Study On The Impact Of Land Use / Cover Change On Agricultural Non - Point Source Pollution At Different Slopes Of Dianchi Lake Basin Based On GIS And SWAT Model

Dianchi Lake, mother lake of Kunming people, has great significance for peopleof Kunming. This lake also is guarantee of booming development of Kunming, whichis a pearl on the Yunnan-Guizhou Plateau. With the economic development and therapid progress of urbanization, the rapid expansion of Kunming urban area, andincreasing of pollution into lake.Finally, resulting deterioration of water quality inDianchi Lake. With the deterioration of water quality in Dianchi Lake, this conditionhas caused the attention of central leadership. In recent years, the point sourcepollution has been got better governance，non-point source pollution increasinglybecome one of the important goals of Dianchi pollution control.LUCC(Land use and land cover change) is one of key research questions ofIGBP(International Geosphere-Biosphere Program) and IHDP(International HumanDimensions Programme on Global Environmental Change), which has become a hotissue with the cutting-edge research on the current domestic and international. Manystudies show that LUCC will affect runoff, sediment and non-point source pollutiondirectly. In this background, we use GIS and SWAT model to research non-pointsource pollution quantitatively in different land use scenarios, and design various landuse scenarios under different slope for diverse purpose. Finally, we try to explore themore optimization land use structure. The main contents and results are as follows:(1) Downloaded2006and2013TM Remote Sensing Images from the USGS(United States Geological Survey), supervised classification and supplemented withvisual interpretation of the data using software ENVI4.8.According to the2007national "land use classification" standard, combined with the actual situation ofagricultural non-point source in Dianchi Lake Basin, this paper classified the DianchiLake Basin land as forestland, grassland, agricultural land, urban land, bare land andwater. The results Kappa coefficients were0.75,0.78and0.74, accuracy able to meetthe model runs.(2) Analysis and statistics of land use changes from2006to2013suggested that urban land and forest land increased and farmland, grassland, bare land and watershowed a decreasing trend. This paper analyzed land transpose after obtaining thetransfer matrix of land use of2006to2013.Using ArcGIS spatial analysis functions,the Dianchi Lake basin was divided into five slope sectors:0°~2°,2°~6°,6°~15°,15°~25°,>25°. Statistics area of different land use types under different slopelevels.(3) After collecting and collating meteorological data, soil, hydrology,agricultural management data, slope data and land use data in Dianchi Lake Basin,this paper established the necessary foundation database of running the SWAT model.Then combined with GIS hydrology analysis functions to divide sub-basin, and runSWAT model. Sensitivity analysis of model by using SWAT-CUP SUFI2algorithm, itidentified sensitive parameters which had greater impact on model results. I usedmeasured hydrologic data of Kunming hydrological station and Ganhaizihydrological stations to calibrate and validate for the model. Calibration results metthe required model running.(4)Changing land use structure in different slope, and designing4land usescenarios based on2013land use structure. The first scenario is convert arable land towoodland which on the slope greater than25degrees, and convert arable land tograssland which on the slope between15degrees and25degrees，others remainunchanged. The simulation results show that total nitrogen reduced to30.21%andtotal phosphorus reduced to13.36%. The second scenario is convert bare land andgrassland to arable land which on the slope less than15degrees, other remainunchanged. The simulation results show that total nitrogen increased to50.38%andtotal phosphorus increased to45.62%. The third scenario is convert grassland toarable land which on the slope greater than15degrees, and convert woodland toarable which on the slope between15degrees and25degrees. The simulation resultsshow that total nitrogen increased to89.2%and total phosphorus increased to88.94%.The fourth scenario is convert arable land and bare land to woodland which on theslope greater than15degrees. The simulation results show that total nitrogen reducedto34.79%and total phosphorus reduced to22.84%. All of the results indicate thatwoodland and grassland can reduce the output load of total nitrogen and total phosphorus effectively, but arable land can increase the output load of total nitrogenand total phosphorus, particularly on steep slope areas this trend will become morepronounced.