Non-point Source Pollution Analysis Of Liangzi Lake Based On SWAT Model

In order to control non-point source pollution effectively,it is necessary to have a competent knowledge of the risk of pollution.Model simulation and prediction analysis are the main meathods to know the risk of pollution.Based on the SWAT model,Liangzihu Basin is used as an example to simulate and predict the risk of non-point source pollution in a basin.The main results are as follows:?1?Based on the remote sensing image of DEM,Landsat7 and Landsat8,basin outline,watershed and landuse type are generated by ArcGIS software,with 42 subbasions divided.As there is little change of landuse in last decade,which is based on the analysis results from2006 to 2015,the impact of landuse change can be ignored in this SWAT model.The losing meteorological data is complemented and the file type is exchanged by Matlab software.The remote sensing images of DEM,Landsat7 and Landsat8 are all from Geospatial Data Cloud.Soil data is from Harmonized World Soil Database.Meteorological data is from China Meteorological Data Web and Hubei Provience Meteorology Bureau.Agricultural management data is from investigation and references.Reservoir data is from remote sensing images and references.Crops growning and fertilization data is from investigation and references.The physical attributes parameters'values of soil are calculated from SPAW software.The area and volume of the lakes both calculated as they are elliptic sinosoids.?2?Database is built with ArcSWAT software,and 481 HRUs are devided in Liangzihu Basin according to landuse,soil and slope.Based on the soil type,slope and the crops'growning characteristics,trillage and fertilization,crops are grew in the model.January 1,2003 is seted as starting data of the model,and 13 years are simulated with warm-up period 3 years?2003 to2005?,calibration period 6 years?2006 to 2011?and validation period 4 years?2012²015?.Observed runoff?from Fankou hydrology station?and concentration of TN,TP?from water quality monitor station of A and B?sampled monthly from 2006 to 2015 are used as calibration and validation by SWAT-CUP software.Digital filter technology is used to devide runoff into baseflow and surface runoff,3 statistical parameters values(Re<20%,R²>0.6andE_ns>0.5)tested in calibration and validation are all acceptable.After the unit of TN and TP concentration is transformed,3 statistical parameters values(Re<30%,R²>0.6 and E_ns>0.5)tested in calibration and validation are all acceptable.So we consider the SWAT model is believable.?3?The yearly precipitations of Wuhan meteorological station from 1994 to 2013?2 decades?are used to meet the Pearson III frequency curve,with the year of 2013,2008 and 2011 are selected as high flow year,normal flow year and low flow year.Compared with the N and P loads in typical hydrology years,N loads in high flow year is the highest one,then normal flow year,and low flow year lowest;P loads in high flow year is the highest one,then normal flow year and low flow year.The seasonal difference is significant of N or P loads,with high loads in Summer and Winter,low loads in Spring and Fall.There is relative high pollution of N and P in subbasins 6,7,8,9,14,16,17,23 and 24,where there are relative large scale of cotton and tea growing.Organic nitrogen accounts for a substantial part of N loads in all of high flow year,normal flow year and low flow year,while mineral nitrogen is more in north basin than south.Mineral phosphorus attached to sediment accounts for a substantial part of P loads,then organic phosphorus,and soluble mineral phosphorus is relative low.?4?Some suggestions are made such as contour tillage,reasonable fertilization of green manure and pricision agriculture based on the results of this SWAT model analysis.Through improving ecological and environmental protection of Liangzihu Basin,reducing the use of quick-acting fertilizer,non-point pollution in Liangzihu basin could be controlled.