Numerical Study On Water Residence Time And Water Quality Of Dahuofang Reservoir

With the rapid development of the national economy,the increasing human activities rely on the freshwater resources of reservoir more closely,and the pressure on the river basin environment eventually converges in the reservoir.Therefore,it is of great significance to study and master the material transport processes in reservoirs for water environment management and sustainable utilization of water resources.In this paper,the material transport and its residence time in reservoirs are studied numerically by taking Dahuofang Reservoir as example.In order to simulate the intertidal zone and floodplain areas,this paper adds the moving boundary treatment technique of wetting and drying scheme in the framework of the 3D hydrodynamic model ECOMSED.The improved model is verified by the free oscillation of water in a parabolic basin.The model computed free water surface,longitudinal,lateral and vertical velocity are in good agreement with the analytical solution.It laid the foundation for the further study of this paper.Based on the improved ECOMSED model,a 3D hydrodynamic model of Dahuofang Reservoir is bulit.The simulated water level and water temperature are in good agreement with the observed values,with the Root Mean Squared Error(RMSE)between the simulated and observed water level at the dam site is 0.459 m.The model successfully reproduces the variation of water temperature vertical structure of the reservoir within nearly one year.Then,the temporal and spatial characteristics of water residence time of Dahuofang Reservoir are studied,and the effects of inflow and outflow,heat flux and wind are analyzed systematically.The results show that inflow and outflow is the decisive factor for the water residence time of reservoir.The stronger the inflow and outflow is,the shorter the water residence time.Heat flux is the controlling factor of the water residence time of reservoir with water temperature stratification.The horizontal density circulation caused by heat flux makes the water residence time distributing more uniform,but the water temperature stratification caused by it enhances the difference of water residence time in the vertical direction.The influence of wind on water residence time of reservoir is the minmum,but it could strengthen the horizontal convection and vertical mixing of the reservoir water and finally change the distribution of water residence time.The above material transport studies are based on general water circulation theory(using conservative tracers to reflect water transport and mixing processes)rather than specific pollutants,so we combine the improved ECOMSED model and RCA model to establish a 3D hydrodynamic and water quality model for Dahuofang Reservoir.The simulated values of dissolved oxygen,ammonia nitrogen,nitrate nitrogen,total nitrogen,total phosphorus andchlorophyll a are in good agreement with the measured values.Then,the temporal variation and spatial distribution of the water quality variables are analysed,and the effects of nutrient inputs on the water quality of Dahuofang Reservoir are systematically studied.The results show that when the input of nitrogen and phosphorus in the rivers are reduced by half,it could significantly increase the concentration of dissolved oxgen in summer,reduce the concentration of nitrogen,phosphorus and chlorophyll a in the reservoir.The phosphorus input of Suzihe River contributes more to the total phosphorus and phytoplankton growth at Fushun Intake than that of the Hunhe River.For the complex mechanism of heavy metal migration and transformation and the fixing difficulty of the deterministic model parameters,an artificial neural network(ANN)simulation model for heavy metal transport in Dahuofang Reservoir is established.The maximum relative error between modeled and observed values of Cd and Cu are 17.5% and17.9%,respectively;the Determination Coefficient of modeled and observed values for Hg and Zn are 0.741 and 0.762,respectively,which even reach 0.96 for Cd and Cu.The results show that the simulated values are in good agreement with the measured data,and the neural network compute quickly and accurately.