The Load Simulation Of Nitrogen And Phosphorus Non-point Source Pollution And Water Quality Assessment In The Three Gorges Reservoir

The decrease of water quality caused by water pollution has become a global concern of environmental pollution.For instance, non-point source pollution, one of the main causes of water environmental pollution and eutrophication, has become a key problem that has restricted national and regional developments. The water environment of the Three Gorges Reservoir Area（TGRA）, which has very special geographical features, is greatly endangered by over-fertilization and serious soil erosion, backwater effect of dam construction and immigration and development in the area. Despite the generous national and regional investments in soil and water conservation, non-point source pollution problems, especially those caused by nitrogen and phosphorus used in agriculture, still exist in the TGRA in the process of construction and development, due to its topographic inequality, increasing usage of fertilizers（pesticides） and the coercion of large engineering projects in the region. These problems have become the ―bottleneck‖ of the reservoir‘s safe operation and economic development. Therefore, focusing the TGRA as its object, this study: Firstly, estimates the five periods(1990-2010（1990, 1995, 2000, 2005 and 2010） of the reservoir sediment load and adsorption nitrogen phosphorus pollution load by using large scale regional sediment transport model, application of the existing soil erosion model, sediment load model and adsorption nitrogen phosphorus pollution load model, taking into account the influence of sediment transport power coefficient and drag coefficient. Secondly, based on the Export Coefficient Model（ECM） of the rainfall and topographic were introduced the contributing index（CI） and vegetation bandwidth and slope into ECM in order to make the construction of the retention index（RI）, Using respectively both indices as weight factors to correct export coefficient for land use in literatures, and then to build an improved export coefficient model, with their pollution load of dissolved nitrogen and phosphorus for land utilization in 1990, 1995, 2000, 2005 and 2010 was simulated. Thirdly, constructedlandscape resistance-force coefficients of non-point source pollution, and finally established landscape resistance-force cost models to identify source-sink pattern of non-point source pollution by embarking from the landscape resistance and force which could affect large scale non-point source pollution and fusing various landscape essential factors（land use, soil, hydrology, the topography, vegetation cover, etc.）. Fourthly, made water quality assessment of the inflow and outflow of the reservoir by analyzing annual and seasonal average statistics of three water quality indices, applying model of TOPSIS（Technique for Order Preference by Similarity to an Ideal Solution） based on the gray correlation coefficient matrix, and explored the future trends of water quality of inflow and outflow by applying R/S analysis. The results show that:（1） The motivation coefficient affecting the sediment delivery ranged from 0.4 to 0.8 in the Three Gorges Reservoir area.Moreover,its spatial heterogeneity was not significant（P>0.05?）. However,the performance of resistance coefficient was complicated. The parallel ridge and valley zone,river alluvial gentle slope zone and mesa zone had high resistance coefficient,due to low gradience of slope,while the northern zone of Qinba Mountain area,and the Alpine and the Valley zone of Wuling Mountain area possessed low resistance coefficient due to steeper slopes.The sediment delivery ratio was in approximately normal distribution in the Three Gorges Reservoir area with a single peak.Its mean value was 0.48.It exhibited significantly spatial heterogeneity.The ratio of sediment delivery in the middle and western parallel ridge and valley zone was much smaller than the Qinba Mountain and Wuling Mountain areas.At the same time,it had the tendancy of the gradient increase from the center of the river to the sides.Moreover,the change trend of mean sediment load was the same as the interannual variation of total sediment load in the Three Gorge Reservoir area.Simulation of the total sediment load（0.9t） versus the value of monitoring from bulletin（0.92t）was the most close in 1995,the relative error was-2.17%,next was-5.00% in 2000,the largest relative error was-57.64% in 2010;At the spatial patterns,the spatial distribution of the low value zone and high value zone of sediment load were relatively stable due to little change in mean sediment load during 1990-2010.The distribution ranges of the low value zone were very wide with a high concentration degree.The distribution of high value area was discreted and broken.（2） In addition, the adsorbed nitrogen and phosphorus load possessed on value year-on-year growth with soil erosion modulus and sediment load in the Three Gorge Reservoir area.The maximum values of the total load in 2010 reached 1.2 tons and 0.6 tons respectively for nitrogen and phosphorus, which had similar spatial distribution with an unbalanced characteristics.Adsorbed nitrogen and phosphorus load gradually presented decreasing tendency from east to west.The simulation result of adsorbed nitrogen load（R2=0.6667） was better than adsorbed phosphorus load（R2=0.4336）when their correlation analysis was made with monitoring data.The mean value of the simulation result of adsorbed nitrogen was bounded by 20 tons,there were 6 basins that greater than 20 tons concentrated on Middle and lower Yangtze River,particularly in pengxi river and modao stream and daning river and tangxi river basin. The division value of adsorbed phosphorus was 10 tons, the division values of daning river and tangxi river basin were greater than 10 tons,in the mean time, the adsorbed nitrogen and phosphorus load had spatial correlation within in the Three Gorge Reservoir area.（3） By improving the coefficient of the output of the model to estimate the dissolved nitrogen and phosphorus pollution load under the differences land use mode: Total load of dissolved nitrogen for the paddy field after improvement exceeded 10,000 tons yearly, but before improvement were less than 10,000 tons. The dividing line of total load of dissolved phosphorous was less or more than 1500 tons, and before or after improvement. The proportion of the total load contribution of dry lands was maximum in all land utilizations, before and after improvement all exceed 30,000 tones, but the incremental total load of before and after improvement was much less in the land than in the paddy field. The total load for forestry and glass land were greater before improvement than after improvement, and the annually total load of dissolved nitrogen for forestry after improvement were less than 4,900 tons, before improvement were more than 5,100 tones, and the dividing line of total load of dissolved phosphorous before and after improvement was 310 tones. The dividing line of total load of dissolved nitrogen for glass land before and after improvement was 3,000 tones. Total load of construction land was smallest, but was increased with time.（4） Landscape resistance coefficients of 16 watersheds reflected a tendency that in the whole more toward downstream direction of the storehouse district, the average value more increases, but instead of landscape force coefficients, therefore, spatial distribution tendency performed for inverse proportion relations between landscape resistance and force coefficients. Compared to landscape resistance-force coefficients curves, landscape resistance-force cost curves behaved more volatility, which was showed that linearly increasing or decreasing tendency more significant, which also was showed that landscape resistance-force cost was more likely to reflect the dynamic process of non-point source pollution, in taking into account of distance cost of non-point source pollutants migration.Due to the result that landscape resistance-force cost difference as small in the high altitude area of reservoir, especially in the dragon boat river basin of Hubei（the 15th） and Xiangxi Brook Valley（the 16th）,landscape resistance-force resulted in a difference averaged by less than 2000.Sink landscape accounted for a large proportion, so the role of ―sink‖ was strong, the effect of ―source‖weak, but not in the low altitude area of reservoir. The landscape resistance-force cost difference average was large, especially the Jialing river basin（the second） had the maximum value（2003.32）,so the role of ―source‖ was strong, the effect of ―sink‖weak.（5） The water quality category of inflow and outflow from reservoir along the Yangtze River mainstream in the TGRA lied betweenâ… and â…¡class among 2004 to 2014, two sections were allâ… class among 2004-2007 and 2012-2014, year of 2008 were all â…¡class, and two sections lied betweenâ… and â…¡class among 2009-2010.The average value of dissolved oxygen（DO） of inflow and out flow section in stem stream of Yangtze River was given first place to withâ… class water quality standard,a few years for â…¡ class water quality standard,with the time change, the volatility of dissolved oxygen index of inflow and outflow cross section existed significant characteristic of stages, and water stage characteristics. The years of average value of Permanganate index(COD_Mn) and ammonia（NH₃-N） that deviated â… class water quality standard were decreased, the number of years nearly â…¡ class water quality standards were increased, the water quality of inflow reservoir was worse than outflow,and as time went on, the water quality of outflow reservoir section gradually became better.（6） The relatively close to the water quality of inflow and outflow section in four quarters each year swinged between I and â…¡classes, but more partial toâ… class, the trend was most pronounced in the years on average, which DO index of the inflow section season average performance in 2004-2014 for middle period（2008-2010）was less than before（2004-2007）, and after two periods（2011-2014）,In addition to the second quarter of 2006 and the first quarter of 2007, the rest of the year in the cross section of the outflow DO index of season average fluctuated in classâ… of the water quality standard. The COD_Mn index of inflow section in the second quarter of 2004（4.8 mg/L） surpass than â…¡class water quality standard,the COD_Mn indexes of outbflow section in 2004-2008 season average in almost all swinged betweenâ… classes and classâ…¡water quality standard, COD_Mn index of the quality of the water started to improve after 2009. The NH₃-N index of inflow section on season average in the first quarter of 2004 and the first and second quarter of 2005 reached only â…¢ class water quality standard, the season trend of NH₃-N index distribution in the outbound cross section of the interannual variability of average in 2004-2014 was similar to the COD_Mn index, the difference was classâ… water quality standard delayed until 2012. The total phosphorus content of inflow section was higher than outflow section, on the time change, when the water storage of the three gorges reservoir area did not reach 175 m in 2009, the total phosphorus contents of inflow and outflow section showed a trend of decline, the water quality standard from class â…¢ up to class â…¡, the total phosphorus content bounced back after 2009, the water quality standard from class â…¡ was changed to class â…¢. Overall, the water quality of two sections in 2005-2007, 2009, 2010 and 2013 in the four quarters presented a opposite trend.On the average annual change, the water quality of inflow and outflow cross section of three gorges reservoir area of the future trend on the average annual change was well consistent with the past, and showed the long-range dependence of the strong and persistent. The hurst index of inflow section was 0.668, and outflow section was 0.6201, they were greater than 0.5, but far away to 1, the change trend of the water quality in the future was not stable,with a certain randomness.Overall, the simulation precision of improved export coefficient model had greater improvement in this paper, and it could more precisely simulated spatial distribution of dissolved nitrogen and phosphorous for land utilization by the characterization of contributing index and retention index, and the increased range of simulation precision of dissolved nitrogen was larger than dissolved phosphorous. The source-sink pattern was divided into three types in this paper by source-sink landscape pattern index. Through the correlation analysis among landscape resistance-force cost difference, source-sink landscape pattern index and the average concentration of non-point source pollutants, it was showed that landscape resistance-force cost model applied in source-sink pattern recognition had certain of feasibility in the process of large scale non-point pollution, source-sink landscape pattern index which combined with the spatial location and landscape resistance-force coefficient could better depict spatial characteristics of non-point pollution. In the cross section of the reservoir area of the Yangtze river main stream in four quarters the future trend of the water quality changing showed from backward to forward sustainable transition to the past, but each section water quality are the future trends to weaker than average degree of future trends, and the future of the outflow section water quality change trend in the four quarters showed the positively and negatively continuous alternation with the past. The future trend of four quarters water quality changing in the cross section of the reservoir area of the Yangtze river main stream showed sustainable transition from backward to forward to the past, but the future trends to water quality of quarter average in each cross section were weaker than the average degree of future trends, and the water qualityof future trend changing in outflow section in the four quarters showed the transformation in positive and negative continuous alternating with the past.