三峡库区非点源污染负荷时空分布模型的构建及应用

为研究三峡库区非点源污染负荷的分布特性,定义降雨侵蚀力影响系数,并将其引入John提出的年均输出系数模型,形成改进的年输出系数模型;针对不同的土地利用类型,基于表征水力联通性的地形指数和植被覆盖度,提出了入河系数的空间分布式;将年输出系数模型与入河系数结合,构建非点源污染负荷的时空分布模型,并对模型进行了验证。应用所构建的模型,对三峡库区2002-2012年总磷总氮负荷进行了模拟。结果表明,库区总磷总氮负荷的时空变化明显,且降雨侵蚀力是影响其年际变化的主要因素;年均总氮负荷在2.6~4.2 kg/hm~2,总磷负荷在0.432~3.186 kg/hm~2;在土地利用、农村人口和畜禽养殖产生的负荷中,畜禽养殖对氮的贡献最大、约占总量的45%,而土地利用对磷的贡献最大、约占总量的57%。该研究成果可为三峡库区非点源污染的控制与治理提供参考,该文所构建的模型也可用于其他大中型区域非点源污染的模拟。

Development and application of non-point source pollution load model of spatial and temporal distribution in Three Gorges Reservoir Region

Abstract: Non-point source (NPS) pollution has been increasingly recognized as a major contributor to the declining quality of aquatic environment in recent years. Three Gorges Reservoir Region (TGRR) covers an area of approximately 58 000 km2, its landscape is dominated by 74% mountains and 21% low hills, and comprises forests, agricultural areas, streams and stretches of steep terrain. In the TGRR, the massive use of fertilizers has caused the releases of large quantities of nitrogen and phosphorus, consequently leading to greater eutrophication. Therefore, there is a dire need for precise knowledge of the loads and causal factors of nitrogen and phosphorus, which remains a high priority for the TGRR and is also essential for the development of the entirety of China as well. NPS pollution is related to uncontrollable climatic events as well as site-specific conditions such as soil, topography, and land use. NPS pollution is difficult to be monitored in entire region, for it is generated over an extensive area and the pollutants enter receiving water in a diffused manner. As a result, many models have been developed for the NPS pollution load estimation, since the quantification is critical to guide decision makers before plans are implemented. Some mechanism models generally require substantial parameterization and calibration data that are often difficult to obtain, especially in large-scale regions and in developing countries. The export coefficient model has the advantages of requiring less data and having fewer parameters, and has been recognized to be applicable and acceptable for modeling the NPS pollution. However, the export coefficient model also has limitations. For example, it does not take into account the uneven spatial and temporal distribution of rainfall, as well as the spatial variation of the process that the pollutants on-side transfer to the water. In this paper, in order to study the spatial and temporal distribution of NPS pollution load in TGRR, a modified annual export coefficient model was proposed by introducing the rainfall erosivity factor into the annual average export coefficient model from John. Considering that there were losses when the pollutants on-side transfer to the water, based on terrain index and vegetation coverage degree, the spatially distributed loss coefficient for different land use was presented. Then, the NPS pollution load model of spatial and temporal distribution, which combined loss coefficient with the annual export coefficient model, was developed, and the model was also validated. Using the load model, the temporal and spatial distributions of the total nitrogen (TN) load and total phosphorus (TP) load in the TGRR were simulated from 2002 to 2012. The results showed that both the spatial and temporal variations of loads were obvious, and their inter-annual variations were affected observably by the rainfall erosivity factor. The annual average TN load was 2.6-4.2 kg/hm2, and the TP load was 0.432-3.186 kg/hm2. Among the loads from land use, rural population and livestock, the largest contribution of TN load was from livestock, about 45%, while the largest one of TP load was from land use, about 57%. The developed model in this paper can be used for estimating NPS load of other large-scale regions.