Study On The Simulation Of The Non-Point Source Pollution In The Upstream Watershed Of The Panjiakou Reservoir

Non-point source pollution is defined to be the runoff transport of constituents from diffuse sources on the land to streams (Lane 1983, Browne 1990, Huber 1993). Some of the constituents composing non-point source pollution include oxygen-demanding substances, suspended solids, nutrients and toxic matter which would deteriorate the water quality and cause the eutrophication in the water body. In the northern area of our country, the water pollution is serious. At the same time the quantity of water resource per person is so little. So it is a very important task to protect the drinking water resource. This project is a part of the Plan of Studying the Prevention and Management of the Eutrophication in the Important Water Source in Haihe River Basin. In the project SWAT model is used to simulate the production and transportation of the non-point source pollutants in the upstream watershed of the Panjiakou Reservoir. The loadings of non-point source pollutants are calculated for different hydrologic years and the spatial and temporal characteristics of non-point source pollution are studied. At first the basic data of the study area are collected and processed for the model. DEM data of the upstream watershed are used to define the stream network which is used to define the layout and number of subbasins. Topographic characteristics of the stream network and subbasins are all derived from the DEM. The soil and land use data are reclassified for the model and the soil physical properties database file is created. SWAT model is calibrated with observed data for the study area and the important parameters' sensitivity is analyzed. The calibrated model is used to calculate the loadings of non-point source pollutants for a wet year, a normal year and a dry year respectively. The difference of different hydrologic years' result is dramatic. The loading of non-point source pollution in the wet year is relatively larger and in the dry year smaller since the non-point source pollutants are mainly transported in the runoff and the most loading in a year is produced in the flood season. Because SWAT is a distributed model, it is possible to view model output as it varies across the basin and the critical areas and reaches are found in the study area. Each land cover represented in the model yields different results. Corn is responsible for 51% of the nitrate and 95% of the soluble phosphorus though its area is only 22% of the study area. RNGB( the mountain range brush) has a higher sediment yield on a per unit area basis because its location's slope is relatively higher and the percentage of coverage is comparatively smaller. Each parameter's output of the forest area is smaller than the basin average on a per unit area basis. The fertilization in rainy season has an important impact on the non-point source pollution in the study area and different fertilization scenarios are simulated. At last the limitation of the model is discussed and the measures for the prevention of non-point source pollution are presented according to the model result and the experience of other areas.