Spatial-Temporal Distribution And Reduction Of Nitrogen Pollution In A Typical Agricultural Watershed

With the intensification of human activities,nitrogen pollution in water bodies have become increasingly serious,leading to a series of environmental problems,such as water quality deterioration and eutrophication of water bodies.Water nitrogen pollution has gradually become one of the environmental problems that need to be solved urgently.Especially in the watershed dominated by intensive agriculture,due to the influence of agricultural activities such as fertilization and livestock and poultry breeding,the river water is usually characterized by high nitrogen load and large spatial-temporal variability of nitrogen loss.Therefore,on the basis of ensuring food security,using models to study the spatial-temporal distribution of nitrogen pollution in agricultural watersheds and related reduction measures is important to ensure sustainable environmental development,which is of great significance to the management of nitrogen pollution in watersheds.As a mechanism model,SWAT model can well simulate the migration and transformation process of nitrogen pollution in the watershed,and can be used as an effective tool for simulation and assessment of nitrogen pollution in the watershed.However,the carbon and nitrogen（C-N）cycle module of SWAT model ignores the dynamic process of decomposition and transformation of soil residues and organic matter,which may have a significant impact on the simulation results of soil nitrogen content in farmland with more residues.Some foreign studies have optimized the C-N cycle module,and the optimized SWAT-M model fully considers the dynamic changes of residues and organic matter.The model has good applicability in the simulation of agricultural watershed,but the model has not been applied in China.Therefore,in this study,the Jurong Reservoir agricultural watershed,which is dominated by intensive agriculture,was taken as the research area,and the preliminary construction of SWAT-M model was completed by collecting and establishing the spatial and attribute database required by the model and modifying the corresponding source code.Based on the observed hydrology and measured nitrogen load from January 2008 to September 2012 and 2021,the SWAT-M model was calibrated and validated on a monthly scale.The temporal-spatial distribution and key sources of nitrogen pollution in this watershed were further analyzed according to the simulation results of the calibrated model.On the basis of this,the year 2021 is taken as the base scenario,and total of8 management measures are simulated by setting three scenarios:scenario FM（fertilization management）,scenario VF（vegetation filtration zone）and Scenario UB（urbanization）,and the reduction rate and effect of each scenarios are evaluated and analyzed.The main conclusions are as following:（1）The model evaluation indexes R² and NSE of runoff,nitrate nitrogen（NO₃^--N）and total nitrogen（TN）from January 2008 to September 2012 are all above 0.77,far higher than the model requires,0.6（R²）and 0.5（NSE）.The simulation results were delightful.In the validation results for 2021 based on data from May 2021 to November 2021,the runoff,NO₃^--N and TN validation results R² and NSE of the three sampling points are above 0.65 and 0.57,respectively,which all meet the requirements of the model.The results show that SWAT-M model has good applicability in the Jurong Reservoir agricultural watershed,and can be used to simulate nitrogen pollution in the watershed.（2）On the time scale,the nitrogen load to the reservior changed with the change of runoff.From 2008 to 2012,the annual nitrogen load to the reservior showed"M"type fluctuation trend,and reached the maximum when the precipitation was the highest in 2009.The results from 2008to 2012 and 2021,the concentration of nitrogen load to the reservoir with precipitation mainly occurred in flood season（June to September）,and the nitrogen load in flood season accounted for more than 69%of the whole year.At the spatial scale,compared with the results of 2008-2012,the total nitrogen load in runoff decreased in 2021,but the overall spatial distribution of NO₃^--N and TN loads in the two periods were consistent,showing the distribution characteristics of"local concentration,adjacent to water body",mainly concentrated in areas with high proportion of farmland.The main source of TN load was farmland fertilization,which contributed more than 49%of the total TN load in these two periods.（3）Among the three scenarios,there is a strong linear relationship between fertilizer application amount and TN load in scenario FM.For each 10%reduction of fertilizer application amount,the reduction rate of TN load to watershed increased by 0.78t and the reduction rate increased by 4.37%.It is proposed that the appropriate value of this scenario measure should be between 10%and 20%.In scenario VF,the width of the vegetation filter zone and its corresponding reduction rate shows a nonlinear growth relationship.The effect is best at 0.5m of the vegetation filter zone in the early stage,following tends to be stable in the last stage.After comprehensive consideration,it is suggested that the optimal width of vegetation filter zone in the study area should be set at 0.5 m.The results of scenario UB show that the reduction in TN of urbanization can reach 0.95t,and the reduction rate can reach 5.30%,however,it maybe need a long time.In this study,the applicability of the improved SWAT-M model in the Jurong Reservoir agricultural watershed was verified by measured data,and the spatial-temporal distribution characteristics and key sources of nitrogen pollution in the watershed were analysed.Moreover,the applicable nitrogen reduction measures were screened out through scenario simulation.The results of this study can provide a theoretical basis for for model simulation and monitoring of nitrogen pollution in intensive agricultural watershed.