Simulation Of Non-point Source Pollution In The Lianshui Upper Basin Based On SWAT Model

As the point source pollution has been controlled effectively, non-point sourcepollution (NPS) became the main source of water pollution.It is necessary to simulatethe NPS process and analyse its temporal and spatial distribution charateristics andcontrolling factors. By the combination of NPS model-Soil and Water AssessmentTool (SWAT) with Geography Information System(GIS), the results from this studycan provide useful instruction for NPS control and watershed planning.This paper reviewed the development of NPS models first. Based on GIStechnology, SWAT model endorsed by USDA ARS (Agricultural Research service),has been applied to simulate the NPS load in the Lianshui Basin (the upper reaches ofLoudi City). The main study contents and new findings are as follows:(1)The watershed NPS database for SWAT model in the study area has beenestablished using GIS, MATLAB and other technology. GIS has been used forprojection transformation, property consolidation and the basin boundary extraction.In order to develop the weather generator in SWAT model, PCPSTAT software wasadopted for the meteorological data statistics. The interpolation method was used tomake the soil grain size conversion, and the process is realized through MATLABprogram. The soil water characteristics were estimated by the SPAW program to buildup soil property database. Using statistic method and field investigation, the otherattribute data were obtained, such as point source pollution data, etc.(2) The measured data were used to calibrate and verify the SWAT model. Theresults from simulation of monthly average natural runoff, sediment and water qualityindicate that the modified SWAT model is suitable for the Lianshui Basin. The dataseries (1999～2002) of average monthly runoff, sediment, and water quality fromLoudi gauging station are applied for parameter calibration of the model, while thedata series (2003～2005) of average monthly runoff, sediment, and water quality wasused to verify the model. The results show that the application of SWAT model toLianshui Basin of nitrogen and phosphorus to study NPS is feasible.(3) The established SWAT model was used to calculate NPS in different years.With the comparison of different years, the nutrition contributing ratio in the studyarea was estimated. The analysis shows that NPS is the main contributor for water environment pollution in the study area. The spatio-temporal features of NPS in thebasin were also analysed. NPS loading is contributed mostly during the rainy season(April～June), while during dry season (July～September) it is less. Spatially, NPSsources are distributed in both banks of Lianshui Main River. The impacts of land-usechanges on NPS are evaluated. The AGRL (farmland) contributes the largest part ofNPS loading whereas the NPS loading from FRST (woodland) is least. Comparatively,the land use in 2005 was more beneficial to reduce the NPS pollution load into theLianshui Basin than that in 1980. Three cenarios of fertilizer adding level wereproposed to find the best management practices. The simulation results show thatadding low fertilizer rates can significantly reduce the agricultural non-point sourcepollution in the study basin.