果表明：岷 江流域 TP 超地表水Ⅲ类水质标准的断面主要集中在岷江干流中下游和局部支流，中游黄 龙溪断面和下游月波断面是两个磷污染较突出的区域；岷江水体中磷形态以溶解性磷（total dissolved phosphorus，TDP）为主，干、支流 TDP 浓度分别占 TP 浓度的 65%和 82%；其 中 TDP 又以活性磷酸盐（phosphates，PO43-）为主，干、支流 PO43-浓度分别占 TDP 浓度 的 72%和 84%。 （2）优化了流域磷污染负荷计算方法，识别了磷污染来源，并实现陆源负荷和断面 通量的相互验证。重点优化了经典的径流输出模型，在分离肥料流失部分单独计算基础上， 增加了地形、降水和土壤背景修正系数，改进后的径流模型比经典模型精度提高了 13%。 研究结果表明：2015 年，工业生产、居民生活、农业种植、畜禽养殖和地表径流对岷江流 域 TP 负荷贡献率分别为 60.2%、12.8%、13.1%、11.3%和 2.6%；工业生产是岷江流域 2005 年以来 TP 负荷增量的主要来源，占总增加量的 99.8%，其次是居民生活，占总增加量的 5.8%；从空间分布来看，岷江中游黄龙溪单元磷污染主要受城镇生活影响，下游月波单元 磷污染主要受磷化工影响。 （3）基于投入产出理论和污染排放系数理论建立了耦合居民生活和消费对磷污染的 负荷驱动模型。研究结果表明：岷江流域 2015 年人口城镇化驱动下 TP 污染负荷较 2005 年增加了 1552.8t/a。其中，居民生活直接排放增加了 755.1t/a，占总增加量的 48.6%；居民 消费间接带动的负荷增加了 797.8t/a，占总增加量的 51.4%；2005—2015 年人口城镇化驱 动导致的岷江磷污染负荷改变量绝对值城镇化效应（811.7t/a）>消费结构效应（797.8t/a）> 技术效应（-500t/a）>规模效应（443.2t/a）；人口城镇化驱动下的增量集中在成都区域，达 1298t/a，占全流域总增量的 84%。 （4）针对传统上水环境容量计算结果偏小的问题，本研究优化了水体 TP 环境容量计 算方法，结合发展情景提出了污染防治对策措施。研究结果表明：为了达到预期水质目标， 2020年和2030年岷江流域分别需要在当年预测情景下削减TP负荷13720.8t/a和13809.3t/a； 其中，2020 年黄龙溪单元和月波单元分别需要削减 TP 负荷 256.8t/a 和 13141.8t/a；2030年，黄龙溪单元和月波单元分别需要削减 TP 负荷 555.6t/a 和 13066.6t/a；在今后一段时期， 控制点源污染仍然将是岷江流域磷污染防治的主要任务；黄龙溪单元磷污染控制需要从控 源减排和调水增容两个角度考量对策措施；月波单元需要削减区域内磷化工污染排放量； 岳店子单元需要从点、面源综合控制方面考虑以实现水质达标。

Phosphorus was an important limiting factor for the eutrophication of the water body. At present, total phosphorus（TP）has become the main pollution factor in the Yangtze River. In order to find out the spatial and temporal distribution of phosphorus pollution and pollution sources in the Yangtze River Basin, the study of the characteristics of phosphorus pollutant load and its driving mechanism in the Minjiang River Basin was conducted. In this paper, a research system consist of phosphorus concentration characteristics in water - land source forward cccounting - reverse inspection of section fluxes - driving specific identification - unit load quantitative limit accounting was established. This study optimized the method of TP load calculation and source identification, a population-based TP load driving model was set up and applied in the Minjiang River Basin. The main achievements were as follow: （1） Characteristics of phosphorus pollution in the Minjiang River were analyzed based on long-term historical sequences and on-site monitoring data. The results showed that the phosphorus-contaminated water concentrated in the middle and lower reaches of the main stream of the Minjiang River and the minor tributaries. Huanglongxi and Yuebo sections had much higher TP concertration as compare to others. Phosphorus was the main composition of dissolved phosphorus (TDP) in the Minjiang River. TDP concentrations occupied 65% and 82% of the TP concentrations in the main stream and tributaries respectively. The majority of TDP was active phosphate (PO43-) , which accounted for 72% and 84% of the TDP concentration in the main stream and tributaries respectively. （2） The method of TP load calculation was optimized and the mutual verification of land-source load and cross-section flux was demonstrated. This study focused on optimizing and improving the classic TP output model. Firstly, the fertilizer loss was calculated separately. Secondly, the terrain, precipitation, and soil background correction coefficients were added. The improved runoff model was 13% more accurate than that of the classic model. The results showed that the contribution rates of industrial production, resident life, agricultural planting, livestock breeding and surface runoff to TP load in the Minjiang River Basin were 60.2%, 12.8%, 13.1%, 11.3% and 2.6%, respectively, in 2015. Industrial production has been the main source of TP load increment since 2005 in Minjiang River Basin, which accounting for 99.8% of the total increase, followed by residents living, accounting for 5.8% of the total increase. In terms of the spatial distribution, phosphorus pollution in Huanglongxi was driven by residents' living discharge, and the phosphorus pollution in the Yuebo section affected by industrial production. （3） Based on input output theory and pollution emission coefficient theory, a load driven model for phosphorus pollution coupled with residents' life and consumption is established. The results indicated that the TP load driven by urbanization was 4393.6 t/a in 2015. It has increased by 1552.8 t/a compared with 2005. The direct discharge of residents' livelihood increased by 755.1 tons/year, the indirect driving load of residents' consumption increased by 797.8 t/a. The absolute amount of urbanization effect (811.7 t/a)>consumption structure effect (797.8 t/a)>technical effect (-500 t/a) > the population quantitative effect (443.2 t/a). The increment of population driven by urbanization was mainly in Chengdu region. The number was 1298t/a, which occupied 84% of the total delta of the whole river basin. (4) This study has optimized the calculation method of TP water environmental capacity, and combined with the development scenario and pollution control proposal. The results showed that the load of TP in the Minjiang River should be reduced by 13720.8 t/a and 13809.3 t/a in 2020 and 2030 respectively. The loads of TP in Huanglongxi unit and Yuebo unit should be reduced by 256.8 t/a and 13141.8 t/a in 2020, 555.6 t/a and 13066.6 t/a in 2030, respectively. Therefore, the point source pollution reduction would still be the main task for controlling of the total phosphorus pollution in the Minjiang River Basin in the future. The control of phosphorus pollution in Huanglongxi unit required both source emission reduction and water adjustment. For Yuebo unit, phosphorus pollution reduction was the significant measurement. As for Yuedianzi unit, integrated control of point and non-point sources was the main step to achieve water quality objective.