Study Of Best Management Practices Of Non-Point Source Polloution Based On SWAT And Information Entropy

At present, Water pollution has been a topic of concern widely. People have gradually recognized the important role of non-point source pollution of surface water pollution. The basic research of water circulation is theoretical foundation to understand the causes of non-point source pollution of watershed. Best Management Practices play an effective role in solving the problem of non-point source pollution. In support of GIS technology, SWAT model has been an effective means of NPS simulation and best management practices through improvement over the years. It has a very broad application prospects.This thesis selected the SWAT model as research tools to evaluate the NPS pollution output of Liaohe River upstream watershed. Using GIS hydrologic analysis module, We established spatial and attribute database database, analyzed model main parameters'sensitivity, calibrated and validated the model using measured hydrological data in 2005-2014, and then completed NPS polution model construction. After that, we simulated the NPS pollution load of Liaohe River watershed from 2005 to 2014. Then we have analyzed the distribution rule of inter-annual and intra-annual production of surface water, sediment loss and total nitrogen (TN) and total phosphorus (TP) pollution load in 2005-2014, as well as the spatial distribution based on the annual means. What is more, we discussed the difference in NPS pollution load among different land uses. The thesis set up 11 scenarios of best management practices, the scenarios are including no tillage, stubble mulch tillage, contour tillage, strip tillage, contour hedgerow, grain for forest, grassland for forest, riparian buffer strips, then we calculated the reduction of sediment and TN and TP. Taking into account the economic costs of the various measures, we obtained comprehensive evaluation through multiple attribute decision based on information entropy. The main conclusions are as follows:(1) While building SAWT model of Liaohe River upstream watershed, we selected determination coefficient (R2) and Nash coefficient (NS) to evaluate the performance of model. Finally, we find R2 and NS which can meet the evaluation criteria of the model.(2) From 2005 to 2014, NPS pollution load has obvious inter-annual variation in Liaohe River. There is abundant rainfall from April to September. the NPS pollution load is also distinct within these months. Thus, we should taken measures to prevent soil erosion during this period. TN load increments from upstream to downstream with high spatial differences. Soil erosion are related to slope gradients and soil types. TN load are closely related to agricultural production and human life. Sediment is the carrier of phosphorus Migration. Reducing the amount of soil loss can be very good control of the amount of TP into the river. Spatial distribution of TN and TP load are quite different, and TN load are higher than TP load. The descending order of NPS pollution load is:arable land> paddy> residential district> grassland> forest> shrubland. The main source of NPS pollution load is arable land, Non-point source pollution control measures should be set up mainly for arable land.(3) We consider NPS polution reduction and economic costs of management measures. For small-scale management measures, "grain for forest above the slope of 15° " and "grassland for forest above the slope of 15°" can control NPS pollution well. Their comprehensive evaluation value is highest. For a wide range of management measures, "contour hedgerow" is best of sediment reduction, and "stubble mulch tillage" can effectively reduce the TN and TP loss. In addition, "strip farming " are good for NPS pollution effects. It have low cost and a higher overall evaluation value. We recommend that different BMPs should be applied according to local conditions. We can also integrate different schemes in order to achieve practical application of different management measures.